A people leader's job isn't to make you happy | Katie Burke (COO, Harvey)

tldr

Katie Burke spent 11 years as HubSpot's Chief People Officer, building one of tech's most distinctive cultures, then joined Harvey as CPO and now COO. Her core thesis: the people leader's job is NOT to make employees happy — it's to create an environment where they can do the best work of their career. She covers treating internal comms as a marketing problem, the two-year Layoff Hangover that follows every reduction, Berry-Gate as a masterclass in feedback filtering, Frequent Flyers as a hiring red flag, AI rewriting the people function right now, and sharing your own Performance Review as a trust-building mechanism. Dense, practical, no-nonsense — the kind of people leader every founder wishes they had.

Key Takeaways

Internal comms is a marketing problem, not an HR one. Employees aren't a captive audience — your messages compete for attention the same way your product competes for customers. Katie Burke hired from hospitality, not corporate HR, to get this instinct.
The resort has to match the brochure. The goal isn't employee happiness — it's creating an environment for career-best work. Be honest in hiring about what the job actually looks like, then deliver on that promise.
Layoffs are a two-and-a-half-year cultural decision, not a 90-day business one. The Layoff Hangover unfolds in waves — shock, survivor guilt, lingering fear — and poisons risk-taking long after the event.
Be Michael Jordan, not his friend. Great exec teammates are selfless under pressure and visibly outwork everyone. The test: would you pass the ball to Steve Kerr in the final seconds?
Separate protein from sugar in feedback. Substantive complaints about things that block winning matter. Berry-Gate taught Burke that naming entitlement openly rallies your best people.
Frequent Flyers are the most dangerous employees. They don't just waste their own time — they recruit others into caring about things that don't matter. A new hire who sends 18 complaints on Thursday of week one will be gone in three to six months.
Employee happiness isn't predictive. "I see myself here in 12 months" is. That single survey question was the only leading indicator HubSpot found across years of trying.
AI is rewriting people ops right now. If you're not asking "how could AI solve this first?" for every process, you won't be a successful people leader in 18 months. Performance Reviews completed in 5 minutes. Onboarding compressed to 3 steps.
Share your own performance review with the team. Burke has done this for eight consecutive years. If the CEO is openly listing things to improve and you can't name one, that's a self-awareness problem.
Negotiate how you disagree, not just what you disagree on. Burke and Winston Weinberg had to explicitly work out conflict mechanics — she needs time to process, he wants to resolve immediately.

[00:05] The People Function as a Marketing Problem

Katie Burke opens with a reframe that most Chief People Officers miss: internal comms is a marketing problem, not an HR one. Her background as a marketer at HubSpot made her think in terms of share of attention — employees aren't a captive audience you can bore into compliance. They choose how to spend their attention, and your internal messages need to compete for it the same way your product competes for customers.

quote

"Most traditional HR leaders think we have a captive audience, so how can we drone on and on with as much information as possible to put people to sleep because they have to listen." — Katie Burke

Three moves she made differently:

Hired from hospitality — restaurants and hotels, not corporate HR. People from service backgrounds think in terms of guest experience, not process. During the pandemic, her team stood up an online Montessori school for employees' kids within three days because hospitality-trained people default to "what does this person actually need right now?"
Treated Employer Brand like product marketing — most employer brands are undifferentiated. She applied the same rigor to candidate messaging that a PM would apply to positioning.
Refused the "seat at the table" framing — instead of lobbying for influence, she built the table and set the menu. Employee programs had to ladder directly to business strategy, not exist as standalone HR initiatives.

quote

"Don't ask for a seat at the table — build the table and set the menu." — Katie Burke

On Radical Transparency post-IPO: when HubSpot went public, lawyers and bankers advised retiring internal information-sharing. Burke did the opposite — made every employee a designated insider and kept sharing at the same velocity. They also published their entire eNPS survey — every comment, including ones that said "Katie was terrible at that company meeting." She introduced a Failure Forum where executives publicly dissected real failures — not "I work too hard" humblebrags but genuine flops.

[05:05] The CPO as Strategic Driver — and the Trust Tightrope

Brett Berson pushes on how the people function becomes a strategic input to hitting revenue targets. Burke breaks it into three layers:

Recruiting for execution + values — not just skills, but people who understand urgency and live the company's operating principles.
Manager enablement — the real leverage isn't the all-hands meeting. It's whether managers are repeating that information in their own team meetings using the same slides.
Succession Planning as strategy, not insurance — stop thinking "what if someone gets hit by a bus" and start thinking "what leadership capabilities do we need in a year that we don't have now?"

The trickiest part of the CPO role: you are the HRBP for the executive team itself. The currency is Trust — destroyed the moment you share someone's private feedback with the board or CEO.

quote

"You cannot take a momentary opportunity to seize power at the risk of trust." — Katie Burke

She drops a line worth remembering: "I don't trust teams where they say we always agree. We're a family." High-performing teams disagree openly. They're a high-performance sports team, not a family.

[10:01] Be the Michael Jordan of Your Exec Team

Katie Burke draws a sharp line between being a great teammate and being a great friend — and argues execs should optimize for the former. Michael Jordan was famously brutal in practice. He agitated people. He pushed relentlessly. But he was loyal, and he worked harder than anyone else on the floor. That combination — selflessness under pressure plus visible work ethic — is what separates a demanding exec from a toxic one.

The test she offers: Would you pass the ball to Steve Kerr in the final seconds? If you're willing to give up the spotlight when stakes are highest, people will follow you into difficult territory.

quote

"He was a better teammate than he was a friend by far. And I think that's part of what your job as an exec is — to be a great teammate." — Katie Burke

When a peer exec isn't performing, Burke's instinct is to go direct first — one-on-one, specific, no ambiguity. Three buckets: your team doesn't trust you, your results aren't there, or you're not showing up as a leader in the exec room. If the pattern persists, escalate. But the first move is always candor, not politics. Not every exit needs to be a fiery exit. People can be excellent leaders — just not for this stage, this company, this moment. The exec team's job is to create as many Graceful Exits as possible.

[12:19] Graceful Exits and the Transparency Trap

Burke reframes transparency not as "everyone gets all information" but as "everyone gets information that is safe for the company to share." When an exec is let go, the why is not the company's information to broadcast. Two risks: you damage someone's ability to get their next job, and you create a precedent where silence in future exits triggers worse speculation.

quote

"Let's pretend that's you. How would you want us to treat you?" — Katie Burke

Her approach at HubSpot: be consistent. Every exit is handled with discretion. When people see you apply the same standard every time, they stop reading into individual cases. For a small team, this matters more, not less. Having a consistent, dignified protocol for exits protects the remaining team's trust.

[16:47] The Layoff Hangover: A Two-Year Cultural Tax

HubSpot's 2023 layoff taught Burke something she initially didn't believe: the cultural aftershock lasts two to two and a half years. Not months — years. The board warned her. She thought her org was too agile for that. She was wrong.

The hangover unfolds in three waves. First, the immediate shock. Then survivor guilt. Finally, the lingering fear that another round is coming, which poisons decision-making and risk-taking long after the event itself.

quote

"It's not just a one-time business decision. It's a cultural decision that has implications for many years to come." — Katie Burke

Burke's tactical takeaway: go in eyes wide open that a layoff is a two-and-a-half-year process, not a 90-day one. Plan Workforce Planning accordingly — during hypergrowth, think about where you can flex with contractors instead of committing to headcount you might need to cut. Every head you add also adds ownership complexity. If two people own anything, the chances it gets done well drop dramatically.

At Harvey, she runs an annual financial plan but does quarterly replans. An annual headcount plan built in January is stale by March — especially in AI.

[23:31] The Resort Has to Match the Brochure

Katie Burke draws a sharp line between two failure modes in employee experience: the Disneyland era of the 2010s, where companies believed their job was to make people happy every day, and the overcorrection happening now — return-to-office mandates, stripped perks, a blunt "work should be hard" posture. Both are wrong. The goal isn't happiness. It's creating an environment where people can do the best work of their career.

quote

"Our job is not to make you happy every day. Our job is to create an environment where you can do the best work of your career." — Katie Burke

The test she uses: does the resort match the brochure? At Harvey, the brochure says intense but reasonable. You won't work 9-to-5, and nobody apologizes for that. But if you need parental leave or a doctor's appointment, you take it without guilt. The interview process is designed to make this crystal clear so nobody walks in surprised.

She extends this to Amazon as a counterintuitive example of good company culture. Amazon pits internal teams against each other — and they're transparent about it in hiring. That's functional. What's dysfunctional is advertising unicorns and rainbows, then running a cutthroat shop behind closed doors.

On severance, Burke's litmus test: one founder told her, "I take care of the people who work here, not people who don't." She pushed back: people took a bet on you, and how you treat departing employees creates a Talent Flywheel — those people talk.

[27:21] Incentives, Titles, and the 20% Who Will Always Be Unhappy

Burke invokes Charlie Munger on Incentive Design: keep it ruthlessly simple. Complexity doesn't prevent manipulation — it enables it. She plays out failure modes with the board for every bonus structure: top line, bottom line, what does gaming look like, which trade-offs are acceptable.

A less obvious insight on manager incentives: most managers don't optimize for themselves. They optimize for their teams. This means promotion calibrations get inflated not by selfishness, but by loyalty. Her fix: before each Performance Review cycle, she pulls up blinded reviews from the previous round and asks leaders to guess the rating.

quote

"Executives who say titles don't matter are delusional. They do matter to people, and so does the opportunity to grow." — Katie Burke

On titles at Harvey: no SVP layer at 650 people, rough guides instead of rigid ladders. She warns that candidates who lobby hard for title during negotiations will be exactly that difficult once they're on board.

The CPO reality: at any given moment, roughly 20% of people are unhappy with you. The skill is distinguishing valid complaints about things that matter from noise about things that don't, and never losing the ability to hear the difference.

[35:06] Berry-Gate and the Art of Filtering Feedback

At HubSpot, Katie Burke used a framework she called "protein versus sugar" to sort feedback: substantive complaints about things that block winning versus noise about snacks and perks. The story that crystallized this was Berry-Gate — the finance team produced a chart showing berry consumption growing exponentially against headcount. Leadership held a serious meeting, then landed on what they considered an elegant fix: kill the fresh berries, open a smoothie bar. Smoothies named after customers. Health-forward.

Within hours, a Slack channel erupted. Employees hadn't been consulted. Someone pointed out the fiber content. Others demanded a town hall. Burke wrote a company-wide note: if you spent 45 minutes complaining about berries in Slack, you clearly don't have enough work to do. The most loyal employees rallied — they'd seen entitlement creeping in and were relieved someone named it.

quote

"We are going to make decisions that have winners and losers and we're going to make hard calls in the interest of scale." — Katie Burke

The skill is separating the person with one legitimate grievance from the Frequent Flyers who show up in every Slack drama.

[42:11] Frequent Flyers and New-Hire Red Flags

Burke's sharpest insight: frequent flyers don't just waste their own time — they recruit others into caring about things that don't matter. They indoctrinate.

She identified a remarkably specific leading indicator at HubSpot: if a new hire sent her a note on Thursday of their first week with 18 observations on what could be improved plus a list of personal requests, that person would last three to six months. Compare that to someone who writes, "My new-hire trainer was incredible, here are three things to keep in mind for next time" — same feedback, completely different orientation.

quote

"The people who bias to negative in their first week — typically that is the leading indicator of how they're going to operate." — Katie Burke

Other patterns she tracked: whether new sales reps actually listened during IT setup on day one, the ratio of complaining to fixing, and how often someone invoked "at my last company" rather than engaging with the current one.

[44:54] The Gap Between VP and C-Level

Katie Burke draws a sharp line between director-to-VP and VP-to-C-level, with different bottlenecks at each.

Director to VP: The failure mode is acting like a camp counselor. The tell is in the calendar: "How is your calendar different now than 6 months ago?" Wrong answer: more one-on-ones. Right answer: fewer one-on-ones and more systems that scale.

VP to C-level: Two things separate them. First, Self-Awareness — you start believing your own press. Second, the ability to recruit people who are better than you. Both are ego problems wearing different masks.

quote

"Your meetings should be more about accountability to the goal versus de facto managing their team. Most directors are just de facto managing the teams for the manager." — Katie Burke

[49:04] Hidden Landmines When Scaling the People Function

Three failure patterns as companies grow from 500 to 1,000+:

1. Program-centric vs. people-centric. Teams get locked into their roadmap and miss the signal. If 17 people from your top-talent list left while you were building a "top talent program," you've lost the plot.

2. Worshipping lagging indicators. Attrition is the classic trap — by the time six people leave, you already have a massive problem. The only predictive measure on HubSpot's survey: "I see myself at HubSpot in the next 12 months." Not employee happiness.

quote

"Employee happiness wasn't actually predictive. What was predictive was people saying, 'I see myself at HubSpot in the next 12 months.' That was the only predictive measure on our survey." — Katie Burke

3. Internal territorial politics. People teams are full of people-centric humans, but turf wars still emerge.

[51:23] HubSpot vs. Harvey — What Changed and What Didn't

Culture as product vs. culture as byproduct. At HubSpot, culture was the differentiator. At Harvey, Burke inverted this: the product and brand lead; culture follows. Winston Weinberg's company wants to be the leading AI platform for lawyers, not the company with the coolest culture deck. This isn't abandoning culture — it's sequencing it correctly.

Time compression. At HubSpot, you could go fast on some decisions and slow on others. At a company like Harvey in the AI arms race, that luxury doesn't exist. If models are getting 20% better, the org needs to get 60% better.

[53:00] AI Is Rewriting the People Function — Right Now

Burke is blunt: if you're a people leader dabbling with AI for personal use but not deploying it at work, you are already behind and will not be successful in the next 18 months.

What's already changed: Harvey uses its own product to generate job descriptions. A team member wrote a script that completes performance reviews in 5 minutes. Onboarding compression: "can we cut it to 3 steps, with AI handling the rest?" Headcount planning must now account for whether work units should be done by people or agents.

What's coming in months: personalized internal comms by team/level, AI-driven workforce planning, CPOs absorbing Chief AI Officer responsibilities because AI adoption is fundamentally a change management problem.

quote

"If you are not thinking 'how could AI solve this problem first' and eliminate hours of work within your team, you're not going to be a successful leader in the people function in the next year and a half." — Katie Burke

What won't change: EQ. If AI eliminates performance-review drudgery, managers should finally have time for mentorship and development. On being demanding: Burke rejects the false choice between high expectations and empathy.

quote

"You can be liked or you can be respected and you have to pick a lane." — Katie Burke

Katie Burke has shared her performance review with her team — elements redacted — for eight consecutive years. Logic: if your CEO and COO are openly listing things they need to improve, and you can't name a single thing you're doing wrong, that's a self-awareness problem, not a strength.

quote

"If your CEO and COO are sharing a long list of things they can do better and you can't think of one thing you're doing wrong, I think that's a self-awareness opportunity." — Katie Burke

Three things she's improved most over eight years:

Overtalking. After promotion to HubSpot CPO, she assumed the job was to say something important at every meeting. Feedback: "Your quality of value-add versus words per minute is going down."
Finger-pointing. When HubSpot fell behind on recruiting, she blamed every adjacent team. Brian Halligan gave her an ultimatum: own it and dig out, or go down pointing fingers.
Letting go of perfection. As an executive, your goal is to bat .400. Getting comfortable with misses without spiraling is a skill built deliberately.

quote

"You are either going to be here in a year because you own this and dug your way out of it, or you're going to go down pointing fingers at other people." — Katie Burke

[66:55] How to Disagree With Your CEO

Katie Burke outlines the mechanics of disagreeing with Winston Weinberg, Harvey's CEO.

Foundation: respect. Even in heated debate, both sides need to feel like they're on the same team. Critical distinction: she and Winston never disagree on where they're going — only on how to get there.

But the real insight is conflict mechanics. Katie needs time to process — a night to sleep on it. Winston would rather hash it out immediately. They had to explicitly negotiate how they disagree, not just what they disagree on.

quote

"We've had to work through how we disagree, not just what we disagree on." — Katie Burke

On decision rights: the CEO doesn't make every call. What matters is agreeing in advance on who decides what. Winston owns fundraising and board. Culture defers to Katie, unless Winston wants to "lay on the tracks." The remaining 10% — where neither the decision nor the decision-maker is obvious — those are the hardest and most important.

Their most common tension: pace. Katie pushes for patience; Winston pushes for speed. This is structural in every startup leadership pair.

On board management: ask for help (your board wants you to succeed), observe before you present, and never present something the board is seeing for the first time — pre-seed every big idea.

quote

"I want the people on my team to say there's no one who wants us to win more as a company and there's no one who's pushed me harder to get there while still having my back." — Katie Burke

People Mentioned

Katie Burke — COO of Harvey, formerly 11-year CPO at HubSpot. The guest.
Brett Berson — partner at First Round Capital, host of Executive Function. The interviewer.
Winston Weinberg — co-founder and CEO of Harvey. Burke's current CEO.
Brian Halligan — co-founder and former CEO of HubSpot. Gave Burke the "own it or go down" ultimatum.
Charlie Munger — Berkshire Hathaway vice chairman, cited on incentive design simplicity.
Michael Jordan — basketball GOAT, cited as the model exec teammate (great teammate, not great friend).
Steve Kerr — current Warriors head coach, former Bulls player; cited in the "pass the ball" test.

One Thing to Act On

Share your own performance review with the team. You run a small team at FeatureOS and SupportWire — this is one of the highest-leverage trust moves available at your scale. Write a short self-assessment: three things you did well this quarter, three things you need to improve. Share it in a team meeting or async doc. Burke did this for eight straight years and says it single-handedly changed the feedback culture on her teams. For a 5-10 person company, the effect is even more concentrated — when the CEO openly names their weaknesses, it gives everyone else permission to be honest about theirs. The "I see myself here in 12 months" survey question is also gold, but you can start the performance-review sharing this week with zero setup cost.

Raw Transcript

Auto-captions from YouTube. Folded by default — expand if you need to grep the source or pull an exact quote.

0:05 We'll maybe talk about this sort of last 0:06 little chapter as you kind of went from 0:09 a chief people officer to a chief 0:11 operating officer. But maybe to start 0:14 when you think about the work that you 0:16 do as a as a chief people officer. What 0:19 are the things that you think you do 0:21 most differently 0:23 >> as a CPO? I think a few things. One is 0:27 because I was a marketer before you 0:29 think about share of attention and that 0:31 applies to recruiting but it also 0:33 applies to employees. So most I think 0:35 traditional HR leaders think we have a 0:38 captive audience so how can we drone on 0:39 and on with as much information as 0:41 possible internally to put people to 0:43 sleep because they have to listen. And 0:45 what I would always tell my team is 0:47 people have a choice with how they spend 0:48 their day and spend their time. We need 0:49 to make sure our internal messages are 0:51 super compelling and interesting. The 0:53 same is true of candidates. most 0:54 employer brands are not that exciting or 0:56 differentiated and so I think some of 0:57 that is thinking about comms as a 1:00 strategic differentiator. Two was we 1:02 hired a lot from hospitality and that 1:04 was very intentional. I believe that 1:06 people coming from restaurants, hotels, 1:09 you name it do a better job at thinking 1:11 about the guest experience and employee 1:12 experience than someone who grew up in a 1:14 more traditional corporate environment. 1:16 And so what you saw was our team come up 1:18 with ideas that might be so as a good 1:20 example during the pandemic we were able 1:22 to stand up an online montatory school 3 1:25 days after the pandemic set in because 1:27 we knew that parents needed the most 1:29 support. We were trying to think outside 1:30 the box. Those are I think some of the 1:32 ideas that come to fruition when you 1:34 have people on your team who come from 1:36 different backgrounds. And then third I 1:37 would just say I think some of the 1:39 traditional HR rhetoric is about getting 1:41 a seat at the table. And I'm a big 1:43 believer in like don't ask for a seat at 1:45 the table, build the table and set the 1:46 menu. And so I think that attitude 1:49 seeped through my whole team, which is 1:50 like the business comes first. Our 1:52 customers come first. Employee 1:55 programming has to ladder up to the 1:56 business strategy first and foremost. 1:58 >> Talk more about what what it means to 2:00 actually build the table like what are 2:01 the stories that come to mind that that 2:03 convey that idea? 2:04 >> Yeah. So I think a few examples come to 2:06 mind. So at HubSpot we had five values 2:09 and the acronym was heart. And I think 2:12 people think of that as a big success 2:13 story. The reality is we had two 2:14 iterations before of failed values 2:16 launches as so many companies do because 2:18 we tried to make eight of them. We had 2:20 an acronym that didn't make sense or 2:21 resonate. And so once we had the five 2:23 values is about like how we actually led 2:25 them. And I think transparency is 2:27 probably the easiest one to talk 2:28 through. When we went public, there were 2:30 a series of things that any normal 2:32 company, especially advised by lawyers 2:35 and bankers, would of course retire. And 2:37 what we decided to do is make everyone a 2:39 designated insider and continue to share 2:41 that information at that velocity. 2:42 That's a total counterintuitive bet, but 2:45 the feedback we got from employees was 2:46 it felt like getting a mini MBA because 2:48 you got so much more exposure. So we had 2:50 a lot more information available to 2:52 people even as a public company. And I 2:54 think that if you were coming from a 2:55 very cautious background and you were 2:57 like, oh, we should follow the rules and 2:59 what every other company has done that. 3:00 We wouldn't do that. And I think that's 3:02 a type of transparency. We would also 3:04 release our entire employee net promoter 3:06 score survey. Every single comment, 3:08 unless there was something that was 3:09 harassing or problematic, every single 3:12 comment, including comments that said 3:13 things like Katie was terrible at that 3:15 company meeting, or I really disagree 3:17 with this program. And I think if you 3:19 wanted to build a team that was just a 3:20 line behind you, you certainly wouldn't 3:22 release that data. That's a good example 3:23 of like we wanted culture to be a 3:25 strategic priority. That means leaving 3:27 living your values even when it's really 3:29 hard. What are the other kind of things 3:32 from the outside seem very weird but 3:34 were definitely the correct thing to do 3:35 at least in that stretchma 3:38 >> during that stretch? So I think a few 3:39 things that we did one in the middle of 3:42 the pandemic we decided to give people 3:45 choice around where they work. So a lot 3:46 of companies did the return to work 3:48 we'll return to work in May or we'll 3:49 return to work in September. And you saw 3:50 them keep pushing that out. What we did 3:52 was we designed a menu of options so you 3:54 picked your work preference much like 3:56 you pick your benefits. And so we had 3:58 three options. home flex and office and 4:00 people opted in every year and your 4:02 stipens, your work allocation, your 4:04 desks, all that kind of stuff aligned 4:05 with that. That's a good example of 4:07 something that no one else is doing at 4:08 the time that they still benefit from to 4:11 this day. So that led to a ton of 4:13 recruiting velocity during a time when a 4:14 lot of companies were struggling to hire 4:16 fast. That was a big example of 4:17 something that allowed us to go faster. 4:19 I think a few other things were we would 4:22 just talk regularly. We had something 4:23 called failure forum and we would 4:25 regularly talk about things that we got 4:27 wrong including execs and it couldn't be 4:29 things like my biggest weakness that I 4:31 worked too hard. It had to be like this 4:32 was a real flop and here's what led to 4:34 it. So one of my favorite examples was a 4:36 VP of product got up and talked about a 4:38 customer launch that went really wrong 4:40 and it created a ton of downstream 4:42 effects for our support reps that 4:44 frankly just completely backlog them for 4:46 3 days. and he basically got up there 4:47 and said, "In addition to apologizing to 4:49 our customers, I want to say to our 4:50 support reps, we got this wrong and mess 4:52 things up for you, which I feel even 4:54 like equally bad about." I think that's 4:56 the kind of humility that people aspire 4:58 to and it creates disproportionate 5:00 loyalty to a company during an 5:02 incredibly tight talent market and that 5:04 stuff really matters. 5:05 >> Think about set any arbitrary goal. We 5:06 want to get to $500 million in recurring 5:08 revenue and we want this type of 5:10 retention and we want this type of usage 5:11 or some metric for value delivered to 5:14 customers. maybe talk a little bit more 5:16 about how the people function becomes 5:19 sort of a strategic input driver for 5:21 that. 5:22 >> Yeah. So, uh I think a few ways. One is 5:25 just the folks that you recruit. So, 5:26 people who are great at execution, who 5:28 live your values, who get the how and 5:30 also who can drive the urgency. Two is 5:32 managers. I think people don't talk 5:34 enough about the fact that managers are 5:36 the folks who set your set your tone and 5:37 experience especially as it relates to 5:39 those metrics. And so part of what a 5:41 great people function does is like you 5:43 get let's say you do an all hands 5:44 meeting every week, every month, doesn't 5:46 matter what the cadence is, the people 5:48 team runs. I actually think your success 5:50 is not just predicated on that. It's 5:52 predicated on to what degree are 5:54 managers repeating that information 5:55 regularly and using the same slides in 5:58 their team meetings to reinforce that 6:00 behavior. So I think a really good 6:01 people team recognizes that you want to 6:03 give people tools to actually advance 6:05 those mechanisms without you in the 6:06 room. And then the third thing I would 6:08 just say is on the leadership side of 6:09 things. Part of your job is just to be 6:10 thinking about the current leadership 6:12 team and the future leadership team. And 6:14 I think succession planning is viewed as 6:16 like what happens if someone gets hit by 6:18 a bus. And what it should actually be 6:20 doing is thinking about what's the 6:21 leadership skill set we need in 10 years 6:23 that we don't have now or in a year that 6:24 we don't have now and where are the 6:26 gaps? What's our leadership team really 6:27 good at now and what are we bad at now? 6:29 And how does our team like executive 6:31 recruiting should really help solve that 6:33 gap? 6:33 >> Maybe share more about that. Obviously, 6:36 if if you're a seale exec, your sort of 6:38 primary team ideally should be the 6:40 executive team. And it seems like at 6:41 least how you're describing it, one of 6:42 the unique dynamics of a of a chief 6:45 people officer is maybe they should be 6:48 much more mindful of everyone else. And 6:50 even if a high functioning executive 6:51 team, everybody's working on the 6:53 business problems together. It's not I'm 6:54 a head of engineering, I'm a this or 6:56 that. There seems like more 6:58 responsibility about who's in those 7:00 seats. How is that evolving? One of the 7:02 trickiest part of parts of being a chief 7:04 people officer is you are the HR 7:06 business partner for the executive team. 7:08 And that includes being able to flip a 7:09 switch between saying, "Hey, let me help 7:11 you with your org and let's do an X-ray 7:13 of your organization," but also to be 7:14 able to turn to you and say, "Hey, the 7:16 way you showed up in that meeting, 7:18 that's not the behavior we all expect of 7:19 each other." And so, you have to be the 7:21 most trusted person in that room. You 7:23 have to be able to have those 7:24 conversations confidentially, but then 7:26 you also have to be able to go to not 7:27 just your CEO, but also your board and 7:29 say, "Here's how our current team stacks 7:31 up, including yourself and your own 7:33 team, and here's where I think we need 7:34 to be to succeed in a year's time or 3 7:36 years time." And I think that role is 7:39 really delicate, really important, and 7:40 really strategic. And I think the most 7:42 important currency you have in that role 7:44 is trust. So the second you have a 7:46 conversation where you and I are talking 7:47 and you're like, "Hey, I just have to 7:49 say I think my board presentation didn't 7:51 land." And I'm like, "Yep, it didn't." 7:52 And here's why. And we have to have a 7:54 conversation. You find that I shared 7:55 that back with our board or shared with 7:57 our CEO. You're never going to come to 7:58 me again. So part of your job is 8:00 discretion and trust. And I think that's 8:02 one of the most important functions of a 8:04 CPU. 8:04 >> But how do you figure that out? It feels 8:06 like it's one is it could lead to just 8:09 insane unproductive politicking where 8:12 you have this like background shadow 8:15 role like you're wearing multiple hats. 8:19 Certainly, you can't hermetically seal 8:21 any conversation you have with another 8:23 exec or you're not doing your job as 8:25 really a a leader and shareholder in the 8:27 business. 8:28 >> I think your first point is the worst 8:30 case scenario of a CPO, the political 8:31 like monster who's advocating and always 8:34 kind of playing chess behind the scenes. 8:36 What I would always say to our team and 8:38 I would say this explicitly to everyone 8:40 is I will not share what you personally 8:42 shared with me. I will share thematic 8:43 feedback. So, hey, the executive team is 8:46 concerned that we're not doing enough 8:47 here. I'm not going to say you as an 8:49 individual said this, but I am going to 8:50 share thematic feedback. It is my duty 8:52 to the to the company and to our 8:53 customers. And so I think some of it is 8:55 just being super clear on expectation 8:57 setting. I think the second part of it 8:59 though is just really taking that 9:00 responsibility seriously. You cannot 9:02 take a momentary opportunity to seize 9:05 power at the risk of trust. And so I 9:07 think the politicking is part of why 9:09 that CPO role is so critically important 9:11 and why the best people at that role 9:13 know that they need the trust to do 9:14 their job every day. What percentage of 9:16 CPOs do you think behave in that way 9:19 versus a politicking way? 9:21 >> I think the better question is what 9:22 percentage of executives generally do 9:24 that? And I think CPOS in particular 9:26 have an additional responsibility. But I 9:28 think it's low. I think that's part of 9:29 why a high performing executive team is 9:31 not just their resumes and CVs of people 9:33 in those seats. It's how they operate 9:35 together as a team. And I think the 9:36 reality is there are a lot of folks who 9:38 play in politics and there are a lot of 9:40 folks who are out for themselves or 9:41 their team. And I think what you said 9:43 about being a first team, that's part of 9:44 why a high functioning executive team 9:46 puts that trust and also the 9:48 disagreement that you have to be able to 9:50 disagree. I don't trust teams where they 9:52 say we don't agree. We're a family. I 9:54 hate when people say we're a family. To 9:56 me, 9:57 >> you're a high performance sports team 9:58 type person. 9:59 >> High performance sports team works. But 10:01 I think high performing sports teams, 10:03 you see when you want to win, like one 10:04 of my favorite things to call example to 10:07 call as an example is like Michael 10:09 Jordan was a really hard person to be in 10:11 practice with. really hard, notably 10:12 hard. He annoyed and agitated his 10:15 teammates. He was a better teammate than 10:16 he was a friend by far. And I think 10:18 that's part of what like your job as an 10:20 exec is to be a great teammate. He was 10:22 incredibly loyal to his team, but he was 10:25 a pain because he tried to make them 10:27 better every single practice. That's 10:29 what you want as an executive. 10:30 >> How would you describe the difference 10:32 between that and an person that 10:34 just nobody wants to work with and is so 10:36 disagreeable? I think one is, you know, 10:38 would you pass to Steve Kerr in like the 10:40 most important high cut? Like, are you 10:42 willing to give up the ball when the 10:44 stakes are highest? I think that 10:45 selflessness is critically important. 10:47 And then two is I think people need to 10:48 believe that you want to win as a team 10:49 and that's not just about you. And I 10:51 think what you hear over and over again 10:52 from interviews with his teammates was, 10:54 "Yeah, he could be extremely difficult 10:56 and confrontational." And would I 10:58 necessarily want to have dinner with him 10:59 every night? No. But I always knew he 11:01 was going to be the hardest working 11:02 person out there. Same thing with Kobe. 11:04 What you see with him is his work ethic 11:06 was unprecedented and I think it makes 11:08 it a whole lot easier to follow someone 11:10 to battle if you feel like they're 11:11 working just as hard if not harder for 11:13 you than the cause. So what about if if 11:15 you think another executive is just 11:17 doesn't have the capability to do the 11:19 job do you then have a conversation with 11:22 the CEO or like what is in in the way 11:25 that you see the world how does one 11:27 navigate that 11:28 >> ideally onetoone so ideally the 11:30 conversation would first be had where 11:31 hey how do you think this is going and 11:33 ideally you have the self-awareness to 11:34 recognize that yourself but if you don't 11:36 give the feedback directly to say I 11:38 think where we're falling shortter in a 11:39 few different ways number one your team 11:41 doesn't have the trust you need number 11:42 two the results aren't 11:44 Number three is as you're showing up as 11:45 a leader within our executive team, I 11:47 don't think we have the bar we're 11:48 looking for. Those are really hard and 11:51 painful conversations, but I would 11:53 always rather someone hear that from me 11:55 than hear it later or wonder about it 11:57 after they were termed. And then to your 11:59 point, if I feel like someone's 12:00 underperforming, absolutely have to tell 12:02 the CEO and the board what that looks 12:03 like. And the idea is not every exit has 12:06 to be a fiery exit. I think a lot of 12:09 times people can be great leaders or 12:11 great executives just not for that time 12:12 period, that stage or that company. And 12:14 I think part of your job as a leader is 12:16 to create as many graceful exits as 12:18 possible. 12:19 >> How does something like that not come 12:21 into conflict with the idea of a 12:22 transparency value? 12:24 >> I would argue that transparency is not 12:26 you having access to every bit of 12:27 information. It's access to as much 12:29 information that is safe for the company 12:30 to share. And so in this case, I think 12:32 hearing directly from your boss and from 12:34 your peer that you're not meeting 12:36 expectations, I think is the 12:37 transparency that most executives 12:39 deserve. 12:40 >> But if someone's fired for performance 12:42 reasons, you think it is 12:43 >> good example. So I think that is not 12:46 your information. So I think the company 12:48 has an obligation to be transparent 12:50 around information. If you share why an 12:52 executive was termed, there are two 12:53 things that could go wrong. Number one, 12:55 it could really impact their ability to 12:56 get a job at another company. Number 12:58 two, it creates rumors and discussion. 13:00 And it also means that during the the 13:02 odd time when you don't share why 13:04 someone is terminated, they make up a 13:06 story that's way worse than whatever it 13:07 was. And so what I would always say to 13:09 our team is we are going to be 13:10 incredibly and radically transparent 13:12 around the things that are our 13:13 information to share. And we're going to 13:15 be super thoughtful and discreet about 13:17 information that is not ours to share, 13:18 including why we've chosen to part ways 13:20 with people. And I think the reality is 13:23 once you explain that to people that 13:24 it's not one situation, that's how we 13:26 always handle exits, I actually think 13:28 people are pretty understanding of it 13:29 because what I always say to people is 13:30 let's pretend that's you. How would you 13:32 want us to treat you? And if we said 13:34 actually this person wasn't cutting it, 13:36 that could impact your job or your 13:38 family moving forward and that's 13:40 unhelpful to our employer brand. 13:41 >> Do you think there's a scenario where it 13:44 does make sense to tell a team that 13:46 somebody was let go? 13:47 >> Yes, I do. I've exercised caution in how 13:50 we do that, but I do think there are 13:51 situations where it makes sense. I also 13:53 think it's totally fine to share that in 13:55 context. And I think there are a few 13:57 examples where sometimes we've been 13:58 clear that it was for underperformance 14:00 or for if someone publicly violates your 14:04 rule, your code of conduct. I think to 14:06 say just in a brief way, Kitty did not 14:08 meet the standards of expectations that 14:10 we have at this organization. I think 14:12 that is totally fair. You don't have to 14:13 go into detail about why. You don't have 14:15 to say more than that. But I think being 14:16 clear with people that it was not our 14:18 expectation is good. 14:19 >> When you think back to the different 14:21 chapters of your career, when is being 14:26 in alignment with values, 14:29 when did you find it was the trickiest 14:30 or the hardest? 14:32 >> I mean, at HubSpot, we like so many 14:34 organizations did a layoff in 2023. 14:37 And I think when you think about our 14:39 values, one of our values is empathy. in 14:42 a lot of the outrage after layoff was it 14:44 doesn't feel like that meets our 14:45 standard of empathy and I think there 14:47 are reasonable I think there's 14:49 reasonable arguments to be said and so I 14:50 think that was one of the the hardest. 14:53 We also had a situation in 2021 where 14:57 like so many people we got through 2020 14:59 and we're rounding the corner in 2021 15:01 and we were kind of feeling like okay 15:02 we're gonna get back to normal and our 15:04 CEO at the time Brian Haligan got in a 15:06 near-death snowmoiling accident and 15:08 thankfully he's okay made a full 15:10 recovery but it was not clear at the 15:11 time how that was all going to go. 15:14 Transparency is tricky in that situation 15:16 because you're dealing with medical 15:18 information. You're dealing with 15:19 someone's family. you're dealing with 15:21 then a public company and disclosure and 15:22 precedent. And so I would say that was a 15:24 really hard one too. So those were the 15:26 two examples that come to mind on 15:28 empathy and transparency that were the 15:29 hardest. 15:30 >> And like if you take the layoffs, what 15:33 how did you work through that tension? 15:36 >> Some people know this, some people 15:37 don't, but when you decide to do a 15:38 layoff as a company, it's a big enough 15:40 decision that your board is certainly 15:41 involved, your entire executive team, 15:43 and you have to be kind of all in it 15:45 together because it's a it's a one-way 15:47 route that you're running, and you have 15:48 to be super thoughtful about it. So part 15:50 of what we did was we asked ourselves, 15:52 how do you infuse as much empathy in as 15:54 possible into a situation that is 15:56 incredibly difficult that included our 15:58 package that included for example we 16:01 tried to be extremely generous with our 16:03 severance package. We got board approval 16:04 to do that and I think even now many 16:07 years later people have said our our 16:10 terms for that were super thoughtful and 16:11 we've also heard from people they 16:13 appreciate how we handled it. So any 16:15 person who was part of the layoff who 16:17 wanted to talk to a senior leader, we 16:19 created office hours. So you got 16:20 one-on-one time. So if you wanted to 16:22 complain, if you wanted to get mad, if 16:25 you wanted to yell, we created space for 16:26 that. And the reason for that is I think 16:28 when you're making a hard decision, you 16:30 never want to be so far from people that 16:32 you don't feel the impact of that 16:33 decision. So I will never forget what 16:35 that felt like. It was incredibly hard. 16:37 It was most hard on the people who were 16:38 impacted by the layoff. Make no mistake 16:41 about that. But I think sometimes making 16:43 yourself available to actually have to 16:44 feel the weight of that decision is 16:46 critically important. 16:47 >> What did that experience teach you? 16:49 >> Layoffs teach you a lot about hangovers. 16:52 So I think a layoff has a hangover of I 16:54 think two to two and a half years at a 16:56 company. And I think you have to be 16:58 aware and cognizant of that going in 17:00 because there's the initial shock then 17:02 there's the survivor guilt for people 17:04 who are still at your organization and 17:06 then there's the ongoing lagging fear of 17:08 people that another one is coming. And 17:10 so I think some of what you have to keep 17:12 in mind is our board members told us 17:13 there's probably going to be a 2-year 17:15 period. And I was like, well, we're 17:16 we're a pretty agile organization, 17:18 pretty fast moving. There is a true 17:20 cultural hangover of a layoff. And I 17:22 think that's sort of a a an important 17:25 warning for people to know is that it's 17:28 not just a one-time business decision. 17:29 It's a cultural decision that has 17:31 implications for many years to come. the 17:33 so what is that you just have to be 17:34 constantly nurturing the organization 17:36 after that or keeping it in mind or 17:38 >> I think the so what is you have to go in 17:40 eyes wide open that it's a two and a 17:41 half year process not a 90-day process 17:44 and then I think the other so what is 17:45 just to be mindful as you think about 17:47 workforce planning just the implications 17:48 of during hyperrowth how are you 17:51 thoughtful about balancing adding a ton 17:53 of headcount which frankly is 17:54 missionritical to achieving your goals 17:56 with balancing where you might want to 17:57 flex up or down with contractors or 17:59 other help and then finally it's just 18:01 recognizing like the decisions that you 18:03 carry as a business leader have 18:05 significant weight on people's lives and 18:06 their famil family's lives and I think 18:07 it's always just important to have that 18:08 at the back of your mind. 18:09 >> Maybe sort of in a in a sort of similar 18:12 vein it feels like the process by which 18:15 resources are allocated 18:19 at least I mean I think still to this 18:20 day but in general is very squishy. Does 18:23 this do we need four people for this, 18:25 nine people for this? We could do this 18:26 with three, but it would be easier with 18:28 10 or it would be and it feels like the 18:30 whole precision around how many people 18:33 do you need for X or Y. 18:36 It's this very messy process and it's I 18:39 think one of the reasons why in general 18:41 you end up with tremendous bloat at most 18:44 companies like the normal just 18:46 gravitational pull whether it be 18:48 executives building thieftdoms and power 18:50 being how many people work underneath 18:52 you or the fact is you know I think we 18:54 could really squeeze to get this goal 18:56 with three but it'll be a heck of a lot 18:58 easier with six so I'm going to ask for 18:59 eight and then we're going to end up at 19:01 six. What are some of the things that 19:03 you figured out around headcount 19:06 planning and what we need for any given 19:09 part of the org? 19:11 >> I think headcount planning I've never 19:13 been an org who's like we have 19:14 absolutely excelled at headcount 19:16 planning. So I think some of it is just 19:17 normalizing exactly as you said that 19:19 it's messy, complicated, inherently 19:21 human. And I say all that with the 19:23 notion that now it's going to get all 19:24 that much more complicated when you add 19:26 agents into the mix. So if you think 19:28 about it, it was already so hard when it 19:30 related to just humans. Now you add 19:31 agents in into workforce planning, it's 19:33 going to get a whole lot messier. And so 19:35 I think a few things that have worked is 19:37 number one is just defining the work 19:39 first that needs to be done. So for 19:40 example, a bottoms up model where people 19:42 just say here's how many heads I need 19:45 without defining what the work is. That 19:46 always falls short because to your point 19:48 the FFTs exist. There isn't necessarily 19:51 alignment on how many folks on EPD or go 19:53 to market are actually needed. For 19:55 example, PLG is a good idea, a good 19:57 example of emotion where you can't just 19:59 spin that team up agnostically. And then 20:02 number two is really pushing on the what 20:04 is the work that needs to be done? 20:05 What's the level that needs to get done 20:07 and having a level of accountability 20:09 between your finance people and business 20:11 teams on it? And then number three is 20:13 just having people really feel an 20:15 understanding of like with every head we 20:17 add, we are adding exactly to your 20:19 point, yes, a great skill and work to be 20:21 done, but you're also adding the 20:23 potential for disperate ownership. So 20:25 one of the things that happens is if two 20:27 people are in charge of anything in a 20:28 company, the chances it gets done well 20:30 are pretty limited. And so part of what 20:32 you have to do is drive discipline along 20:33 with every time you're adding people, 20:35 you have to add a little bit more 20:36 discipline to the process to make sure 20:38 that everyone is clear on what your 20:39 goals are and that everyone is clear at 20:41 the speed at which they're expected to 20:42 achieve them. 20:43 >> Walk us through maybe in the case of 20:45 Harvey, you're doing headcount planning. 20:47 What does it actually look like? Who's 20:48 involved? Like what does the process 20:50 look like? 20:51 >> I'll give you a recent example. So we 20:54 just hired a chief product officer, 20:55 Anique Drum Rate. She joined us from 20:56 Ripling. She is currently with my team 20:59 starting tomorrow for 2 days. Part of 21:02 what we're doing is new executive comes 21:04 on board. You want to be thoughtful to 21:05 make sure that the headcount plan that 21:07 we did, you know, all of 6 weeks ago is 21:10 still relevant for what she's thinking 21:11 she needs. And obviously, she spent a 21:13 ton of time with our team already just 21:14 getting to know us and getting to know 21:16 our customers a bit and also has gotten 21:18 feedback from her direct reports. And so 21:19 part of it is if you have an annual 21:22 plan, especially in the world of AI, 21:24 that is going to get stale very fast. 21:25 I'm saying this to you now as someone, 21:27 you know, 5 weeks in, we're already 21:28 doing a replan. That doesn't mean 21:30 starting over from scratch. It just 21:32 means for the next quarter ahead, what 21:34 do we need to change or adjust to make 21:35 sure we're thoughtful and calibrated? 21:37 Number two is you need to make sure just 21:38 at the very basics that you have enough 21:40 ramped recruiters to achieve those 21:42 goals. So, one of the mistakes people 21:43 make is for example putting a number 21:45 into Q2 headcount in AMIA when notice 21:48 periods simply will not allow that to 21:49 happen. And then three, I think, is just 21:52 aligning to business need first and 21:53 foremost. So another every once in a 21:55 while you'll hear from someone like I 21:56 Katie just want to hire this person and 21:58 part of what you need to push is like a 22:00 level of business need without adding a 22:02 ton of process. So in other words what 22:04 I've seen people do is in the age of AI 22:06 now people are like I have to write a 22:08 full wreck I have to get 17 people's 22:10 approval to get headcount open. I don't 22:12 think that works either. So it's finding 22:13 the balance between being agile and 22:14 ownorous in the planning. And then I 22:16 think when you add a new leader you have 22:18 to be open to letting them give feedback 22:20 on the current plan versus keeping it 22:21 rigid. Otherwise, I think the gap 22:23 between what you were intending to do 22:25 and the people you have on board doing 22:26 it just gets too too wide. So, at 22:28 Harvey, for example, we do a financial 22:30 plan for the entire year. We do a 22:31 headcount plan for the year, but we do a 22:33 mini replan on a quarterly basis based 22:36 on business needs, based on geos, all 22:38 that kind of good stuff. Maybe switching 22:39 gears slightly, I I think if you look 22:41 back certainly in the late 2010s and 22:43 into the pandemic, a lot of people 22:45 organizations sort of became some 22:47 version of a nanny state that it was 22:49 about employee happiness that it was 22:51 about everyone comes to work dancing 22:53 down the street. It's it's and that it 22:56 it felt like many obviously not all but 22:58 many of the organizations became like 23:00 some version of a coddling or happiness 23:04 function. And maybe on the other end of 23:06 the extreme there's sort of some dynamic 23:09 where the the people function really 23:10 drives performance and at times there's 23:13 maybe a tension between those things or 23:14 a misunderstanding about those things. 23:16 What are your reflections on that? And 23:18 is it company specific? Certain 23:19 companies are nanny states and that's 23:20 just if you want a nanny if you want to 23:23 be nanny excuse me that's a great 23:24 company to go to and that that it's very 23:27 company dependent where a lot of people 23:28 just have this wholesale wrong. Our VP 23:31 of talent, Maggie Landers, has a great 23:33 expression which is the resort has to 23:34 match the brochure. And I think when it 23:36 comes to the coddling versus performance 23:39 culture, what you say you do versus what 23:41 you do has to match. I do think so many 23:44 companies went too far in the 2010s. I 23:47 do think people thought our job was to 23:49 make people happy every day. And I think 23:51 the Disneyland approach to employee 23:53 experience is gone. On the flip side, I 23:56 think we've overcorrected, which is to 23:58 say work should be hard. get back to 24:00 work, return to office mandates, taking 24:02 away perks, taking away the humanity of 24:04 it. I don't want to jerk the wheel too 24:06 far. So, when I was talking to founders, 24:08 one of the questions that I asked them 24:09 as part of my interview process was, 24:11 "How do you think about severance 24:12 packages for employees who don't work 24:14 out?" And one of the founders I spoke to 24:16 said, "I take care of the people who 24:18 work here, not people who don't." And I 24:19 was like, "Right, I hear you, but people 24:21 took the bet on you. And don't you feel 24:24 like you should at least be fair and 24:25 ethical on that?" And the feedback that 24:26 Winston and Gabe had was we should do 24:28 whatever is fair and market. And by the 24:31 way, we know that those people talk to 24:33 other people in the industry. And so we 24:35 want to make sure we're thoughtful about 24:36 the flywheel we're creating for talent. 24:38 And so I think when we talk about 24:39 Harvey, we talk about the fact that 24:40 we're incredibly intense. We don't make 24:42 any apologies for that. But we're 24:44 intense but reasonable. So in other 24:46 words, if you have a doctor's 24:47 appointment, if you're taking parental 24:48 leave, we want to make sure that you 24:49 feel like you can actually take it. We 24:51 want to hire great humans, but also at 24:53 the same time, we want to make sure 24:54 we're clear that this is not a 9-to-f4 24:56 culture. And I think balancing those 24:58 two, all you have to do is be clear on 24:59 who you are and make sure that your 25:01 interview process matches what people 25:03 experience dayto-day. 25:04 >> But the part of what you're saying is 25:05 the way that you describe that is is 25:07 really only correct for Harvey as an 25:09 organism. It's not globally correct. 25:12 >> I don't think there's a globally correct 25:14 way to do things. So, for example, when 25:15 I was at HubSpot, we spent so much time 25:17 on culture and employee experience. And 25:19 people would often bait me to say bad 25:21 things about Amazon because they viewed 25:22 it as antithetical to what we did at 25:24 HubSpot. I'm like, I actually think 25:26 Amazon has a good culture. They're clear 25:28 on who they are and you know what you're 25:30 signing up for. So, for example, they 25:32 pit teams, internal teams against each 25:34 other. They're clear about that in the 25:35 interview process. So, I think that is a 25:37 functional culture. I have a much bigger 25:39 problem with people who advertise it's 25:41 all unicorns and rainbows and when you 25:43 come in it's cutth through. 25:44 >> Do you think you at different points in 25:47 your career created too much of like I'm 25:50 in charge of h employee happiness and 25:53 all of that type of stuff or did do you 25:54 think you sort of threaded the needle 25:56 appropriately? 25:57 >> I think everyone did at that time and I 25:59 think HubSpot was no different and I was 26:00 certainly no different as a leader. I 26:03 think what we tried to get people back 26:04 to was our job is not to make you happy 26:07 every day. Our job is to create an 26:08 environment where you can do the best 26:09 work of your career and we said that 26:11 explicitly to people. I think how that 26:13 actually takes shape dayto-day is really 26:15 hard to enforce and get right and there 26:18 were certainly days when we overindex 26:19 too high to the unicorns and rainbows 26:22 and I think disappointed people and I 26:24 think the reality is we had to find a 26:25 better balance. 26:26 >> What have you figured out about the 26:27 tricky things related to incentives 26:30 right? Right? If you go back to the idea 26:31 of the Charlie Mongerism, incentives 26:33 rule the world, that has to express 26:35 itself in so many ways in a people 26:37 function, compensation, bonuses, 26:40 promotions. Do you have any sort of meta 26:43 thoughts or like specific things that 26:44 you figured out that are like 26:46 particularly tricky about the way that 26:49 incentives works? particularly because 26:51 in so many ways incentives are some 26:53 proxy metric that like at the end of the 26:56 day you're trying to deliver something 26:57 for customers that in turn creates 26:59 shareholder value in a particular way 27:01 when you sort of think about it actually 27:02 as a business and then you have all the 27:04 human beings andor agents now in pursuit 27:08 of sort of that flywheel of shareholder 27:10 value creation and then you have sort of 27:12 all of these things that try to 27:14 approximate those and it seems like it 27:16 gets even more complicated at each 27:18 chapter of sort of company growth and 27:19 more people and sort of that type of 27:21 thing. 27:21 >> Charlie Munger was preient about this as 27:23 he has done so many different things. I 27:25 would say a few things from having 27:27 having watched this up close that stick 27:29 out. Number one is you have to keep your 27:31 incentives as simple as possible. So a 27:34 good example would be as you're thinking 27:35 about executive compensation a bonus 27:38 program for your entire company. The 27:40 temptation is to make it six different 27:42 variables that folks can then 27:43 manipulate. And the idea is then it 27:45 makes it harder to manipulate. That is 27:46 fundamentally untrue. I actually think 27:49 simplicity is really important when it 27:50 comes to comp and incentive. 27:52 >> Well, what normally happens to your 27:53 point is it feels like, okay, we're 27:54 going to have some simple bonus 27:56 structure and then you're like, wait, 27:57 no, no, no. It's going to be gamed in 27:59 this way. So, we're going to then 28:01 measure these nine things. 28:03 >> Anytime we would do a bonus structure or 28:05 anytime I do a bonus structure, what I 28:07 often play out including with the board 28:09 is what's the failure mode for each of 28:10 these. Usually, there's a top line, 28:12 bottom line. There aren't many creative 28:13 optics on all things bonus. And so as a 28:16 result, you go, okay, what would failure 28:17 mode look like and what trade-offs are 28:19 we comfortable with given that? So I'm a 28:21 big believer in simplicity wins. Two, on 28:23 incentives is I have not met, 28:25 counterintuitively, most managers are 28:27 not actually out for themselves. They 28:29 think of their first team as the people 28:31 below them and they view their job as to 28:33 protect and lobby for the people below 28:35 them. And so often times what I would 28:37 see is not prioritizing themselves over 28:40 the customer of the company, but 28:41 prioritizing their team. And so I think 28:43 when you know that part of what you have 28:45 to think about is when you go into 28:46 promotion calibrations, when you go into 28:49 giving managers autonomy over 28:50 performance ratings, giving managers 28:52 autonomy around compensation and equity 28:54 adjustments, that's the frame you have 28:56 to keep in mind is not will they solve 28:58 for themselves as individuals, but how 29:00 do they view their role as lobbyists on 29:03 behalf of their team and how do you 29:04 potentially intermediate that? And in my 29:07 experience, that should be the exec's 29:09 job of that function versus the people 29:12 function to be calling that out. The 29:13 people function obviously plays a 29:14 critical role. But to me, that's holding 29:17 the exec accountable for holding their 29:19 directors accountable, for example. 29:21 >> But how do you then not just how does 29:23 the problem just not flow up to the the 29:26 well normally you have the manager, they 29:28 have a pot of engineers. In that case, 29:29 they're incentivized to get whatever 29:31 they can for them. How does it just not 29:33 end up with this cascading problem where 29:34 then at the very top that person has 29:36 that same cognitive bias to think about 29:38 the thousand engineers that sit sort of 29:40 in multi multiple layers beneath them. 29:42 So every time perform a performance 29:44 cycle, what I try and do is before we 29:46 even get into the how we're going to do 29:47 it, I try and go, what are we solving 29:49 for? And so for example, I will make 29:51 leaders around the table say we want to 29:53 have a high performance culture. We want 29:54 to retain the top 10% of people. And 29:57 what I have found is super helpful is 29:59 pulling up reviews. So often times I 30:00 will pull up a blinded review. And I'll 30:02 say, what do you think this person got 30:04 on their rating in this team? And I will 30:06 include someone from my own org so I'm 30:07 not throwing shade at anyone. And 30:09 oftentimes that's a humbling moment for 30:11 people to go, okay, last cycle I let 30:13 that go on my team. Do we want to have 30:15 an environment where we do that? Okay, 30:17 maybe that's fine if we do. If we want 30:18 to be more permissive, that's okay. But 30:20 if we want to be a high performance 30:22 culture, we need to be honest about the 30:23 culture we create in our own teams and 30:25 then we set the objectives together. So 30:27 I will come in with a straw man and say, 30:28 "Here's what I think we should roughly 30:30 do." But often times I hold up a mirror 30:32 from the last review cycle. I find using 30:34 verbatims is super helpful because I 30:36 think if you talk in the abstract or 30:38 pull from some performance management 30:39 book, you're probably going to fall 30:40 short. If you pull examples from real 30:43 managers saying, "Here's what I heard or 30:45 from better yet from your HR from the 30:47 last cycle." I find that is much more 30:50 likely to actually change people's 30:51 behavior. 30:52 >> What have you figured out about titles 30:54 when you're at a fast growing company 30:56 and like if someone is they're 100% 30:59 company, their plan is to get they're 31:01 going to be 250 people this year. 31:03 they're kind of have now a career ladder 31:06 and that type of thing. What's like your 31:08 important advice for thinking through 31:11 what happens with titles as this company 31:13 starts to really grow. 31:14 >> So I would say a few things. I think 31:16 executives who say titles don't matter 31:19 are delusional. They do matter to people 31:21 and so does the opportunity to grow. 31:23 However, if you are someone who in the 31:25 interview process is mostly focused on 31:27 what your title is going to be, your 31:28 behavior during your entire duration of 31:31 the process. So, one of my pieces of 31:33 insight is people are on their best 31:34 behavior as candidates. I can't even 31:36 tell you how many conversations I've had 31:37 where I'm like, he was a little 31:38 difficult in the negotiation, but I'm 31:40 sure once he's here, he'll be a real 31:41 peach, and that has not been my 31:43 experience. So, I think if someone is 31:45 lobbying for title, they are going to be 31:46 that difficult once they're on board. 31:48 However, I believe not surprisingly on 31:51 titles that you should solve with a 31:52 level of simplicity. So, for example, at 31:55 Harvey, we don't have an SVP layer yet. 31:57 That's pretty intentional on our part. 31:59 We have head of layers, we have VP 32:01 layers, we have sea level layers. We 32:03 don't feel like at 650 people, we need 32:05 to add that level of complexity. With 32:07 that said, we also don't have a 32:08 perfectly rigid career ladder. So, what 32:10 I often say for people interviewing is 32:12 they'll say, "What's your exact ladder?" 32:13 I'll say, "We're growing so fast. We 32:15 have guides. We have rough leveling, but 32:18 we don't have this clear. If you do this 32:19 in 2 years, you'll be promoted." That's 32:21 part of the joy of joining a fast 32:24 growing AI startup, but it also means 32:25 you got to be comfortable with the 32:26 ambiguity that comes along with it. And 32:29 in my experience, you need to make sure 32:30 you set some federal guidelines on that. 32:32 So if for example you one of the things 32:34 you could do at a startup is say, hey, 32:35 you do your own thing, I'll do mine and 32:37 it will get figured out later as we 32:38 scale 32:39 >> just push it down to sort of some 32:41 manager or this exact can do whatever 32:43 they want in that. 32:43 >> That gets exactly as messy as you would 32:45 think based on what we talked about with 32:46 headcount planning. And so I'm a big 32:48 believer that leveling at least rough 32:50 guidelines should be a federal decision 32:52 versus state. At what point do you think 32:55 that should be made much more explicit 32:57 or legible versus like, yeah, we roughly 33:00 have it, but we're going to figure out 33:01 as we go. 33:02 >> So, what we did at Harvey was we did a 33:04 federal system, but then the states are 33:06 responsible for figuring out how they 33:07 level each individual. That to me feels 33:09 fair. Doing the actual leveling at the 33:12 federal level was pretty painful at our 33:13 stage and size. I would say we did it 33:15 around 300 people. The reason we did it 33:17 was we were about to double or triple in 33:19 size. So given that, it just felt like 33:21 as we're bringing in a bunch of new 33:22 people, this is only going to get 33:23 messier. Leveling is exactly as messy no 33:27 matter what stage. So we did a leveling 33:29 exercise at HubSpot at 8,000 people. We 33:31 did one at Harpy at 300 people. It's 33:33 always really messy because it targets 33:36 people's selfidentity and awareness and 33:38 how they align with a title. And so I 33:40 think just going in knowing that you 33:42 have to appreciate the emotional depth 33:43 that goes behind a leveling decision. 33:45 Then you have to pick a lane that 33:46 includes trade-offs and then set the 33:48 course. All executives experience this, 33:51 but certainly the chief people officer 33:52 probably more so than anyone, which is 33:54 basically by definition when you're 33:56 operating a large company, you're going 33:57 to have a subset of people that no 33:59 matter what you do or what the approach 34:01 is or what the philosophy is will not 34:03 like it. You could have the greatest 34:05 snacks in the world and there's going to 34:07 be someone that's, you know, has all 34:09 these different allergies and they don't 34:11 like the snacks and you don't get the 34:13 snacks that I You could come up with any 34:14 possible policy and it just feels like 34:17 you're constantly dealing with a decent 34:20 chunk of people that are frustrated 34:21 about something. Did you just learn to 34:23 be at peace with that or why does it not 34:26 drive you insane? I don't think being a 34:29 chief people officer is for the faint of 34:31 heart because I think you have to be 34:33 okay with the fact that at any given 34:35 moment roughly 20% of people are unhappy 34:38 with you the decisions you 34:39 >> look at a large organization there's a 34:40 lot of people 34:41 >> a lot of people but the other thing 34:42 about it is if you the worst thing you 34:45 could possibly do is say that doesn't 34:46 matter they're wrong that's terrible 34:49 because then you shut down to important 34:51 critical feedback that I think really 34:52 matters and so I think the balance 34:54 between listening to things that people 34:57 have very valid complaints about on 34:59 things that matter is critical. Ignoring 35:01 the things that don't really matter. So 35:03 a good example, our old COO at HubSpot 35:06 had an expression called protein versus 35:07 sugar. And the best way to think about 35:09 it is people's substantive complaints 35:11 around things that really matter that 35:13 stand in the way of us winning that are 35:15 at odds with our culture. Those you got 35:17 to really pay attention to. The 35:18 complaining about the snacks really 35:20 doesn't matter. We had an episode that 35:22 will forever be famous in the halls of 35:24 HubSpot called Berrygate. And basically 35:26 what happened was we used to provide 35:29 fresh berries as a perk and the fresh 35:32 berry the finance team created a chart 35:34 where the berry consumption versus 35:36 headcount was disproportionate. We had 35:38 this important management team meeting 35:40 to figure it out. 35:40 >> It was like an exponential or something. 35:42 >> Yeah, it was exponential and the cost of 35:43 berries was going up and you can imagine 35:45 like how much time 35:45 >> double exponential curve. 35:47 >> Yes. And then the price of berries was 35:48 going up and what could happen and how 35:50 could this obviously a ridiculous 35:52 conversation but nonetheless 35:53 >> but it was it was not a joke 35:54 conversation. It was not a joke 35:55 conversation at all. We had a serious 35:57 conversation about it and we figured out 35:58 what we considered at the time to be an 36:00 elegant solution to it, which was we 36:02 were going to take away the fresh 36:03 berries, but we were going to open a 36:05 smoothie bar. And the smoothie bar was a 36:07 win-win, at least from this is like 36:09 executive leadership, but it's fin 36:13 because customers could come in. We 36:14 named the smoothies after customers. It 36:16 was health forward. It provided a 36:18 non-drinking option. Like, what could go 36:19 wrong? 36:20 >> You leave this meeting and you guys are 36:21 like, we 36:22 >> killed it. Like honestly, like pat on 36:26 the back, round of applause. No one has 36:28 better elegant solutions than us. I get 36:30 back to my desk and a woman on my team 36:33 who I really trust will not make eye 36:34 contact with me. Nothing to see here. 36:36 She's literally avoiding me at all cost. 36:38 And I was like, tell me exactly what's 36:39 going on right now. She's like, well, 36:42 there might be a bit of it a bit of an 36:44 issue. And I was like, talk to me about 36:46 what you mean. And basically there was a 36:48 whole Slack channel discussion around 36:51 how employees were not consulted in this 36:53 decision. There were a few other things. 36:55 Do I not know that the fiber content of 36:58 berries is better when you consume the 37:00 berries whole than in a smoothie? Why 37:03 was our C why did our CFO not hold a 37:06 town hall meeting for people's input on 37:08 this? I pretty rarely at HubSpot I 37:11 always was sort of like what could I do 37:12 better? What could I do wrong? I wrote a 37:13 note to everyone and I was like, I think 37:15 we've completely lost the plot on stuff 37:18 like this. We are not going to consult 37:19 the entire company. We are going to make 37:22 decisions that have winners and losers 37:23 and we're going to make hard calls in 37:25 the interest of scale. And in this 37:27 particular case, if you have spent the 37:29 last 45 minutes complaining in the Slack 37:31 channel about Barrygate, you clearly 37:32 don't have enough work to do and we need 37:34 to have a different conversation. And 37:36 what's interesting about that moment in 37:37 time is it became a rallying cry for 37:39 what people actually cared about. our 37:41 most loyal employees were like, "Great. 37:43 I'm so glad you said something because 37:45 we've had entitlement issues for a 37:46 while." And so, I think part of what you 37:48 have to do as a chief people officer is 37:50 be clear on the feedback that matters 37:51 versus the feedback that doesn't. I got 37:54 a lot of really tough feedback on things 37:55 that made me better that I had to listen 37:57 to that were really important. So, a 37:59 good example would be we had one 38:01 engineer in Dublin that had a comment 38:03 for almost everything that we did. The 38:06 reality was he was right on like let's 38:08 call it 75% of things that he emailed me 38:11 about. And so when I got an email from 38:13 him I'd always go what can I be doing 38:14 better or differently and take it 38:16 seriously. And so I think the temptation 38:18 is just ignore all the noise and I think 38:20 that's a mistake but if you pay 38:21 attention to the noise it will there's a 38:24 reason that the average tenure of a CPO 38:26 is short. Some of it is because of you 38:28 know maybe not the right fit for 38:30 business value or stage but some of it 38:31 is because it's a really hard job. It 38:33 takes a big emotional attack on people. 38:35 Do you think based on sort of all of 38:37 your different experiences that if 38:39 someone is complaining about things that 38:41 are just, you know, berries 38:44 that you should just get rid of them? 38:46 That like you that it is highly unlikely 38:48 that high performance the most important 38:51 people in the company are complaining 38:52 about berries or no actually there's a 38:54 lot of star players that complain about 38:55 berries. 38:56 >> There are a lot of star players who 38:57 complain about berries and what we have 38:59 to figure out is the balance between the 39:01 two. So, for example, there were people 39:03 who talked about or complained about 39:04 things. A good example would be the 39:06 office environment, temperature, 39:07 lighting. Lighting is a really good 39:09 example of something engineers are super 39:10 passionate about because they feel like 39:12 if I'm going to stand in front of a 39:13 screen for 8 hours a day, you want to 39:15 make sure you don't have glare. There 39:17 could be health issues and reasons 39:18 behind it. And so, I think the easy 39:20 temptation is to kind of just say 39:22 everyone who complains about what you 39:24 consider to be arbitrary is ridiculous. 39:27 I think that is a huge mistake. And so 39:29 part of what you have to do is figure 39:30 out what's the ratio of people doing 39:32 great work versus complaining. So to me, 39:34 I care less about the berries and more 39:36 about many of the people that were 39:37 complaining in that Slack channel were 39:39 people who are frequent flyers. And I 39:41 care more about the frequent flyers of 39:42 people who I'm like feels like eight out 39:44 of 10 times we have an issue. There are 39:46 these five people that are always around 39:47 the water cooler on it. That to me is a 39:49 more concerning pattern than people 39:51 having fickle complaints. I think most 39:53 human beings have a fickle complaint 39:55 about something. And if you have a 39:56 specific dietary restriction, a lighting 39:59 thing, if you have one thing that you're 40:00 passionate about, I think there are a 40:02 ton of talented people that you can make 40:03 accommodations for, I think the pattern 40:06 of behavior of being a nuisance is a 40:07 bigger red flag. 40:08 >> You said something in passing a second 40:09 ago that I I think a couple people have 40:12 written about maybe in great detail, 40:13 which is basically this idea that 40:15 whenever there is a not enough work at a 40:17 company, all sorts of shenanigans start 40:20 to happen. And that you could argue in 40:22 the late 2010s through the pandemic it 40:24 was at an extreme where like there was 40:26 just not enough demanding work for 40:28 people. When environments are demanding 40:31 and demand a lot there's just no time to 40:33 spend 45 minutes ranting about a banana 40:38 or something. Do you think a lot about 40:40 that in in your role? Do you think 40:41 that's that's generally correct that you 40:44 do need to match the amount of work 40:47 relative to people or it's there's not 40:49 that much to be learned from it? 40:51 >> I think people are at their best when 40:54 things feel demanding but not completely 40:56 overwhelming. And I think that balance 40:57 is really hard. It's also just hard to 40:59 achieve companywide. So, everyone's been 41:01 in sprints where the engineering team is 41:03 working super hard or or times like end 41:05 of year sales and finance are working 41:06 super hard. And so a lot of running a 41:09 company is energy management and getting 41:11 people motivated for sprints and then 41:14 also getting people motivated to take a 41:15 breather when they need it. That's 41:17 really hard I think right now in this 41:19 chapter when it feels like it is growth 41:21 at all costs. And so I do think that 41:23 there's something to like you should be 41:25 you should have enough demanding work 41:26 for everyone to focus on. Actually 41:28 allocating that work across teams and 41:30 geos is really hard to do. In the point 41:32 about the frequent flyers or frequent 41:34 complainers, one of the things that I've 41:37 noticed is they also have a way of of 41:39 indoctrinating their ideas into other 41:41 people and getting other people rallied 41:45 around nonsense. Are you hypervigilant 41:47 of those type of people and you try to 41:49 exit them because you do think it could 41:51 kind of spread? the people with specific 41:54 opinions around certain snacks or things 41:56 or temperatures or items, those people 41:59 don't bother me that much because every 42:00 single one of us had them. The frequent 42:02 flyers are the problem because they get 42:04 people to focus their energy around 42:06 those minute things that don't matter. 42:08 And so to me, I pay a lot of attention 42:09 to frequent flyers. I'll give you a few 42:11 examples. At HubSpot, whenever anyone 42:13 started a new hire training, and if on 42:15 the Thursday of new hire training when 42:17 things wrapped up, they sent me a note 42:18 that said, "Here are my 18 observations 42:20 on things that could have been better or 42:22 differently, and oh, by the way, here's 42:23 my list of requests moving forward." I'd 42:25 always go, "This person will probably 42:28 last 3 to 6 months." The difference 42:30 between that and someone who writes and 42:32 says, "Hey, I just want to say my new 42:34 hair trainer was incredible." Or, "I 42:36 feel so proud to be part of HubSpot, and 42:38 I think these three things just to keep 42:39 in mind would be feedback for next time 42:40 around. look forward to working with 42:42 you. Those are two totally different 42:43 things and the people who bias to 42:45 negative in their first week typically 42:48 is the leading indicator of how they're 42:49 going to operate. 42:50 >> Are there other patterns like that that 42:51 you've observed? 42:52 >> Listening to directions. So, 42:54 interestingly enough, one of the biggest 42:56 predictors of the highest performing 42:58 sales reps was them listening to the 43:00 basics from IT facilities in day one. 43:03 So, long before you get to selling, just 43:04 actually listening to people's 43:06 instructions. And not surprisingly that 43:08 had an impact on performance but also on 43:11 attrition. So that was a leading 43:12 indicator that most people like computer 43:14 setup would not be the intuitive thing 43:15 that you would think would make a 43:16 difference and it does. I think the 43:19 ratio of complaining versus fixing is a 43:21 big one at any organization. And I think 43:23 we can all name someone we work with who 43:25 is complaining 75% of the time and 43:27 fixing 25%. And then the other thing I 43:29 would just say is just how much people 43:31 bring up at my last company I did this 43:33 or I was known for this. I think 43:35 bringing up past experiences, learning 43:37 from other companies is super valuable. 43:39 No one wants to hear you regail everyone 43:41 with the tales of your greatness at your 43:43 other organization. So the relying on 43:45 your resume versus what you're doing at 43:47 your current organization is another red 43:49 flag. 43:49 >> Do you think the manager of a team does 43:53 not need to be the best in that 43:54 discipline? 43:55 >> I don't think they need to be the best, 43:57 but I think this notion of professional 43:58 managers who have no idea what their 43:59 team do all day is also a huge miss. I 44:02 don't think anyone wants to work for 44:03 someone who doesn't understand anything 44:04 about their craft. But I don't think you 44:06 need to be better than the people you 44:07 manage. I manage a ton of people who are 44:10 way better at their area of function, 44:11 but I need to have enough subject matter 44:13 expertise that they can learn something 44:14 from me or at least that I can be 44:16 dangerous in conversation with them. I 44:18 think that credibility is important. 44:20 >> Play devil's advocate. If you're the VP 44:22 of engineering, is there a case that you 44:24 actually should be the most strongest 44:26 technically in that organization? 44:27 >> I think you have to be incredibly sound 44:29 technically. So in other words, I think 44:30 the folks who were like, "Yeah, I did 44:32 engineering for a few years and then 44:34 I've been a professional manager ever 44:35 since." I think it's about that level of 44:36 credibility and I do think that matters. 44:38 And so I think you have to be the best 44:40 is very different from you're extremely 44:42 good and you're enough to get the 44:44 principal or staff engineer to be able 44:46 to talk shop with you. That's an 44:47 important threshold. 44:49 >> Switching gears back a little bit to the 44:50 craft of what it means to be an 44:52 effective chief people officer. explain 44:54 in your mind what is the difference 44:56 between a VP of people and a chief 45:00 people officer in terms of capability. 45:02 Not obviously you're in different roles. 45:04 One you're on the e staff or executive 45:05 team but like in your mind if somebody's 45:08 caught and they're going to really 45:10 struggle they're very good as a VP of 45:12 people but they're going to struggle. 45:14 What's the gap in your mind between 45:15 those two? 45:16 >> I'll answer this generally which is like 45:18 VP versus sea level executives. Where do 45:20 people get caught generally? Because I 45:22 actually think a lot of VPs 45:24 I have unfortunately become the poster 45:26 child for CPOs and so as a result CEOs 45:29 are like when should I hire the CPO 45:31 versus VP of people. I actually think 45:33 both of them can be important exacts at 45:36 a company regardless of stage. I think 45:37 the answer to your question of what 45:39 stands between a VP of people becoming a 45:41 seale executive or really any function. 45:43 Number one is self-awareness. So once 45:46 you hit the VP level, there's this thing 45:47 that happens where you start going, 45:49 okay, I'm actually kind of good at this. 45:51 I've hit this level of competency where 45:52 I'm actually feeling comfortable. You 45:54 start taking in a lot of praise and kind 45:57 of filtering out constructive feedback 45:59 on how you can get better. And so I 46:00 think self-awareness is a big one to get 46:02 from VP to C level. 46:03 >> Why is that not not important to go from 46:05 director to VP? Again, not even getting 46:07 to level lengths, but like squishy sort 46:10 of up the org. Why do you highlight that 46:12 so much more than maybe even director to 46:14 VP? So director to VP, I think it's 46:16 actually relevant for the next one, 46:18 which is I think your ability to recruit 46:19 people who are better than you. I think 46:20 that's applies to both directors and 46:22 VPs. Director toVP, it's how you talk to 46:26 your direct reports. So oftentimes what 46:28 you find directors doing is almost 46:30 acting like a camp camp counselor. Okay, 46:32 friends, here's what we're going to do. 46:34 And everyone knows what it feels like to 46:35 be on the other side of that 46:36 conversation. That's how they treat 46:38 their managers. And that's singularly 46:40 unhelpful. So director to VP, it's how 46:42 do you elevate your operating system so 46:44 you can have a conversation that doesn't 46:45 feel like you're speaking to people like 46:46 a camp counselor. VP to Calele, it's 46:49 self-awareness and then recruiting 46:50 people are better than you. 46:52 >> That sort of note about director to VP 46:53 in terms of managing the team. Like what 46:55 does it sound like if they're doing it 46:57 correctly and it's not a camp counselor? 47:00 >> It's number one trusting and empowering 47:02 the people that are below you. So in 47:03 other words, your meetings are more 47:05 about accountability to that goal versus 47:07 de facto managing their team. So most 47:09 directors are just de facto managing the 47:11 teams for the manager. And what you want 47:13 to hear is your time is more spent 47:15 going, hey, what are the patterns you're 47:16 seeing in our team and how can I best 47:17 help versus, okay, how is this person 47:20 doing? How is person one doing? How is 47:22 person two doing? Number two is that 47:24 they've changed their operating system 47:25 from being a manager or director. So 47:27 what I often ask for is how is your 47:29 calendar different now than it was 6 47:30 months ago when you were director? And 47:32 people often say, well, I have more 47:33 one-on- ones. And I'm like, that's 47:34 actually the wrong answer. What I would 47:36 like to see is fewer one-on- ones. and 47:38 more systems that help you scale. And so 47:40 that's I think for a VP what I see or 47:42 for a great director, you see them go, I 47:44 changed my operating system. I said no 47:46 to interviewing IC's on my team. I'm now 47:48 only focused on interviewing managers 47:50 and on equipping my managers to make 47:52 better decisions for their team. And as 47:54 you move to VP, it's like, hey, now I 47:55 have five directors reporting to me. I'm 47:57 thinking about an operating system that 47:58 matches the urgency we need. And so I've 48:00 switched my team meetings, my skip 48:02 levels, everything to reflect that. 48:04 >> Let's take the scenario. You're a 2500 48:06 person company. you're the chief people 48:08 officer in in the way that you think 48:10 about operating a company. What are the 48:13 types of decisions that you think you 48:14 are singularly responsible for and what 48:17 are the types of decisions that you are 48:18 pushing off or you are sure other people 48:21 should be responsible for? 48:22 >> Yes. So I think company goals, team 48:25 objectives and then what the shape of 48:28 our organization at the end of a given 48:29 year looks like. I think that falls to 48:31 the executive decisions on I see hiring, 48:34 how we're going to build a certain 48:36 campaign, things like that, I think 48:37 should fall to your team and ideally 48:39 you're finding a mix of those two so 48:41 people feel level. I I'm a big believer 48:43 the best people in the world want some 48:45 level of autonomy. I don't think 48:46 anyone's ever been like, "Wow, I can't 48:48 wait to work for her. She's the best 48:49 micromanager around." And so I think 48:51 ideally you're clear on where you play 48:53 so that people understand, hey, this is 48:55 yours. This is your decision to make. 48:57 And so to me, it's the company goals, 48:59 the team goals, and then here's what I 49:00 want us to have achieved by the end of 49:01 the year. And then everything else falls 49:03 to your team. 49:04 >> When you're sort of building a company 49:06 and and you're at scale, you talked 49:08 about this a little bit, but at least in 49:10 the people function, what are the most 49:12 common problems that aren't always 49:15 apparent when you're at the sort of top 49:17 of the org, but like doing this for a 49:19 long time, there are like these hidden 49:21 landmines as you go from 500 people to a 49:23 thousand. and the people or goes from 49:25 two people to 10 people that like what 49:27 are the types of things that maybe go 49:29 overlooked but most people leaders will 49:32 eventually run into these 49:34 >> one is you get so focused on your road 49:36 map of we should be implementing this 49:38 talent program or we should be rolling 49:40 out ACR at this time you get to program 49:43 centric versus people centric and so for 49:45 example if you're rolling out this top 49:48 talent program but 17 people from your 49:51 top talent radar have left during that 49:53 quarter you should stop and be able to 49:55 be agile enough to actually go hold hold 49:57 our horses. Let's do something 49:58 different. I think what I've noticed at 50:00 that scale is often times people are so 50:02 like but we said we were going to do 50:03 this. You have to make sure you have the 50:05 ability to flex the plan based on what 50:07 you're seeing. Number two is just being 50:09 too reliant on just the numbers. So for 50:11 example, attrition is a great example of 50:14 something that is a lagging indicator. 50:16 You often don't realize you have a huge 50:18 problem in a team until six people 50:19 leave. And so part of what you want to 50:21 look at is your analytics for what's 50:23 actually predictive. At HubSpot, we 50:26 found that employee happiness, so just 50:27 saying on a scale of 1 to 10, how likely 50:29 are you to recommend HubSpot? Great 50:30 measure. Feels really good when it's 50:32 high. It wasn't actually predictive. 50:34 What was predictive was people saying, I 50:36 see myself at HubSpot in the next 12 50:38 months. That was the only predictive 50:39 measure on our survey. So as a result, 50:41 you got to get really religious about 50:43 paying attention. And so we would do 50:44 that every single quarter. and you'd pay 50:46 attention to behaviors by geo, by team, 50:48 by level, and you'd have specific 50:50 actions based on how you interact with 50:53 those groups because otherwise you're 50:54 going to miss it. It's too late. So 50:55 looking at leading indicators is another 50:57 one. The third one though is your own 50:58 team. So what you find on people teams 51:01 is everyone is peoplecentric. But the 51:02 same temptation to become more team 51:04 versus company is still there within 51:06 people organizations. So territory 51:09 territorial approaches for example 51:11 learning and development and HR business 51:13 partners are a good example of teams 51:14 that often and talent development teams 51:17 they tend to have conflict and part of 51:19 what you have to do is be an active 51:21 proponent of helping squash that. 51:23 >> You had this very long stretch at 51:25 HubSpot of call it 11 years. It seems 51:28 like a very formative time in your 51:30 career and you formed all sorts of 51:32 important ideas. Have you changed your 51:35 mind on any of those important ones 51:37 since joining Harvey? 51:39 >> HubSpot was absolutely formative and I 51:41 think one of the things that we did 51:43 incredibly well as a team was that focus 51:46 on culture as a key differentiator for 51:49 the company. So for example, we had 51:50 people at HubSpot who knew more about 51:52 our culture and employer brand than they 51:54 did necessarily about our product. And 51:56 so part of what I wanted to do at Harvey 51:58 was a little bit do something slightly 51:59 different, which is to say we have a 52:02 great culture. I'm very proud of the 52:03 culture that we're building, but 52:05 ultimately at the end of the day, we 52:06 want to be the leading AI platform for 52:07 lawyers. So, I want our product and our 52:09 brand to lead and our culture to follow. 52:12 And so, I think I don't know that that's 52:13 necessarily changing my mind, but it's 52:15 the focus on the work and what we're 52:16 building and then investing in building 52:18 our brand first and foremost. I think 52:20 that's probably one. And then I think 52:23 secondarily, the degree to which pace 52:25 matters. So, I would have said at 52:26 HubSpot, we have to match the like going 52:29 fast and going slow on certain 52:30 decisions. I don't think you have that 52:32 luxury. So the time compression I would 52:34 say at Harvey has been a big thing. I 52:36 don't think any opinions I have with 52:38 regard to like how to run an 52:40 organization have meaningfully changed 52:41 other than the need for for rapid speed 52:44 and then making sure that you lead with 52:46 the brand and your customer value above 52:48 the company and culture. When you think 52:50 about what's happened in AI and called 52:52 the last bit of time post chat GBT and 52:55 maybe even most recently over the post 52:57 last 6 months where it feels like 52:58 there's been just such a step function, 53:00 what do you think if anything has 53:02 changed as it relates to the people 53:03 function? Not you think in the next 5 53:07 years it was X and it's going to be Y, 53:08 but like right now if you're running a 53:10 people or do you think there have been 53:13 really important changes? 53:15 >> I think you said a few important things 53:17 there. One is the inflection point you 53:19 you mentioned even in December I think 53:21 what you saw from the model outputs was 53:24 oh boy things have already changed from 53:26 even 6 months to a year ago given that I 53:29 think you're seeing two things one is 53:31 for non-technical folks the ability to 53:35 use for example claude and meaningfully 53:38 automate or create systems that 53:39 otherwise would have taken you an hour 53:40 is hours is pretty staggering and I 53:43 would say anyone who is missing that as 53:46 like an a holy cow moment is missing the 53:48 point. But secondarily, I think part of 53:50 what you see is if the models are 53:52 improving that much, then any AI company 53:55 needs to improve the operations of the 53:56 company at that same pace or beating it. 53:59 And so part of what I think you need to 54:01 do is go, okay, if we were for example 54:03 getting 20% better at Harvey, we need to 54:05 get 60% better because that's what the 54:08 model output. So it's become a new 54:09 threshold for quality and for pace and I 54:12 think it drives a lot of the urgency 54:13 around what application layer companies 54:15 are building. It certainly is driving 54:17 urgency within Harvey to make sure we're 54:19 keeping pace with where the industry is 54:20 going. 54:21 >> So, but like if you look at the people 54:22 or specifically if I were to look at how 54:24 you're running it, what's the tooling 54:26 and infrastructure? What is your 54:28 expectation for a director of people 54:31 versus four years ago or 6 years ago 54:33 when you were the same size at HubSpot? 54:35 What's the contrast in your mind? 54:37 >> So, one thing you're looking for is as a 54:40 leader, what process are you eliminating 54:42 or fundamentally changing with AI? So, 54:44 I'll give you an example. 54:46 Yesterday we launched a process where 54:48 people kick off all of their job wrecks 54:50 to open up. So for example, let's say 54:52 you're opening up a wreck. Writing a job 54:54 description used to sit down, you'd have 54:55 a cup of coffee, draft it. We use our 54:58 vault product in Harvey to do that on a 55:00 regular basis. It's repeatable. If you 55:02 are not thinking how could AI solve this 55:04 problem first and eliminate hours of 55:05 work within your team, you're not going 55:07 to be a successful leader in the people 55:09 function in the next year and a half. So 55:11 in other words, if you dabble regularly 55:13 in one of the models occasionally for 55:15 personal use, but you're not using it at 55:17 work, you are already way behind and 55:19 need to catch up. The best people 55:20 leaders, I think, are leaning into 55:22 eliminating or reducing processes using 55:24 AI. Second thing is just thinking about 55:27 what we can no longer be wed to. So a 55:29 good example would be we have as part of 55:31 our onboarding process tried to 55:33 eliminate as much friction as possible. 55:35 So in other words, what I would view as 55:36 successful is can you take a 15step 55:39 process and consolidate it. Before if 55:41 you got it down to 12, that would be 55:42 good efficiency. Now it's like can you 55:44 get it down to three and where can AI 55:46 play a critical role so that we're just 55:48 focused on the onboarding process and 55:49 creating a really human experience. So 55:51 that's where you're infusing the 55:52 hospitality versus the others. Finally, 55:55 I would just say as you think about the 55:57 headcount planning, you need to be 55:59 thinking about the work units to be done 56:00 and the skills needed for that and 56:02 whether or not it should be people or 56:03 agents. And I think that's a critical 56:04 part of the conversation that people 56:06 leaders should be leading not following. 56:08 >> What are some of the things that you 56:09 think are now possible that you had 56:11 always dreamed of doing as a people 56:13 leader if there is anything? 56:15 >> So I'll give you an example. We are 56:17 currently in the middle of performance 56:19 reviews and one of our team members just 56:21 on her own wrote a script for completing 56:24 our performance review in a matter of 5 56:25 minutes and a bunch of other people 56:27 tried it and they were like this is the 56:29 process that used to be so ownorous took 56:31 a lot of self-examination and it was 56:33 completed in 5 minutes. Performance 56:35 reviews are just like so incredibly 56:37 painful and if you can make those much 56:39 less painful I think that's a huge win. 56:41 I think next year AI will be critical in 56:43 the workforce planning side of things 56:45 and that's always felt like pulling tea. 56:46 So that's a big one. And then third, I 56:48 would just say the building of decks and 56:50 internal comms. That's something that I 56:52 used to spend I mean hours if not days 56:54 on. And I think the reality is that will 56:56 end up being much more AIdriven even in 56:59 the next 2 months um versus the next 57:01 year as well. 57:02 >> Like right now I mean I assume if you're 57:04 doing a companywide communication, you 57:05 may be getting editing advice or things 57:07 like that, but like where do you think 57:09 it ends up going 57:10 >> on internal comms? you used to sort of 57:12 do like I used to for example do a loom 57:14 for my team every single Friday and it 57:17 was highly viewed at HubSpot but part of 57:20 it is people's attention spans are just 57:21 going way down so I think what you'll 57:23 see is in the next 6 months instead what 57:25 you'll do is have a level of 57:26 customization you're getting something 57:27 that is relevant to you by team by level 57:30 and it's all automated based on AI's 57:33 understanding of where you are in the 57:34 organization what you need to hear I 57:36 also think you'll see more nurturing 57:37 based on what you've already consumed or 57:39 not it's going to get a whole lot 57:40 smarter so for example Right now, 57:42 everyone gets the same exact, hey, you 57:44 haven't completed your performance 57:45 review, even if you have. And I think 57:47 part of what you'll see is people going, 57:49 hey, last time around it took you 35 57:50 seconds to finish your performance 57:52 review. This is what we want to see from 57:53 you, and here's your customized report 57:54 this week on things you open and engage. 57:57 Oh, by the way, here's the gong call 57:58 from the most successful sales rep so 58:00 that you can you can actually study it 58:02 yourself. And oh, by the way, here's the 58:04 best plugin on cloud code, for example, 58:08 that someone used that was super 58:09 effective for them. What's your current 58:12 strongest held belief about the next few 58:15 years in terms of the people stuff that 58:16 is not going to change? 58:18 >> I think the importance of EQ is not 58:22 going to change. And so I think the if 58:24 you have to worry less about all the 58:26 things that people have always faded 58:27 about being a manager. So ownorous 58:29 performance reviews, ownerous internal 58:31 comms, ownerous ACR processes. If that 58:33 gets easier, ideally you have more time 58:35 for the things that make a great 58:36 manager, which is a strong level of 58:38 empathy, but also a level of actually 58:41 training people to do their jobs and 58:42 work. So ideally, you see a return to a 58:44 little bit more mentorship. Most people 58:46 I know don't have a manager that spent a 58:48 lot of time actually getting them ready 58:49 for the next stage of their career. And 58:51 so I hope that that becomes the case. 58:53 Number two, though, is I've always said 58:55 that the people team needs to set the 58:57 pace. AI adoption is another area where 58:59 I think people leaders need to set the 59:01 pace. And so I think you'll see that as 59:03 well. And then I think the other thing 59:04 that won't change is the importance of 59:07 emphasizing what you care about and what 59:08 you don't. So as an example, I think as 59:10 we see engineering is a good example of 59:13 something that I think is going to 59:14 change tremendously not just in the next 59:16 two years but the next two weeks it 59:17 seems like given that what do you care 59:20 about? Are you hiring junior engineers? 59:22 Are you not? How do you make sure the 59:23 best people are focused on the most 59:25 technical problems and what does your 59:26 resourcing look like? I think having 59:28 honest conversations about that so 59:29 people are clear on what they're signing 59:30 up for is going to be important. 59:32 >> Do you think what a great VP of people 59:36 looks like and does in a year or two 59:38 looks relatively similar or very 59:40 different than today? I think that 59:42 there's a reason that for example Carmel 59:45 Galvin at Clavio, 59:47 Jackie at Service Now, there are 59:49 multiple chief people officers who are 59:51 playing like chief AI officer, AI 59:53 transformation roles. I don't think 59:55 that's an accident because I think the 59:57 reality is AI is a technology, but it 60:01 also requires a ton of change management 60:03 and human behavior. And so I think what 60:04 you're seeing is the best chief people 60:06 officers understand what it takes to get 60:08 people past fear on the change curve. 60:10 They understand what it means to put 60:11 champions up to actually make sure we're 60:13 rewarding the behavior. Your incentive 60:15 should be aligned to using and adopting 60:17 AI in meaningful ways. Not just in a 60:19 like I tried this and it didn't work 60:21 kind of thing, but actually to transform 60:22 how you run your teams. And I don't 60:24 think it's an accident that the 60:26 forwardleaning people in AI are getting 60:28 rewarded for that with additional 60:30 titles, responsibility. I do think that 60:33 people totally misunderstand that when 60:34 it comes to rolling out and adopting AI, 60:36 we see this at Harvey. We sell to 60:38 lawyers. Lawyers are not the people that 60:40 you think of as most traditionally to 60:42 lean into AI, but what we've seen is 60:44 that lawyers are actually pretty eager 60:46 because they're sick of doing document 60:48 review. Nobody goes to law school to say 60:50 like, I've always wanted to redline 60:51 documents for a really long time. And so 60:53 part of what you have to figure out is 60:55 how you get them on the change curve to 60:57 consider adopting AI for the tasks that 60:59 they absolutely hate and dread. And how 61:01 do you appeal to them? That includes 61:03 change management. That's not just our 61:04 platform getting better. That's going 61:06 into an organization, understanding 61:07 their special sauce. And I think the 61:09 reality is what we see is the best 61:11 people are really good at doing that and 61:13 understanding what it takes to get 61:14 through to an organization to champion 61:16 people. People leaders will be the same. 61:18 >> What is the role of being demanding as a 61:20 manager? 61:21 >> I think it's a huge part of the job. I 61:24 think you can be liked or you can be 61:26 respected and you have to pick a lane. 61:28 And the reality is, I think the best 61:31 players, the best teams, all the 61:32 champions that we admire are much better 61:35 at being admired than liked. And you 61:37 have to be okay with sacrificing some 61:39 level of congeniality, comfort to get 61:43 people to deliver their best, the best 61:46 work of their lives. And so I think most 61:48 people who work for me would say I'm 61:50 incredibly demanding. They would also 61:51 say that I'm incredibly loyal, caring, 61:53 and kind. And I do I think you can be 61:55 both. 61:55 >> Explain. Most people think that those 61:57 two things are at odds and particularly 61:59 when you talk about one of the things 62:01 that is likely to survive what is 62:04 happening in AI is human empathy. I 62:06 think when a lot of people hear that 62:08 again you go back to the coddling of 62:10 someone and it almost feels antithetical 62:14 to being demanding or having insanely 62:15 high expectations and it feels like part 62:17 of your management philosophy is 62:19 bringing those two things together. 62:21 >> Absolutely. I think the best leaders or 62:24 coaches demand more of you than you ever 62:27 thought was possible. And I still think 62:29 they're the first person to come around 62:30 if you're having a tough day and check 62:31 in on you as a human being. And so I 62:34 view those not as at odds with one 62:36 another, but in fact that the best 62:37 leaders have to be able to hold place 62:39 for both of those. And so what I always 62:41 say to people is I think if you're a 62:43 people leader, you have to be a queen of 62:44 awkwardness. And you have to be willing 62:46 to sit in really awkward conversations 62:48 and just say do you think that was your 62:50 best work out there and to sit in the 62:53 awkwardness of going like it wasn't and 62:55 here's why and here's why I think you 62:57 can do better and I have higher 62:58 expectations of you. How do we chart 63:00 that course together? That is really 63:02 awkward. It's so much easier to just say 63:04 you did an amazing job out there. And so 63:06 being willing to sit in that awkwardness 63:08 and what I try and do is remind myself 63:10 regularly that first of all the best 63:12 coaches I ever had growing up were 63:14 people that demanded that of me. And 63:16 then on the flip side the best 63:17 compliments I ever get are from people 5 63:19 years down the line who say remember 63:21 when you gave me that feedback that made 63:23 me better and I think about it often. 63:24 And so I think you have to remind 63:26 yourself of that versus just that desire 63:27 to be liked. So, what what's sort of an 63:31 example of what it feels like to sort of 63:35 work for you 63:37 and have that sense of both being pushed 63:42 really hard and also having the level of 63:45 support that you aspire to give someone. 63:48 >> I hope it feels like being part of the 63:51 best team you've ever been a part of. 63:53 And it feels equal parts challenging 63:57 and really really hard and rewarding and 64:01 worth it. And so I think I hope that on 64:03 any given day on my team that if people 64:05 are working 5 days a week that you know 64:08 who who cares how you divide it up but 64:09 that three days feel really challenging 64:11 and hard and two days just feel like 64:13 you're completely winning and crushing 64:15 it. I think that balance is really 64:17 important. So one of our values at 64:19 Harvey is jobs not finished and that's 64:21 really intentional. We want people to be 64:23 super clear on how they could be better 64:25 and to know the difference between 64:26 what's good and what's great. And part 64:28 of what I do with my team is I share my 64:30 performance reviews. I'll share with 64:31 them here's what I said I was good at 64:33 and here's what I said I could do 64:34 better. And I think part of it is them 64:37 knowing I'm not just asking that of 64:38 them. I'm I'm expecting it of myself. 64:40 And I think that then it's easier to 64:42 open the door. If your CEO and COO are 64:46 sharing a long list of things they can 64:47 do better and you can't think of one 64:49 thing you're doing wrong, I think that's 64:52 a self-awareness opportunity. And so 64:53 part of what I try and do is lead by 64:55 example. And so I've shared my 64:56 performance review with my team with 64:59 elements redacted for the better part of 65:00 the last 8 years. And so I think one of 65:02 the things people would say is I'm super 65:04 transparent around things that I could 65:06 be doing better. And then I'm also super 65:07 transparent with you about things you 65:08 could be doing better too. In that 65:10 theme, what's the most important thing 65:11 you've improved on in the last eight 65:13 years? 65:14 >> One was I overtalked. So when I first 65:17 got promoted to be CPO, I thought my job 65:20 was to say something important at every 65:22 meeting. That is not your job. Your job 65:24 is to know when you add value and when 65:26 you don't. And so I overtalked and that 65:28 was really tough feedback to get early 65:30 on was in your new role, it feels like 65:32 you're trying to grab the mic versus ad 65:34 to the conversation. and your quality of 65:36 value ad versus words per minute is 65:39 going down, not up. That was tough to 65:41 hear, but incredibly important. And it 65:43 taught me to really sit back and only 65:44 speak up when I had something valuable 65:46 to add. Second thing was I had a really 65:49 big failure. We were really, really, 65:52 really behind on recruiting at HubSpot 65:54 many, many, many years ago. And I 65:58 started pointing fingers left and right. 65:59 I was like, "It's this team's fault. 66:01 This team's fault. It's this team's 66:02 fault." And Brian had a really hard 66:04 conversation with me, which is you are 66:06 either going to be here in a year 66:07 because you own this and dug your way 66:09 out of it or you're going to go down 66:11 pointing fingers at other people. And so 66:12 I learned a really hard lesson on just 66:15 owning your own mistakes, but also at a 66:17 certain point it doesn't actually matter 66:18 whose fault it is. It matters that it's 66:20 within your team. So that level of 66:21 ownership as an exec and just being the 66:24 person who in front of the board is 66:25 willing to say that was me is a big one. 66:27 I think the third thing is letting go of 66:29 perfection. I was very much someone who 66:31 is an achiever and like doing a great 66:33 job and like getting, you know, good 66:35 grades. And I think part of it was also 66:37 just the embracing of like when you're 66:39 an executive, you know, your your goal 66:42 is to like bat 400 at at best. That's 66:45 like what best-in-class people are 66:46 doing. And so getting more comfortable 66:47 with like, yeah, you're right. That 66:49 meeting was a miss could have been 66:50 better. And just getting more 66:51 comfortable with not always knocking it 66:54 out of the park. 66:55 >> What does it look like when you disagree 66:57 well with a CEO? I think it starts with 67:00 a foundation of respect. When Winston 67:01 and I are talking, even in the most 67:02 heated debate, there's a level of 67:04 respect there and then it feels like 67:06 you're lining up on the same side of 67:07 like you're still on the same team even 67:09 when you vehemently disagree. I also 67:11 just think it's understanding that you 67:13 have the same long-term goal. So, for 67:15 example, on the things about about which 67:18 Winston and I disagree the most, we 67:20 never have a disagreement around where 67:22 we're going. It's always just about how 67:23 we get there. And I think it's that 67:25 understanding. I also just think it's 67:26 having a mechanism for it. So, one of 67:28 the hard conversations we've had to had 67:29 to have is he's very comfortable with 67:31 conflict and so am I. But I need more 67:33 time to process my own role in it. So, 67:35 for example, if I'm the one at fault, I 67:38 need a night to sleep on it to go, I 67:40 have to understand why this conversation 67:42 bothered me so much and I need to 67:43 process then discuss it. He would always 67:45 rather just like just yell at me and 67:46 we'll figure it out. And so, we've had 67:48 to work through how we disagree, not 67:50 just what we disagree on. 67:51 >> Do you ultimately think the CEO has to 67:54 make the final decisions? I don't think 67:56 the CEO has to make all the decisions. I 67:58 think part of what you have to agree on 68:00 is who gets to make the call when. So 68:02 Winston is really clear on, hey, you're 68:04 the decision maker on these five things. 68:06 I'm the decision maker on these other 68:07 things. And I think we're clear on that. 68:09 And then when we're not clear on both 68:10 what the decision should be and who the 68:12 decision maker would be, those are more 68:14 complex conversations. But I think 68:16 they're some of the most important ones 68:17 we have. 68:17 >> What's an example of like one of those 68:19 where it's really nuanced and not clear 68:21 and easy? 68:23 So I would say often times it's about 68:25 the pace at which we do something and 68:27 the pace at which we can expect change 68:29 and whether that's hiring or open up new 68:31 markets or deciding whether or not to 68:33 turn over a team. often times I'm the 68:36 one saying we need a little bit more 68:37 patience and time and he's the one 68:39 saying no we don't have any time and so 68:41 I think part of it is just agreeing like 68:43 as a good example what we have agreed to 68:45 is like on fundraising 68:47 his call CEO's job his call on the board 68:51 meeting like he is the person who gets 68:53 to make the call because they are his 68:54 relationships to manage and I'm of 68:55 course supporting him what that looks 68:57 like if we are making a big culture 68:59 decision we'll make it together but he 69:01 will ultimately defer to me on a lot of 69:03 recommendations so long as he doesn't 69:04 want to lay on the tracks on it. Part of 69:06 what you have to do is agree on what 69:07 those parameters are. And I think we've 69:09 done a decent job at for the most part. 69:11 I would say 90% of things we agree on 69:13 both the decision maker and how we'll 69:14 get there. 69:15 >> What's the the advice you have? Somebody 69:17 step stepping into the chief people 69:19 officer role for the first time. In a 69:22 lot of cases, their experience with the 69:24 board is a net new thing. Maybe every 69:27 now and again a VP of people comes in 69:29 and presents something to a board, but 69:31 the dynamic of a chief people officer 69:32 with a board is just very different. 69:35 What are things to keep in mind or 69:37 pieces of advice that you would give 69:39 somebody stepping into the role 69:40 specifically as it relates to boards? 69:42 >> I hit the lottery on this one. So at 69:45 HubSpot, I had I think presented to the 69:48 board one time before I became the chief 69:50 officer. So to your point, I would have 69:51 fallen on my face badly. We had a board 69:54 member named Lorie Norington. She's 69:56 still there. She's a board member at 69:58 some of the best companies in the world. 70:00 And she basically saw that I was getting 70:02 appointed into a role that was a stretch 70:04 for me. And she set up that I would go 70:06 to eBay and shadow their CHRO at the 70:09 time and shadow and meet their people 70:10 team. And so the first thing I would do 70:12 is rather than trying to go in all 70:14 confident and pretend you know what 70:16 you're doing, ask for help. Your board 70:18 is usually very willing, especially 70:20 given the relationship between the chief 70:21 people officer and and especially the 70:23 compensation committee. they have 70:25 incentive for you to succeed and for you 70:27 to have a personal relationship. And so 70:29 I would say part of it is just ask for 70:30 help and ask for exposure. Ask them to 70:32 meet the best chief people officer you 70:34 know. That is often the way in which you 70:36 get the best context on what resonates 70:38 and what doesn't versus falling on your 70:39 face. Number two is ask to observe 70:42 before you present. So I think one of 70:44 the best things I've seen people do in 70:46 the board work that I've done is kind of 70:47 go, hey, it's my first time around. I'm 70:49 just going to take in the cadence here 70:51 and then for next time around I'll come 70:52 in prepared. Third is just the meeting. 70:55 The meeting is just a meeting. Part of 70:57 what you have to do is build the 70:58 relationships before and after. So for 71:00 the most important decisions we made at 71:01 HubSpot, the board presentations, there 71:04 was never a time when the board was 71:05 seeing something for the first time that 71:07 I was presenting. I had preeded the 71:09 conversations and big ideas. I had 71:11 solicited their opinion early on on 71:13 recommendations. I had asked what other 71:15 portfolio companies they were working 71:16 with had done as a best practice. And so 71:18 part of it is realizing it's not 71:20 actually about the meeting. It's 71:21 actually about your ongoing relationship 71:23 with the board and it took me a while to 71:25 learn that your your job is to manage 71:27 that dynamic and to interact with them 71:29 regularly. So the meeting is just 71:30 another touch point. 71:31 >> What about you hinted this a little bit 71:34 but what is different about the CPO role 71:38 and its working relationship and 71:40 responsibility with the board relative 71:42 to a CRO or uh a CTO or sort of any 71:47 other executive function. 71:48 >> I mean it's so different. you have at a 71:50 public company you have a few pretty 71:52 critical roles. So number one the 71:56 compensation committee nom and gov both 71:58 you have pretty deep relationships with 72:00 and oftentimes you have a core 72:02 responsibility to both on 72:04 recommendation. So, for example, the 72:06 compensation committee recommends the 72:08 comp for your entire executive team, 72:10 including your CEO, and you're the CPO 72:12 who's in the middle of that discussion. 72:14 That's a little awkward to manage. And 72:15 so, part of what you have to do is have 72:17 a really deep relationship and trusted 72:19 relationship with them and with your 72:21 compensation consultant, who is an 72:23 independent adviser to the board. Second 72:25 though is you manage succession 72:26 planning, which depending on the board, 72:28 is either a nom comp depending on what 72:30 it looks like, but usually a nom. you 72:33 have to have really awkward 72:34 conversations with them around the 72:35 readiness of the leadership team for the 72:37 next level of growth, for the next level 72:39 of global expansion, for multi-product, 72:40 you name it. And so I would say if 72:42 you're the CTO or CRO, you of course 72:45 have an obligation to the board and your 72:46 relationships matter. But as the CPO, 72:49 you have kind of an awkward dual 72:50 relationship where you have a duty to 72:52 your investors, especially at a public 72:54 company, to be super thoughtful about 72:56 the balance. And so I think it's really 72:57 navigating and understanding the nuance 72:59 there is critically important. I guess 73:01 maybe to wrap up, what is it that you 73:04 want the the people on your team to say 73:08 about you behind your back? 73:10 >> I think I want the people on my team to 73:12 say there's no one who wants us to win 73:14 more as a company and there's no one 73:17 who's pushed me harder to get there 73:18 while still having my back. 73:20 >> Good place to end. Thank you so much. 73:22 >> Of course. Thank you for having me.