Growth in the Age of AI - Yamini Rangan & Brendan Tolleson | Southbound 26

Hello, Atlanta. How are you? That's some southern hospitality.

Yeah, that is amazing. Welcome to Atlanta. Yeah, you know, the first day that I landed in the US, I landed in Atlanta. This was in 1994. I was actually... I just got into Clemson. Go Tigers. Yes. Is that Georgia bull? Sorry, sorry, but neighboring state, and I landed here in Atlanta. So Atlanta always has like this real meaning for me. It's very nice to be back.

Oh my god, yes. We're going to get into it very soon.

Yeah, three and a half years. Does it feel like... it feels like every day we're running to catch up on what AI can do, should do, how should we as leaders, you know, operate in this, and it's just been this one massive race that we... the speed only increases. It does not decline. So, you know, I have a few slides and I just used it as a prop to kind of like, you know, share some of the thoughts. And so I'll say first off, three and a half years, it is a platform shift. And platform shifts basically have a pattern, right? Every 15 to 20 years you've had a platform shift.

Yeah. Now generative AI. And every one of these platform shifts, they carry a pattern to it, which is a new technology emerges. Right now, it is generative AI, the ability to kind of... for AI to generate insights and now even these things. And then the next thing that happens is that people start building towards it. Sometimes overbuilding. I think we're kind of in that spot of overbuilding right now. And then, you know, you begin to see what the value is of the new technology and winners and losers of that new technology kind of emerges. This is literally the pattern. And you can apply it to internet. You know, when internet came, it was more about how do you get information, you know, across any place that you're in? How can you search that information? That was what happened. People had to build capacity, fiber, you know, routing, switching, all of the local area, wide area, all the networks that were built in the early 2000s for this new technology. There was a new leader in that technology. Anybody who controlled search became super powerful. They were kind of the gatekeepers of that wave. And then a lot more new models arrived, like DoorDash couldn't have happened without both internet as well as mobile, Airbnb, and, you know, Uber. All of these things actually emerged. And so there were clear value generation from that. I think what is distinct about this particular platform shift is that it is happening faster than we were all used to. Right? What used to take five or eight years is taking like two or three years. And it feels like every day you wake up and you're like scanning for the 10 new advancements this technology has created, and so you're trying to catch up. The other thing that is also different this time around is that everybody's got a megaphone, right? You know, again, you get up in the morning and there's like, you know, 20 different, you know, threads on LinkedIn or X or YouTube podcasts that you have to catch up with, and everyone has got a megaphone, which means that it's particularly noisy. And for leaders like us that are navigating a whole platform shift, this has been much harder because of the pace at which things are moving and how noisy, you know, it is to navigate. And so it's even more important for us to be very clear about what our goals are in our business, how we can use this new technology in a productive way, and how to lead our organizations through a pretty massive platform shift. That's kind of how I think about it.

It is so challenging. It is so challenging. We have to give ourselves a little grace because we're navigating really, really critical times and at a pace that we're not used to.

Yeah. And so I think if you look at any news channel, any podcast you listen to, you're probably hearing this narrative that the SaaS apocalypse, if you will, that...

Oh, absolutely. My perspective from the very beginning has been that this is complementary. And again, like the pattern that happens is that when a new technology comes in, people are not clear about who's going to win and who's not going to win. And, you know, you take Anthropic. Dario was at our Inbound conference last fall. You were there. And, you know, it is just pretty incredible to see how they've gone from 100 million in 2020 to then a billion, and then 10 billion, and then they're on pace... they've already crossed like 30 billion. So it's just incredible to see. But it's also very clear that they are a platform. And they're a platform that's kind of like enabling a lot of other technologies to be built on top of the platform. They are the new gatekeepers. When I say they, I'm saying like LLMs are the new gatekeepers of this platform, just like, you know, search companies were, and they provided a whole new place for us to like interact. I do think that there is a whole game, which is the AI operating system, and they are now the gatekeepers of it. And so, you know, you talk about SaaS apocalypse. Every day, you know, I get the same question: is... what, you know, are you going to like get wiped out by vibe coding? And, you know, of course, coding has become easier, but it's not just about coding. You're all leaders, you're all technology leaders, and you know this. It's not like just writing code; who's going to maintain it? How are you going to integrate it? How do you make sure that your teams are actually adopting that technology? Just because coding got easier, the rest of change management and driving something and maintaining it and scaling it for thousands of users didn't get easier. So I again, like, I would break it down into maybe, you know, use this, which is the question that we're going through as an industry is, you know, who is going to drive incremental value? Is it LLMs? You know, is it the new startup that has just cropped up and they're moving at a pace that we've never seen before? Or, you know, for the first time, HubSpot is being an incumbent. Like for the last entire cycle, we were like the disruptors. And, you know, we're still as a company getting used to being called the incumbent, but we are in this whole new pattern. And what is the value that is, you know, getting delivered? And who's got the real moat? Is it the data? Is it, you know, the context? Is it just the foundational models' capabilities that are, you know, kind of continuing to improve? And so there are lots of questions. To me, it goes back to what you just said, which is I do think LLMs and, you know, application software are complementary because LLMs can provide insights, but the context of what you need to provide insight actually lives within an application. And, you know, LLMs can write an email and of course send an email out, but if you provide the growth context of the last thousand conversations, which, you know, email has actually driven more of an open rate? How, you know, does that email reflect your company, your value proposition, your brand voice, and how you compete against your own competitors? If it can reflect that, then that email with more context is actually going to drive better outcomes. And so you do need this new technology that can deliver insights, but you also need the context of what your application can provide across your business that then drives outcomes. And so we are at the stage where a lot is getting figured out. This is again the pattern, you know, of every new technology cycle. It's very, very volatile. It's very choppy because no one knows where the value is. But from HubSpot's perspective, we are clear. We're not doing this just for the technology's sake. We're doing this to drive outcomes for customers and prospects like you, and the way to do it is to really understand your business context so that we can help you grow.

Look, it is a huge, you know, opportunity in the sense that if you fundamentally break down the last couple of decades of software, software we provided helped humans within all our teams get more work done. And that was a common place that you would go to, your teams would click and point and be able to then direct the work that they have to do in order to drive growth. What is fundamentally shifting with AI is that AI is now our teammate. An agent can actually work as much as someone in a team can work, and that means all of that context needs to be there for that particular agent, and if not, that agent is not going to be successful. You wouldn't like hire an employee and basically say, 'Go. Go, you know, grow my brand' without talking about, you know, onboarding that employee and actually having the employee understand what your product is, what your brand stands for, and how you want to compete, and how you want to communicate. You wouldn't do that, right? When you hire someone, you're not just going to be like, 'Yeah, here's a quota and go hit your number.' You're going to train that. You're going to provide them the context and you're going to help them understand what your company does. That's exactly what we need to do with AI before we drive AI use cases and before we have agents actually start autonomous work or even human-in-the-loop work. For both of those, you got to provide a lot of the context, and that's the stage that we're at.

Yeah, Brendan, I'll tell you it's just been wild. You know, in November 2022, ChatGPT, you know, was the big moment, and I think all of us were fascinated as consumers of the technology. Interestingly, we came into 2023, we just finished our annual plan. And we just like figured out what the roadmap needed to look like and what the plans were for 2023. And in January, we basically said we would just have to completely pivot. And I think like, you know, the question for us was, is this an incremental technology or is this a leapfrog of capability that technology can do? And the problems are exactly the same, but I think for us that moment was like, we can now do things that we were never able to do before. I mean, you and I have been in sales, you know, and when we started our careers, we were given X number of accounts. We could only cover maybe 10%, 20%, 30% of those accounts. Like that, you know, we've not been able to cover all of the accounts that we've been given, and we're not able to follow up with our customers the way we want to. But here's a new technology that is now capable of doing that. So I think for us it was very clear that this was not incremental, it was a big platform shift. And so we pivoted. The first set of, you know, AI capabilities that we launched were in March of 2023. And I'll tell you we got a lot wrong, right? And the first set of things that we did was all conversational because ChatGPT was conversational. We're like, 'Okay, everything is going to be conversational immediately.' And then our first set of features that we launched were all conversational, but then, you know, once you roll it out to customers, they're like, 'Well, I don't know what to ask. I'm not exactly sure what to ask. And if it gives me a response, I'm not exactly sure if that is going to help me grow.' And so I think since then we've really refined what it takes to leverage AI within companies so it really up-levels the company-level productivity. And, you know, I kind of walk through this particular slide which, you know, basically says if we were a customer relationship management platform, we had a basic job to do before, which was manage your company contact, customer information, deals, tickets. Like we were, you know, very much a structured database. The change that AI actually led us to is to capture all of the unstructured information that happens. 80% of the information within go-to-market teams is conversations that we have with customers. Conversations as a prospect, conversations as a customer, support conversations, pre-sales conversations. So 80% of the information was all of this unstructured conversation. And before, we just... even if we could capture it, we couldn't do much with it. But AI gave a completely different way to now capture all of this information. And that's exactly where we started. And the core of what we've built is really to drive much more context. You know, our real kind of like point of view is that there's a lot of AI, I would call it slop, or just AI output, you know, for the sake of output. And what we care about is driving outcomes. And in order to drive growth outcomes, you have to know every customer in a much deeper manner. What their business is. And, you know, we should be able to go to your website and pull information about you so that we can tell you what kinds of prompts you should be writing. We should be able to look at your conversations, the 80% of conversations that's happening outside of your application and what you're tracking, and be able to say, 'Here's what works for your customers.' And that's the job that we now have to do. And so that's really what we have expanded to. And then obviously, when you try to, you know, manage all of the agents, what is changing at the interaction layer is that, of course, we will continue to go into applications, use these applications, but we now have agents, and those agents can do work on your team's behalf. It can write content for your marketer. It can actually set up a meeting for your sales team. It can resolve a support ticket. And in order to do that, we have to make sure that we're delivering both HubSpot applications where your teams can use it, as well as agents that can get work done for your teams. And that's the journey that we've been on. And, you know, this week was Spring Spotlight, which is a pretty big, you know, set of launches for us. We do product launches twice a year, one before Inbound, that's coming up later this year, and one this week. And for us, it's driving and improving the quality of what we deliver to customers.

Yeah, thank you. And we're just getting started. I think for us, it is about driving outcomes. And it's not just about technology for the sake of technology. It's driving outcomes, driving outcomes for growth. And so, you know, we're also changing our pricing. You know, this week we announced that we have a customer agent, and that customer agent resolves support tickets, and we'll only charge credits if it resolves a support ticket, not if it's having a conversation for you, or not if it's just handling some, you know, call volume. It's actually, you know, we want to be able to tie ourselves to the outcomes that we deliver for our customers.

Yeah, I think that's a great question. I was having some really great conversations at breakfast with many of you here. I think like if you're a go-to-market leader, you have to think about AI as, you know, new ways of solving old problems, right? That's the way that I would be thinking about it. Like now that there is a new technology, I mean, your goals are still the same. You wake up in the morning thinking about growth. You know, how many of you are thinking about a technology? You're not. You're waking up thinking about how do I grow my EdTech business? How do I grow my FinTech business? How do I drive, you know, my insurance company to, you know, adopt new things so that we can get much more productivity? That's what you're thinking about. And if you are in the growth business, start with things that were not possible before. And I'll give you like maybe two examples. The first, we were talking about this example of, you know, all of us in sales, we've been in the place where if you have 100 accounts, you would have to like manually research those accounts. And I remember my first job in sales, you know, I would read 10-Ks. I would read 10-Qs. I would read like any executive within that company talking about the company initiatives. Like create notes and track those notes in a Word document and then set up a call. Yeah. You can now do that, right? The place to start is how do you leverage this technology to say if you have, you know, a thousand new customers that you want to reach out to, you can get information about that company immediately. You can enrich the data. You can understand who are the right people within that company that your salespeople have to talk to. And you can get intent signals. And what I mean by intent signals is, you know, is there a new initiative that they announced? Is there someone visiting your pricing page? Is someone asking in a community like Reddit what is happening in a particular space? So you can get all of these intent signals and enrich it so that you are reaching out to that prospect at the time where their intent is really high. That's the place that I would start because it's kind of solving old problems with a new technology. Then the second thing that I would do as a growth leader, and we all know this, is like, you know, you have five calls in a day and you have the best intentions of capturing the notes and taking all of those notes and sending it to maybe your customer success manager or someone else within the team and sending a response in an email, but you never get around to it. Because, you know, your day is already maxed out. That has been the sales problem. You can now solve it, you know, pretty easily. You can have like a note taker join the conversation. It's captured all of the conversation that is happening. At the end of the call, as soon as you're done with the call, it says, 'Here are the four highlights. Here are the action items. And here is an email that you can send to the customer or the prospect as a follow-up.' Do you want to change it? Of course, you have the ability to change it. And then when you send it, like you've now made it, you know, much, much easier for salespeople to do their jobs. And so I would start with two or three of these new capabilities that were just not possible before. But then I wouldn't stop there. I think what is also what we're finding with, you know, AI in general is that people are getting individual AI productivity. How many of you here are using LLMs every single day? Yeah, all of us. My emails have gotten, you know, faster. What I write has gotten faster. But how many of you...