Yamini Rangan，HubSpot CEO：HubSpot 如何与 Salesforce 竞争

When you’re in wartime or a hard time, you’re optimizing for speed of execution. When you’re in peacetime, you’re optimizing for getting a consensus around what you’re trying to do. You’ve got to be good at both, and many other modes of operating.

B2B apps are not winner-take-all markets. Whether you look at HR, finance, recruiting, or CRM, they all have multiple players. The best competitors make you better, and I think Salesforce does a really good job of that.

Yeah, listen. I spoke to Brian, I spoke to Pat, I spoke to Jay Simons, Claire Hughes Johnson. I think I stalked the shit out of you, bluntly, before this show. I know—terrifying—but thank you so much for joining me.

Thank you for having me, Harry. I’ve been a longtime listener, first-time talker to you.

1. Taking Over the CEO Role from the Founders

That is fantastic. I want to start with the movement into the role of CEO. It’s a $31 billion public company. The size of HubSpot is incredible. The question that I want to start with is: the founders are still fairly involved. What’s it like being CEO when the founders are still fairly involved?

It is a privilege and an honor to take over from a founder, and it’s hard and tough work to make it work. For me, I had been at the company for about a year, and I always looked at the decisions that Brian and likely Dharmesh made in starting and scaling the company: deep conviction in SMB, deep conviction in building a platform and crafting the platform, and deep conviction in having culture as a product. Those are decisions that I just loved and completely believed in.

That is the easier part, and that’s the pro of working with founders who know the company, who have taken it from a very small startup to scaling it to a public company and beyond. On the flip side, it’s a lot of work. When the transition happens, you have to think about a lot of things during the transition and throughout what happens afterward.

What’s the cadence? Do you want to send 20 emails a day or 2 emails? What’s the altitude? Do you want to talk about the strategy and the product, or do you want to talk about the whole business? If it’s one or the other, you really need to make sure that you’re covering the bases across all of those conversations.

Then what do you do when something bad happens, which something will happen? If you hit a pothole or if you make a bet that does not work, what happens? How are you going to make it work? There’s a lot of work that you need to put in. Just like any relationship that you have, first of all, you have to commit to it, and then you’ve got to put the work in to make it function. So, yeah, there’s a lot of work.

Oh, that’s why I’m single. Now I get it.

You said about the deep conviction there across—I mean, it’s amazing when you look at the deep conviction to, as you said, SMB, the GTM, doing CRM all from nothing when it was such a mature space. Are you able to have such deep conviction as a public company CEO when the Street does react very strongly to decisions?

You have to have deep conviction in order to run a public company, and if those decisions and the choices that you make are counterintuitive, even better. You’ve got to forget that you’re a public company and start with the basics of what you believe in.

One of the things that I just came from was talking to investors all day. We had a whole non-deal roadshow today, and we spent a lot of time talking to investors. They actually appreciate your conviction in your business. That’s what they like to hear. When we talk to them about why we make these decisions and why we focus on the segments that we do, they get our story.

You said the more counterintuitive, the better. What do you think is the most counterintuitive decision that HubSpot is making today that may not be obvious?

A couple of things come to mind. I would say pricing, and we’ll talk about AI pricing. When AI came on the scene in November 2022, we fairly immediately moved into product building. We changed our roadmap—we actually stopped our roadmap, changed the roadmap, and started building product.

2. Wartime vs Peacetime CEOship

Since that day, people have asked, “How are you going to monetize it? What are you going to do with the monetization?” We have a deep conviction that there is one product, which is an AI product, and it’s going to be built into HubSpot and every part of the platform. That means monetization is the entire platform.

Harry, I can’t tell you: every day I get asked this question about 10 times. “How are you monetizing it?” We’re monetizing it with the current product, and that’s how we plan to build.

I think it’s counterintuitive because today everybody has an AI SKU. They slap a premium on that SKU, or they put a usage metric or an outcome metric on it, and there’s just a lot of swirl about how you price products. But I think the first thing that you’ve got to do when a big technology shift like this happens is focus on customer value.

I’m a fan of making bold statements unsupported by data because I’m a venture capitalist. When you look at outcome-based pricing, it is impossible unless it is completely non-human-in-the-loop, because otherwise there’s always going to be this continuous debate: “Well, actually, Yam, I did 80% of the sale and your agent only did 20%, so I don’t want to pay 100% for your outcome.”

How do you think about how quickly we’ll be able to move to outcome-based pricing, and whether that even is the solution for the next generation of AI?

There’s a lot of swirl, isn’t there? Is it seats? Are seats going to completely go away? Is it outcome? Is it some form of consumption, credits, or usage?

If you’re a horizontal-based product like HubSpot, you have some customers that will use you for LPs—for example, a fund, which is a very high ticket—versus a florist, which is a very low ticket. You can’t just have the same outcome-based pricing.

We believe that, in the fullness of time, we’ll have hybrid pricing. We’ll have some seat-based pricing and some usage-based pricing. I think it’s really hard to think about outcomes for exactly the reason that you’re talking about: how can you validate an outcome?

Especially if you offer a platform that, in one spot, can resolve a ticket; in another, can help you set up a meeting; and, in another, can write a blog or a social post for you, the outcomes are so different. How are you going to validate all of it?

I think the simpler way, at least for SMB customers, to think about it is results, and results are more about usage. Are you using a certain credit? Are you using something in our AI platform? It’s a credit-based mechanism rather than an outcome-based mechanism. That’s how I think it’s going to evolve.

I do want to go back, just before we move on to the leadership element, where we mentioned that some elements of the transition were hard and some elements were easier. Can you describe one thing that you’ve changed at HubSpot culturally or strategically since becoming CEO that might not have happened under Brian’s leadership?

I want to know who asked that question, Harry, but I will hold it. That’s my enterprise version.

3. How to Scale Into Enterprise: What Everyone Gets Wrong

I think we’re going from a scale-up to a grown-up stage. One of the things at HubSpot that I heard when I took on the role was that we don’t like process. Process is a four-letter bad word. We don’t want to talk about process. You need a certain amount of process as you grow and become a bigger, distributed company across multiple offices.

One of the things that I’ve done is really focused on tying strategy to priorities to outcomes. When you run a large, grown-up company, clarity around what is important is really needed. Strategy is just that: what’s the most important work that we do as an organization? The second part is priorities off that strategy: what are you going to do this year that’s a priority? The third part is what outcomes do you commit to based on the work that you’re doing this year, and are you accountable for those outcomes?

This strategy-to-priorities-to-outcomes framework is something that I’ve obsessed about for the last couple of years.

Just so I get an example of what strategy means, what is the strategy?

The strategy is threefold, Harry, and super simple. First, focus and obsess on our customer. Everything starts with the customer. Second, make our product easy, fast, and unified. Third, have a growth mindset and become grow-getters ourselves.

That structure of going from strategy to priorities to outcomes is something that you need as you scale and become even bigger and more distributed. I can’t speak for Brian—Brian is awesome, and he probably would have done a lot of things. I don’t know if he would have put a framework like that, with the process that we’ve driven.

Can I ask you about “have a growth mindset”? I always think you want something where someone can take the alternate side. For example, I believe working unreasonably hard yields outcomes. Work-life balance is a complete myth, but many people disagree with me, and that’s fine. No one wants someone who doesn’t have a growth mindset, do they?

That’s not true. I’ll push you on that.

People love the comfort of what they’ve done every single day. Look at AI, which is coming. One of the things that we’re talking to the whole company about is: can we become AI-first ourselves? Can we learn how to write prompts? Can we start using AI every single day?

The answer is not everybody wants to do that, because they’re comfortable. Some of us are comfortable doing whatever we’ve been doing for the last decade. I think the alternate to a growth mindset is comfort in how you do what you do, and maybe even comfort in the expert level that you have based on what you’ve done for a long period of time.

Asking questions, being curious, learning, and adapting—that is not a given in any company. It’s not even a given in the mindset. In the larger you become, you do have people who like the comfort of, “I’ve done this before, I’ve done it a certain way, I know how to do it, and I’m very comfortable doing it the way that I’ve done it before.”

I totally agree with you, and I think that’s absolutely fair. I probably just don’t like those people, but you’re right. That is, I’m sure, a large portion of large-company workforces.

Can I ask: what were you comfortable with in how you did things that you’ve had to change with the leadership change and becoming CEO? It’s a big transition. What did you have to get comfortable with that you weren’t comfortable doing before?

One of the biggest changes when you go from any other position into a CEO role is the number of stakeholders that you have. You now have the board, investors, founders, and a broader employee population across a lot of teams. The number of key stakeholders who judge you, focus on what you’re doing, and look at you is just broad.

What I got comfortable with is that, no matter how hard I try, I’m not going to perfectly hit the approval of all those stakeholders. At some point, I’m going to do something that’s super beneficial to customers. At some point, I’m going to do something that employees are happy about. At some point, I’m going to do something that board members are satisfied with, and maybe investors are not.

You’re never going to hit that perfect equilibrium where you’re satisfying, and maybe even overachieving, the expectations of all of your stakeholders.

How do you make decisions? That’s the job. How do you think about decision-making when you have so many different stakeholders? Do you think from a downside-protection mindset, an upside-maximization mindset, or something else?

I think you have to have some very strong north stars. What are you optimizing for? Within HubSpot, we’ve said, “Solve for the customer,” or SFTC. Solving for the customer is the north star.

If there is ever a decision between whether we’re solving for the company or the customer, it’s always the customer. There’s something that we say: customer over company, over team, over self. That is a guiding framework and principle, which means that when in doubt, always do the right thing for the customer, then for the company. You shouldn’t be solving for yourself.

There are guiding principles in terms of how you think about decisions, which will shift things one way or the other, but you’re still not going to get every one of those decisions right.

When you’re a $31 billion public-company CEO, I mean, respectfully, it’s almost like you’re a politician in a way. You have so many different constituencies and you have to make them happy. There are so many different offices and people. To what extent does it sometimes feel like that, versus the belief that resource allocation is the sole job, or the main job, of a CEO?

Politician? No, I don’t want to be a politician. I don’t think of my job that way. I would not do this job if I had to be very political. In fact, I want none of the politics or those considerations in terms of how we make decisions.

I took on this role as CEO of HubSpot because I think we have such a big opportunity. The mission of what we’re trying to do—to help small and medium businesses grow—is so personal to me, and it’s so motivating and inspiring. That is the focus for me.

In terms of how you make decisions on a given day and how you pivot, you said it comes from resource allocation. I actually think about it as resource alignment. Have you heard the term “culture eats strategy for breakfast”? I said to our team, “Alignment eats strategy for lunch.”

What I mean by that is that the biggest job—job number one—is setting strategy. But I think what’s even more important and critical in terms of a CEO is aligning all of the resources to accomplish that strategy. As teams grow and become distributed and much bigger, vectors are totally not aligned.

Can you help me understand where they were misaligned, where you aligned them, and what the outcome was?

Think about this, Harry: you have a sales organization that’s optimizing for this quarter. You might have a customer-success organization that’s optimizing for customers staying and not churning. Those 2 things are not always the same. Then you might have a product organization that’s trying to launch a product.

How do you align every one of these functions that are trying to do what they think is right? They’re trying to optimize for subsystems. How do you get them to align on the system?

Most of the work that gets done in a larger organization is cross-functional in nature, not functional in nature. Most organizations grow up with a functional mindset: what am I doing in my team, what am I doing in my function, and how do I optimize?

The alignment is really about whether we can be clearer on what moves the needle for the company. I gave you an example of our strategy earlier: we want our product to be easy, fast, and unified. That’s not just the product team’s job.

It is the job of marketing to communicate that value proposition to our customers. It is the job of sales to be able to demo that and help our customers understand it. It is the job of customer success to make sure that we’re delivering on it. Those are very cross-functional priorities.

If I go to our team and say, “The number-one priority for HubSpot today is to make our product easy, fast, and unified,” that aligns the entire set of subfunctions to optimize for the outcome that we want to drive based on that strategy.

In most organizations, it’s not just about resource allocation. When you allocate the resources, is there duplicate work? Is there clarity on what the resources are going to accomplish and in what period of time? Do we know what the goals are? If something does not work well, is there clarity on who takes accountability to get the right outcomes? That’s the job. Once you set that up, you begin to see that work.

When I think about alignment, I think about facing a challenge, bluntly, and aligning against it. It makes me think of Jay Simons, who I spoke to last night, and he said I had to ask you about wartime versus peacetime CEOship. How do you reflect on wartime versus peacetime CEOship?

First of all, I love Jay Simons. He’s amazing. He’s been on our board for a long time, and he’s added so much value. We go to him to soundboard almost any part of the strategy.

I don’t know if I like the words “wartime” and “peacetime.” There’s real war going on, so I don’t know if I want to use those words. But I think what you’re getting at is good: are there different modes that a CEO needs to operate in, maybe in hard times or not, with the modes optimized for different things?

When you’re in wartime or a hard time, you’re optimizing for speed of execution. When you’re in peacetime, you’re optimizing for getting a consensus around what you’re trying to do. Those modes are completely different.

The question is whether you should be good at one or the other. The answer is that you’ve got to be good at both, and at many other modes of operating. There are years where you get into the year and you’re like, “This is go, go, go.” You’re optimizing for speed, which means we’re going to be very clear in terms of what we want to accomplish this year. We’ve got to get people to follow. That’s the north star for this year, and we’re just going to go.

That’s completely different from when you need to enroll everybody in your strategy. You’re doing roundtables, getting everybody to align on what the priorities are, and the mode of execution is really different. You’ve got to do both. Sometimes you’ve got to change the mode within a year or even in a given quarter.

You said speed is important there. Can you take me to a time when you moved very fast, maybe should have moved slower, and what you learned?

There’s a moment where I know something in my head and it’s like, “Do I say that story?” I laugh, Harry, because I’ve been getting performance reviews forever and ever. The consistent theme across 25-plus years of my performance reviews is, “Go slow,” because my natural mode is really going super fast.

Very early on, I was in sales and had just gotten promoted to a sales-manager role. That was the first time I got the feedback: “Go slow, because your team is not behind you. You need to explain things to them. You need to ask questions rather than give answers.”

I have a sticky note that is probably a decade old, and I can’t take that sticky note off my computer because it says, “Slow down and ask questions.”

My mode of operation is always speed and speed of execution. I’ve found that when you spend the time providing context to your team, ask more questions so that they can get to the answers, and enroll them in what you want to accomplish as a whole organization, then you go far.

Sometimes you opt for speed, and you turn around and you’re the only one there. What I’ve learned is that there is a ton of context-building when you’re making a pivot or changing the direction of something. You have to provide the context and the “why” behind what you’re asking people to do. That is a process of slowing down in order to go far.

4. Why is B2B Not Winner Take All

I’m intrigued when you think about making those big decisions and providing context for them. One thing that everyone actually said we should talk about was the transition—not the transition, but the moving upmarket—of your customer ICP and embracing a larger enterprise client.

Why did you decide not to move away, but to add, midmarket and enterprise to the purely SMB HubSpot?

That’s a great question. When I joined HubSpot in January 2020, I looked at how we were running the company and how we were running marketing, sales, and service. We were running HubSpot on HubSpot. At that time, we were a 4,500-person company, so we were well out of our target segment and our ICP, and it was working beautifully.

I could get all the reports and insights. I didn’t have to ask anybody; it was right there. We were customer number zero for HubSpot, so the product was absolutely ready. The question was whether we could get the partner ecosystem and the whole company behind moving upmarket.

We’re talking about 2 to 2,000. We’re not talking about a 20,000-person company or a 200,000-person company. We’re still talking about the segment that we were originally focused on, which is 2 to 2,000. But the opportunity to go to a 1,000-person company and help them grow across marketing, sales, and service was huge.

Our product was already there. The next step was to state that as part of the strategy and get the whole company behind serving an upmarket customer in that space. It’s worked out.

Is it not an entirely different business in the way that you’ve suddenly got to scale an SDR, BDR, and AE team, a customer-success team? These are really complex go-to-market motions and functions.

Is it not an entirely different company that you’ve got to build?

No, I don’t think it is that different, especially if you think about how we went about doing it. We were already serving companies with 200 employees, then we started serving companies with 500 employees, and then we started serving companies with thousands of employees.

There is a myth that there’s a completely different motion and you’ve got to hire all these enterprise-class reps to come and sell. In fact, if you do that, it will break the system and break the culture.

A more incremental approach is to get the product to be a perfect fit for the next-higher segment. Go from 200 to 500, then look at the kind of sales and support organization that you need and uplevel them incrementally. The next year, make sure they can serve a 500-person company, then take it up to 1,000 and to 2,000.

That is a much better way of scaling upmarket. We’ve been able to bring our reps along by upskilling them, bring our partner ecosystem along by upskilling them, and continuously drive innovation in the product as we go upmarket. That’s where we’ve been able to set the flywheel in motion.

Where does it feel most different? When you go from 200 to 400, then 400 to 800, and 800 to 1,500, where is the point where you think, “This is different from the last phase”?

It’s when the complexity of the buying process on behalf of the customer goes from, “I’m making decisions and I’ve got the budget,” to, “I have a committee of people who now need to make the decisions.” That includes IT, procurement, security, and maybe the board.

That’s where we had to upskill our sales team. How do you sell to a committee versus an individual owner, a sole proprietor, or a VP of marketing? How do you make sure that you’re driving a process that makes sense and works for customers upmarket?

There is an organic way of upscaling that you can do to make sure that your team is enrolled in the process, the skills are getting better, and then you begin to set that flywheel in motion.

When we think about the movement up to a higher employee base, you nudge more and more into Salesforce’s territory. You’ve always—not you, but HubSpot has—been like, “No, we’re not really competitors.”

No, they’re not competitors.

You kind of are, right?

It’s a very large market, and B2B apps are not winner-take-all markets. It’s different from consumer apps. With consumer apps, you see one use case, an app comes in, and then 90% of the market is just one provider.

That hasn’t been the case with almost any B2B app. Whether you look at HR, finance, recruiting, or CRM, they all have multiple players. The way we look at it is that, for the 2-to-2,000 segment, we have a great product and a great go-to-market fit. We’re really good at that.

Salesforce has an amazing enterprise segment that they’re really good at. Is there overlap in the upper end of our upmarket segment, in a 2,000-person company? Sure. We see them, and they probably see us. There’s a whole space in enterprise, and there’s some overlap there.

Are we competitive in one portion of the segment? Yes. But we also focus on really delivering for that entire 2-to-2,000 space.

What do you think they do better than you, where you say, “We can learn from that and respect that”?

I always respect and learn from competitors. The best competitors make you better. Salesforce does a really good job of marketing, telling the story, and creating the narrative. We look at that and we’re very impressed, and we learn from it. Hopefully, they learn something from us.

What do you think they should learn from you?

Making the products easy to use. If you look at a 1,000-person company that’s using CRM, we want to be able to put a customer in the application, and the next week people should start using it, adopting it, and intuitively getting value out of it.

We focus on making our product super easy, super intuitive, and unified. That’s not what customer platforms and B2B applications have been. Part of what we’ve been doing is training the market and our customers that ease of use is a really important attribute.

Pat Grady told me that I had to ask you this one. When you have an incremental dollar, how do you think about where to invest it? Is that investing in the product to make it easier, making the product-led growth motion more fluid, or investing in a more sales-led motion, which obviously increases the ACV and touches a larger customer base? How do you think about where to spend that incremental dollar?

If we really had to make a choice, it would be product first. The more you can invest in the product for it to be easy, intuitive, unified, and crafted, the easier it is to get the sales and marketing engine going.

I do want to talk about SMBs in particular with regard to AI, because you said something fascinating before. You said AI is history’s greatest equalizer for SMBs. I was like, my word, that is the title of a podcast. What do you mean by that?

5. How Does HubSpot Compete Against Salesforce

50% of companies that get started fail within the first 5 years. Starting a company is hard—you know this—but scaling a company is absolutely brutal.

If you think about why scaling is so hard, especially for small and medium businesses, they operate under constraints. In particular, 3 constraints: they are headcount-constrained, capital-constrained, and expertise-constrained. They can’t add the next headcount, they don’t have big budgets, and they don’t have an army of consultants and resources to bank upon.

They operate under constraints but have grand ambitions. When I said AI is an equalizer for SMBs, it removes each one of those constraints.

Think about headcount. If you’re using AI for support, you may not need to hire the next support person for a bit longer. It unblocks the headcount constraint.

Think about marketing. If you have a huge budget, you can personalize every single message that you send out. If you don’t have a huge budget, you’re not going to be able to do that. But now AI can help you look at your ICPs and look at segments—not as broad segments, but as individual customers that you want to prospect—and get a much better conversion rate. There’s no need for huge budgets if you’re leveraging AI to send personal messages.

Then think about sales and the expertise that you need to sell. You need a presales consultant, a demo, a solution consultant—you want all these experts to be able to scale selling. Well, you can get demos done through AI. AI can help in a lot of those areas.

When generative AI came on the scene, we got super excited—not because it was a shiny, cool technology that we wanted to hype, but because it unblocks SMBs in the core areas that we focus on: marketing, sales, and service. That is exciting.

Taking a pretty sophisticated technology and making it super simple—you mentioned sales tools that can do all these things. If you speak to most of the customers, they’ve churned and said they’re not very good. This is true for point solutions that are purely tailored to a specific part of the workflow, whether it’s demoing, interviewing candidates, or whatever that is. What is hype and what is reality?

There’s a lot of hype. This is why the north star for us has always been customer value. We talked earlier about companies slapping on an AI SKU and pricing it, and then you begin to see churn.

One of the things that we’ve focused on is building a great product and making sure that there is repeat usage and repeat value with customers. When we see that, we can scale it, price it, and do all kinds of things with it.

To your point, it’s early, but we are seeing clear signals in terms of agents.

What signals do you think we’re seeing that are most exciting?

Support is a very clear signal. We have an agent—a Customer Agent—and we have 1,500 customers using it. Within weeks, we saw 42% to 45% of tickets being deflected by AI. That’s within a handful of weeks.

Internally, we’ve been using AI to resolve 35% of tickets in year 1, and now we’ll be on to 50%. That’s a very clear signal in terms of leveraging AI to scale support.

6. Where Does Value Accrue in a World of AI

I’d say the same thing in marketing. We’ve used AI in marketing to drive conversion by having more personalized emails. That has driven conversion rates and helped us scale marketing. It’s early, but at this point in the technology, we’re looking for signals.

In terms of slapping on AI, what’s the difference between slapping on AI and the way it should be done for a lot of scale-ups? Think about your Notions and your Canvases—they’re in that realm of adding AI to an existing product base. What’s slapping it on, and what’s doing it the right way?

Think about this from an end-user perspective. When a product is neat and novel, people are going to use it for 1 day, and then they’re probably going to check it out again once every 2 weeks—or maybe they’re not going to—and then they churn out.

It’s really the difference between neat and necessary. Necessary means you’re using it every single day; you cannot get your job done without it. The best way to think about leveraging AI and building it into your product is: can that feature go from neat to necessary? Can it become essential as part of the workflow of your end user, such that they cannot live without it?

That’s the thing that we have to do. You can’t get it right away. You can’t just launch an AI feature the next day and expect millions of users. There’s a lot of iterative work that needs to happen to figure out which features are going to go from neat to necessary.

That’s the way to think about building it into the product: a product pathway where you start off neat and, over time, move to necessary with incremental improvements in data, improvements in accuracy, and customer feedback.

It’s very difficult to start as just necessary, isn’t it? You have a good path.

What is unique about AI is that every day you’re getting customer feedback and it’s getting reinforced by whether a product is being used. I’ll give you an example with content creation.

When we launched Content Agent, one of the first things people attempted to do was create blogs, then move those into different social posts and social media. We can get the feedback: “This was good; I’m going to use it,” or, “This was not good; it didn’t work.”

That gives us the opportunity to iterate. As we iterate on some of these agents, we begin to see people start using them every week. Then there are some features where we begin to see repeat usage and a pattern of usage, and that shows that we’re on the right path.

This is the time when you’ve got to dream big and iterate small. You’ve got to start with a couple of use cases and continue to build patterns in terms of customer usage.

Do you worry about margin degradation with the integration of AI into your tools? You sit on top of models and have to pass a lot of the revenue through to them. It’s expensive to use them. That is more expensive than just running traditional HubSpot.

First of all, model costs and the cost of inference have come down pretty significantly. You look at where we were in 2023 compared with where we are now, and it’s come down significantly. I think the cost curve is going to continue to decline.

I wish we had a margin problem. That would mean the usage is really high. I think we’re still in the early stages of getting the features out and getting people to use them on a daily basis, so I don’t think there is margin degradation. I continue to believe that the cost of inference is going to continue to come down, so we won’t see that. There will be a lot more value that we can generate for our customers.

Who is in the most advantageous position? Is it startups with no legacy code, which are much more nimble? Is it scale-ups—say, 500 to 2,000 people? Or is it your HubSpots of the world, the incumbents?

Companies with speed will win. It doesn’t matter whether you’re a startup, a scale-up, or a grown-up like HubSpot. Anybody who operates with speed and can innovate will win.

At this point, HubSpot is operating like a startup. We have the speed, agility, focus, and urgency of a startup.

That’s what you’ve always done, though, no?

No. I think this is a pretty big shift in our culture. As soon as we pivoted the organization toward building an AI-first platform and building all the features for AI, we also said that we needed to become agile and move with urgency.

What specifically do you do to move with urgency and agility where you didn’t before? Is it devolving decision-making, removing one-on-ones, or whatever it is?

It’s speed and cross-functional decision-making that need to get faster. We talked a lot about alignment in a company like HubSpot. Getting work done is getting cross-functional work done, and getting cross-functional work done means asking: who’s making the decision? Is there clear accountability for the decision? What happens if the project goes south or hits a pothole? How are we going to solve for it?

That is the speed that we’re trying to get people to rethink. I also think it’s about the fluency of what AI can do for us internally as a company, as well as for our customers.

Part of driving that urgency is asking: are you AI-fluent? Are you getting better at writing prompts? Are you getting better at use cases within your function? The more conviction you have in using AI every single day, the more conviction you’ll have in educating customers and getting customers on AI.

7. How Does Yamini Use AI Everyday

I got told I had to ask you this one: how do you personally use AI every day?

My workflow has changed quite a bit. I have Claude with a few projects set up.

You have Claude? Why do you have Claude? I have all of them. I have Gemini Gems. I’m just intrigued why Claude.

I’m going to lose friends because my CMO was just on the show. I put in a prompt for Yam, and I do that for every single show. I do it across Grok, Perplexity, OpenAI Deep Research, and Claude. Claude comes out with amazing ones, and it says, “Yamini said this, and Yam said this, and Askar said this.”

I say, “Amazing. Where did Yamini say that?” And it comes back, “Oh, we didn’t know you actually needed real statements that she said.” I’m like, “Whoa, whoa, whoa. I could have said to her, ‘You said this,’ and she’d say, ‘No, I didn’t.’” These are very, very bad.

You’re talking about sourcing and citation, which is really important in certain use cases. I use it to write my LinkedIn posts. I find it better than OpenAI in certain cases.

I have a project for earnings-script writing, and I’ve literally fed it the last 16 quarters of earnings scripts. I put all of the documents from the last 16 earnings scripts into the project. When I’m about to write the next one, I give it the parameters of the story that I want to tell, ask it to use the approach that I’ve taken in the past 16, and help me write and think through it.

It thinks through it, and then I’ll ask, “If I need to make sure that I communicate this message, can you improve it?” Reasoning, thinking, sorting, analyzing, and then writing—it’s super useful. It has really shortened the time it takes me to write each earnings script.

I now use a combination of Claude and a couple of other models. I do the same thing: I’ll try a couple of different outputs with the same prompts and see which one I like better. That has helped me refine my thinking.

What is your process for writing an earnings script? Do you sit down and do it in one session, or do you do it over multiple sessions?

It’s one session. It’s a long Sunday afternoon, probably 3 to 4 hours. It’s much more about what clear narrative you want to provide.

You put the prompts in and it produces something. Now I have a thought partner, a reasoning partner, and a creative partner for doing it. I used to do it all by myself, and I’d think, “Oh, get me through this Sunday afternoon.” Now I’m actually enjoying it, because I’m asking, “Did I get this message across in this paragraph?” It analyzes it and says, “No, you could probably say it this other way.”

It really helps with being a partner in this process. I actually enjoy it. As much as I say it’s a lot of work, I really enjoy it because it’s a time for me to reflect on the past quarter or the year, think through where we made progress and what has worked, and clearly articulate that story to a really wide audience.

8. What Does HubSpot Do When It’s Core SEO Channel Dies

This may be way too in the weeds.

Sure, go for it.

When we think about SEO today, I’m sure a lot of HubSpot’s traffic is SEO.

Yes, it is.

You have an unbelievable content engine. SEO is slowly transitioning from Google and search engines to ChatGPT and other tools. How do you think about that? Is that a concern?

It’s something that has been happening for a few years. If you look at SEO traffic, especially after AI Overviews and what Perplexity and some of these sources provide, the big shift that has happened is that search engines used to provide blue links that you could click and then reach someone’s website. Now they’re providing answers, which means you no longer need to click.

And then you get sponsored links beneath that, and suddenly you’re on page 2 or page 3.

Exactly. I think there’s a whole process that’s going to happen. This is not surprising to us. We’ve been looking at diversifying our content strategy for the last 3 to 4 years, even before AI came in.

That means expanding into podcasts. We actually bought a podcast network, and we get millions of users listening to our podcasts. We’ve also bought a couple of email newsletters—old-style email newsletters—that get daily coverage and provide us another way to expand our distribution.

What’s the ROI of those media acquisitions? I know they’re fantastic properties, and well done to them, but how do you think about the ROI in terms of conversion and attribution versus brand and marketing?

It’s difficult to measure. It’s distribution. It’s top of funnel. I don’t think there’s one user who’s just going to listen to a podcast and immediately buy a product.

It’s about awareness-building. How is our overall awareness increasing across these multiple channels? What is the level of awareness of our brand, and how many touches does it take for someone to look at our brand, continue the process of becoming more curious about it, and then get converted into sales?

That’s a multitouch attribution model. I don’t think we look at it on a single-property basis. We look at how our overall awareness is increasing across these multiple channels.

That totally makes sense.

When we think about the shift away from SEO, the way we think about it is: where are our customers? Customers used to just visit our website, but today they’re on YouTube, Instagram, TikTok, LinkedIn, and other channels.

9. Quick-Fire Round: Satya Nadella, Parenting Advice, Biggest Concern

You need to be where your customers are, not just expect them to come to where you are, which is your digital presence—the thing SEO helped us build. That means it’s a much more customer-centric marketing strategy, which I love.

One of my favorite segments is the quick-fire segment. I’m going to say a short statement, and you’re going to give me your immediate thoughts. Does that sound okay?

That sounds good.

Let’s go. What do you believe that most people around you disbelieve?

Pricing strategy should not be about chasing revenue; it should be about attracting revenue.

Most people think pricing is a way to maximize revenue. That’s why you get all these questions about AI: how are you going to monetize it, and what’s the next big thing you’re going to do?

We always look at how we provide value for customers first. We’ve consistently lowered pricing to increase the value that we provide, and the compelling value that we provide attracts more customers.

A lot of people think pricing is about maximizing every dollar and therefore achieving revenue maximization. We think about it as market-share maximization, which means you approach pricing very differently.

What public stock could you buy and hold for the next 10 years—not HubSpot?

Microsoft. I’ll tell you, Satya Nadella is one of my all-time favorites. He’s done such a phenomenal job.

What do you learn from Satya? How he bet all in first on cloud and now on AI? How he used culture to pivot the company?

How he has instilled this growth mindset and curiosity within a company that seemed super siloed before he started. I read his book, Hit Refresh, and I was obsessed with it. I thought, “This is amazing.” It’s a whole lesson in company-building.

He’s done a phenomenal job. He talks to customers, partners, investors, and new companies. I think that’s incredible.

What board member do you not have that you’d most like to have?

likely Dario Amodei, the CEO of Anthropic. I think he is phenomenal: deep research, an exceptionally well-thought-out process for safe AI, and a researcher of incredible caliber.

I’m going to get in trouble, definitely.

What have you changed your mind on in the last 12 months?

I have a personal story. My son is applying to universities, and I wanted him to stay in California, where we live. I told him, “You’ve got to stay here. There are plenty of great schools.”

He was really focused on the East Coast. We went there, looked at the colleges, and he fell in love with the East Coast. I changed my mind. I saw him there, feeling really at ease in a completely different part of the country, and I was very supportive. He’s going to university on the East Coast.

Does being a parent make you a better leader?

Definitely. A more patient leader.

I have 2 teenagers, and a lot of times what I go through with my teens is exactly what I’m going through at HubSpot. We also happen to be in our teenage years.

There’s a brilliant book, How to Talk So Kids Will Listen & Listen So Kids Will Talk, which talks about how to communicate with children and teens effectively. It’s very much the same.

I’ll have to buy that. I’m going to go there. Why not?

Unlimited vacation, mental-health days, and no-meeting Fridays: were these just signs of woke times, or are they part of the culture?

We actually just made a change from unlimited vacation time to flexible time off, and that’s something that we’ve moved the culture around.

A lot of what you talked about happened in 2020 and 2021. Those were times when, as a humanity, we were going through a pandemic for the first time. People were losing their lives and livelihoods, and parents were trying to figure out how to work from home. I think it was a time when support was necessary.

We’ve now really refocused the company on customer outcomes and mission orientation. We just made this change to flexible time off. Don’t ask me how it goes—I’m actually not going to leave it there.

What concerns you most in the world today?

I read a statistic recently that gender parity, at the rate at which we’re improving, is going to take 134 more years. It’s going to take 134 more years to get to gender parity across broad health, longevity, the kinds of work that we do, pay, and all of that. That’s a really long time.

Do you think we’re seeing a reversion from the work that was done over the last few years? It’s one of those sad truths, but when you look at good times, you see things like gender parity being more of a focus. When times are hard, inequality worsens and conflicts happen. Respectfully, it is not a priority, and it is instantly deprioritized.

I know, I know, and that’s scary. That’s very scary. It’s similar to climate.

What’s the biggest surprise, or what is something that people don’t understand about being a public-company CEO?

It’s a roller-coaster ride. Every year there’s some new challenge, and you don’t know what the challenge is going to be. It’s a roller-coaster ride, but it’s a beautiful one.

Do you find it hard to regulate yourself when it’s very volatile?

You have to have good practices for yourself to ground yourself. For me, it’s been meditation for a very long time, meditation and yoga. Having my family around grounds me a lot.

There are things that ground me, and that’s important to me. It centers me. There’s a lot of volatility and uncertainty in the role. Every year feels very different from the previous one, but you’ve got to ground yourself.

One thing I find hard is that when companies reach a certain size, no matter how good everything is, there’s always one thing that’s bad.

Of course.

In every company of a certain size, there’s something that’s wrong with HubSpot today.

Of course.

You can always choose to be down. Do you know what I mean?

Absolutely. It’s hard, absolutely. If you’re a perfectionist, you focus on that one thing that’s wrong. You know 98% is good, but that sales motion is not working. I’m struggling with that.

What question are you never asked that you wish people would ask?

The number of times I get asked the question, “Are you going upmarket?” That’s a shit question.

Who would ask that?

The number of times I get asked that question is unbelievable. There’s a bit of a myth that the only way to grow, the only way to serve millions of customers, and the only way to become a very large company is by going upmarket.

I wish someone would ask why we have deep conviction in the midmarket. We would say, “Look, there is an enterprise market that’s very large, and there’s a market for SMBs. We want to define a category of midmarket software.”

Five or 10 years from now, when you have guests on your podcast, I want you to ask them, “Are you like Salesforce, are you like HubSpot serving the midmarket, or are you like Intuit serving SMBs?” That’s a question I’d love to dig into.

That is a terrible question about moving upmarket. I can’t believe people ask that. You’re right.

I’ve so enjoyed doing this. I heard so many good things from everyone, and this has been such a pleasure to make happen. Thank you so much for having me.

It’s such a fantastic experience to do it live with you here. It is so much better. This has been amazing.

Thank you.

Thanks a lot. Take care.