The Founder Selling to Professional Sales Transition w/ Dmitry Shevelenko (CBO, Perplexity)
35:46 Share

I think where Aravind was not able to make it happen on his own is he did not tie the story together for them about how doing this is going to either improve that other partner's customer acquisition or their retention. The year is 2025. You walk up to 100 people on the street and ask, what is Google? They all know search engines where I go to get my questions answered.

The year's 1985. You walk up to 100 people and you ask, what is Google? 99 say, I don't know. And one, who's a PhD in mathematics, says it's the number one with 100 zeros after it. It's crazy in tech. It's one of the things that attracts me to tech is like how we change the world. We shape the world with products, workflows, innovation, societal changes, and new words like Google. Well, today we meet another potential new word called perplexity.

And they're going after Google Space. Today we're speaking with Dmitry Shevelenko, the chief business officer at Perplexity. He met their brilliant founder, Aravind, a couple years ago. Dmitry was advising a bunch of... He didn't even have enough conviction to join full-time. It was one of five adviserships that he took on.

But wow, what energy and pull he found from this vision. And he's going to unpack how they successfully transitioned from founder selling to professional sales. I'm Mark Roberge, and this is a Science of Scaling. Dimitri, welcome to the show.

Thanks for having me. Excited for this. Gosh, every quarter or whatever, when I check in, it's like perplexity is that much more famous. So what a ride. And unpack that first conversation, I guess, both with the VC as well as the founding team. Like, what intrigued you? I think the audacity of taking on Google, which is something that just on the surface, you know,

still sounds crazy, but sounded even crazier in 2023. It also piques your curiosity in a landscape where Google has everything locked down. You know, what would be the interesting angle? And Aravind, you know, he's a AI PhD. He's a researcher. He does not

have experiences in the business world. I kind of, in the first meeting, you know, shared some ideas on, on things the company could do. And then, uh, I think he reached back out and, um, and said, Hey, like I tried everything you, you said, and none of it worked. Uh,

And that's when we started getting the details of like, well, how did you do it? You know, who did you talk to? What exactly did you say? And in that conversation, it was kind of like, well, if you think you're so smart, why don't you come try to do it? And so so shortly after that, I kind of became an advisor to the company. Why did you want to start part time? Like, you know, yeah.

Were you just not quite sure at that point? Or you're just like, it just made sense for your career? Or you had an opportunity to try before you buy? Like, unpack that. So I was a, you know, coming off the emotional wounds of winding down a startup I had founded that failed, which was a robotic startup called Tortoise. And I really liked working with founders, kind of an advisor model. So I had, I think at the time, you know,

four or five other startups where I was working closely with founders, you know, just helping them think through growth partnerships and fundraising. Hey, folks, just Mark here. Yeah, I'm listening to Dimitri navigate this career transition timeframe. And there are three things that are popping out here that I just love. Like I've done them myself. And I find myself giving people this advice often when they come to me.

between gigs. All right. The first one is like, he just mentioned like, okay, he's robotic startup, unfortunately failed or didn't go as well. And he's, it's fine. Get over it. We are so blessed, especially if you're listening here in North America, United States, this isn't true in all cultures, but we in a way reward failure. Like it's okay. I, you know, I have a lot of ECs and I share this myself is I'm, if I'm looking at someone that

is entering the entrepreneur ecosystem and has never done it before and comparing them to someone who did a startup and it failed. I put the latter person much higher in terms of their abilities. Now, of course, we got to unpack why it failed, that kind of stuff. But like, it's this risky stuff.

So, you know, kudos for him picking his head up, just getting out there and, you know, you're going to fail. I've never met a successful entrepreneur that hasn't failed, hasn't been fired, hasn't gone bankrupt in one of their plays. So accept that reality. Number two, he's doing a try before you buy. Now, not everyone can afford it or figure it out, but if you can, I just don't like it if you're going to go through like five to 10 hours of an interview with a company and pick your next career move.

He's actually advising four to five startups at a time. He's understanding how it's like to work with the founder. He's understanding what the customer conversations are like. He's understanding how the pull with the product is. He's understanding how to sell the thing. He gets the culture. He understands all the skeletons before he dives in and picks one. And thank goodness one of them was perplexity. He gave himself that optionality.

So, you know, if you can, two, four, six months, try before you buy and be an advisor to a few. And the third point is just get into AI. It's the future. It's very hard for me to justify a career move right now if AI is not at the cornerstone of it in terms of the offering and the experience you're going to get, especially if you're in tech, especially if you're in, say, B2B software like many of you are. All right, let's get back to Dimitri.

I enjoyed kind of the intellectual variety of working with a lot of different companies. So I wasn't looking for it. I didn't know enough about AI and so much of kind of early stage startups is chemistry with the founders, right? And that obviously takes time to develop as well. Biggest question I get every day. How do you build the next unicorn?

How do you build predictable, scalable revenue growth? Luckily, the folks at HubSpot have put together the Science of Scaling Database, real playbooks from companies that have gotten to the scale of like $40 billion in market cap. The playbooks cover key decisions like...

hiring, compensation, go-to-market strategy, ICP development, even stuff on AI in sales. All from the folks behind the fastest growing companies in tech. It's not fluffy. It's super tactical. You know my content. You know my brand. It's stuff that you're going to pick up, download, and put to use within an hour. So head to the description, click on the link, and download your free copy to start scaling your business today.

Even before the advisory, it's like, it sounds like you all had like a one hour brainstorming call. What were some of those tips that you gave him that he liked and then tried and it didn't work? Like why didn't it work? And then how did you kind of fix it or do it correctly?

Everybody else like would tell him, you know, go work with a browser, like basically like do what Google did. One thing I think I suggested in the very first meeting that has become a big part of our growth strategy is focusing on mobile carriers and focusing on the fact that we have a subscription version of Perplexity, Perplexity Pro. Those are great.

you know, potential Lego blocks where you could create a new type of partnership. What do mobile carriers care about? They want customer acquisition, customer retention. And so any kind of exclusive differentiation is going to matter to them. The fact that I guess I didn't suggest trying to work with a browser helped me stand out. Is that that you were

Trying to, from the beginning, being app-based with a premium subscription. It was just more a realization that there's no, there's nothing you're going to do to get Apple to drop Google from Safari. There's no way you could land the plane on that deal. And so why even talk about it? Google comes out, whatever, 25 years ago, and they build a better approach to search engines. It seems to be driven off like,

inbound links and the quality of these inbound links to decipher whether a piece of content is authoritative. And they take an ad-based model, which means build an awesome product. They were notorious for like that very clean look, which drove a ton of folks. And then

I think we're a main driver in the CPC approach, just an inventor of that. And that led to one of the most valuable businesses in the world. And it seems like perplexity is a very different approach. Obviously, the advent of AI is challenging the notion of search. And in this case, the revenue model doesn't appear to be like...

ad based, but is going to be, you know, sort of a freemium that leads to a subscription. Those are two different strategies. Were those right for the time? I mean, Google definitely played the right set of cards at the time. It's just, you know, you didn't have...

the technology breakthroughs to make natural language questions and natural language answers possible. Even if you had tried to disrupt Google by just having, you know, better results that are links or better results with fewer ads, you know, Bing tried that and it didn't work, right? And also like the internet in, you know, early Google days, I don't think the subscription model, uh,

And to be clear, like we are also a perplexity experimenting with ads in perplexity. Like I don't like our position isn't like you will never have ads. AdWords style performance advertising is going to be less relevant in the future of the Internet than it has been in the past.

What we think will be evergreen is brand advertising, right? So we think questions are a very natural way of doing that. So we have some views on that. And that's actually why it was somewhat controversial, but why we started with CPM advertising, because the kind of focus on clicks, again, will, I think, be

be a smaller part of the internet of the future. And that has a huge impact on how I think B2B marketers, if we move in that direction for a second, have been kind of retrained and pioneer over the last 20, 25 years, you know, moving away from the

billboard impression events. You know, you hear from everybody that the organic clicks are plummeting because this is just such a better experience for the user. It's like, I have a question. Just tell me what the specific answer is and give me citations to the best content so I can do my validation of like whether I believe it or not. Don't make me crawl through like 20 different results in a SERP to figure out my answer. But again,

It has such a dramatic impact on like, what is the future of SEO? What are the future of websites? Yeah. I often get asked like, how do I show up more often in perplexity answers? Right. If you can just stuff keywords into a website and all of a sudden that changes the perplexity answer, then that perplexity answer isn't that valuable. So the...

unsatisfying but true answer is if you want to show up more often in perplexity answers, build a better product. You know, have people say the right things about it. Right. So it's the the signals we look at are things like reviews, you know, earned media. And it's the sources that, again, people trust that end up influencing the grounding for for the answer.

Yeah, I'm kind of working this one through my head. It's always been about creating great content. Don't try to trick the algorithm. And he's given us that flavor too. And I agree that's going to happen. He's definitely leaning toward like product reviews. And maybe this means like third party opinions and data will carry more weight than the content marketing SEO wave where we kind of controlled more of that narrative. And then just paying attention to where authority is perceived with those

sources, which I'm not quite sure. He's basically saying like the LLMs deciding a bit. So yeah, I mean, I was excited to kind of walk that through them. He gave us a couple nuggets, but maybe that's just where we have to work together. Maybe in the comments, if you have some questions, advice, et cetera, on SEO, search, even SEMs changing. It's going to change quite a bit. I don't know if his narrative there inspired a viewpoint from you.

All right, let's get back to him. But I think a lot of where this goes is it's like a focus on quality rather than kind of this gamification of the web. Which in the end, I think that's always been the mantra is like just build great content and good stuff will happen. How do you all think about who to trust?

You mentioned we're going to go look at reviews and obviously we'll weight those toward the more trusted sources. How does Perplexity think about that? There's a number of algorithmic systems and there's also just a lot of feedback loops. A really interesting one is we have our own search ranking system.

And then we get a signal of, you know, what did the LLM choose to incorporate in the answer? Right. Because you pass through a lot of snippets. It doesn't necessarily use all the sources. Right. So you kind of have this, you know, constantly improving intelligence. But ultimately, what keeps us honest is like.

Our singular kind of metrics that we look at is queries per day and retention. Right. So if we're doing our job, people will keep coming back to perplexity because they found value in the answer. And so that's kind of, you know, the true North Star for us.

Okay, we're back in that one hour powwow. You drops great knowledge on, you know, Aravind that that's, it's different than what other folks were telling them. It's not about the browser. It's a different approach. I don't know if the specifics was about this like mobile carrier partnership.

Can you give us like, what were the one or two things he like? Yeah, I like that a lot. Let me go try. He tried it. It didn't work. Yeah. Getting back to that. How did that not work for him? And what did you come in and do once you were an advisor? I mean, the crux of the strategy is like, I mean, you can think about mobile carriers as a subscription product. And so it's basically bundle yourself with other subscription products and be willing to, uh,

effectively subsidize a free trial period. You could write it out as a math equation where you're

arbitraging a lower CPA over your costs, your cogs on higher inference for a pro user versus a free user. Knowing though that having a critical part of just growth in consumer is earned media and kind of embedded in the insight was if you have a lot of these things start popping in a somewhat concentrated time window,

you just create a lot of chatter, right? This was only the right strategy for perplexity because what I did know going into it is the people that were using the product really liked it. You can't like, you know, nobody's going to bundle like a

a shitty product. This was a strategy that worked because the product demos. Well, we talked about different, whether it be like music subscription services, whether it be the mobile carriers, like there's other subscription products that you can kind of bundle into. Um, I think where Aravind was not able to make it happen on his own is he did not tie the story together for them about how doing this is going to either improve that, that other partners, uh,

customer acquisition or their retention? How do you get the person on the other side of the table promoted? That's your most important task in trying to do something zero to one and have them see that story emerge. It's important to unpack some of these building blocks of effective founder selling. It's not just like, look at how cool my product is.

but just the dimensions of like getting to the right person and telling the story in the right way. So now he says, he goes out and does it, he gets stuck in the bureaucracy. He comes back to you and says, "Dimitri, like I tried, it didn't work. Either you don't know what you're talking about or I don't know what I'm doing on this thing. Can you help?"

So you come back, as much as you can tell us, like, who'd you go talk to? Is there, you can anonymize the carrier. You can talk, if you can tell us what it's like, just so we can get into that meeting and how that story unfolded. It wasn't all carriers. I mean, carriers in some ways, you know, I knew that would be a much longer trajectory. So it was like other, you know, other kind of, you know, consumer or prosumer subscription products that I thought we could make quicker progress on.

Unlike a pure B2B product, like we're gonna, most companies that we try to work with, there's already somebody there that loves us, right? And it's about having that champion kind of, you know,

help us fight their own bureaucracy, uh, to, to, to get to the other side. Right. So I think it was, you know, quickly just even like scanning, if I asked him to like forward me all the email chains you've ever had. And I could kind of read in between the lines of like, there's nobody there that really wanted it to happen. And so like, let's not, let's not focus there. Uh,

And what were the ones where there's clearly a champion and we just hadn't empowered that person to make this exciting. In the science of scaling model here, we're at that point of transitioning from founder selling to professional selling. And geez, do we see the same patterns all the time? So if you haven't seen some of the other, the past episodes, let me just kind of catch you up. Okay. First off,

I mean, these technical founders, they're brilliant, but they've never sold before, right? So they're naturally gravitate toward a show up and throw up, feature selling, very technical selling. Look how awesome my product is. And that's how I did it too, coming out as an engineer. Not understanding like some of the just basic fundamentals of sales. And maybe you're a product founder listening too. So you hear Dimitri come in, right?

And so Dimitri is all about like, what's going to get the person promoted, right? So he's not coming in and being like, look at how fancy perplexity is and all of our, you know, AI and the neural network and like the, you know, the quality of the data and the algorithm, all that kind of stuff. He's just coming in saying,

asking about like, hey, I know your business models around retention and acquisition. Can you talk to me through some of the strategies that you've been using? He hears how they're talking and he's going to find a way to help that person with their perceived problem to make them look good and get them promoted. Right. So like he's completely flipping on his head.

He's also really aware of where power is, where decision power is, what a true champion is. I've never heard this before. He has Ervin send him all the email threads that Ervin had with these potential partners to detect whether or not we have a real champion in here. So that's just brilliant. I'm sure you could just feed that into perplexity and summarize it and help identify the strength there.

The other thing that he's walking through is the importance of an early adopter organization, right? There are organizations that are going to like require 10 references before they buy and socks compliance. And like, guess what? They're not an early adopter. In Dimitri's case, his sense of an early adopter is a huge fan person of perplexity.

And then you can start getting into like their history of adopting newer technologies to get ahead of the market. We've got to sniff that out. And then the final piece that he's teaching us on is in the early days, you can't start with Verizon, in this case, with a carrier. Even if they love it and you have the early adopter and a strong champion and all that stuff, it's still going to take you 18 months to get through and your champion might quit in that time.

You got to start small. Yes. Okay, fine. Replit. Now they're crushing. But at the time they probably weren't. And that's just a great place to start and test out these messages, test it out with a couple of users and slowly graduate up to the higher potential, high scale folks that might take a little more sophistication and a little more time in the sales cycle. The same patterns on this transition from founder selling to professional selling.

let's get back to dimitri sometimes you do a partnership not because the partnership itself is going to deliver value but the people who work at the other companies that you know really do matter they'll notice it and it'll create some fomo and it's an essential trait for so many founders is impatience uh like and kind of from a first principles like why can't this be done faster

And so I knew we needed, you know, kind of a barbell strategy of like, let's just let's do something. It's not going to be earth shaking, but it might play well on social media and like introduce this concept. And there's like kind of almost like a law here, right? Like the bigger your potential distribution partner, uh,

uh inherently more employees work there more bureaucratic right like you can always move quicker with another startup than with a bigger company but the startup itself will not have that many users or customers and so it will be a smaller distribution partner you've been talking from the perspective of business development you know i think there are a lot of like early founders and early sellers here that are confused on those terms you know i do think some people

call their sales team business development. Sometimes they call them account executives or sales. How do you frame those different roles in your mind so that a founder knows like what the difference between a true business development person is versus like a true account executive, maybe direct seller and which one they should use first?

I wasn't like boxed into like a function. It was like, Hey, you know, your job is to figure out how we grow. Cause I don't fundamentally see a meaningful distinction between social media, comms, business development, sales. It's all getting more users, getting more queries. You know, it's, it's, you're deploying different strategies, potentially tied, tied to some, some of those function names. But,

But the most effective campaigns and initiatives are inherently combined multiple elements. Like we were doing partnerships in some cases just for the tweet. And so you have to then you're optimizing for a different set of variables than, you know, you are with other partnerships. And so you have to think about these things holistically. If you're bringing on someone functional, you better optimize.

yourself have an idea of how their function fits into the larger puzzle. If you don't, then bring on someone who can actually, you know, cross many functions and help you think through those dynamics. Yeah, we haven't talked about this that much. And there's certainly a pattern here, again, with the transition from founder selling to professional selling. On one hand of the spectrum, you know, Ervin's doing a brilliant move here.

in like making this an abstract role, figure out how to grow. Like not saying like figure out our ad strategy or figuring out our partnership strategy or figure out our earned media strategy. That is like leading the witness. Just like figure out how to grow. And Ervin avoided it. But at the same time, he didn't just go and say like, go figure out who at Google did this.

Like I kind of feel like he had a sense of what the right strategy was and leaned in there. Like if he had sensed that this was going to be like a, an ad model, I think he would have gone with the Google person. But I, I think the founder had a sense and you kind of do, you need a directional sense because not everyone is like so intellectually broad that they can just like move around. Like,

Dimitri did have a background in certain areas. He had a background building partnerships at Facebook, could have been on the mobile side. He had a background doing business development at Uber, which is like a mobile first company. I think there was a directional hypothesis that Ervin had in selecting Dimitri and that led them to this optimal growth strategy.

I hope that makes sense. Let's get back to him. Probably the most difficult transition for any early stage startup is like going from founder led sales where it seems to be working and then you hire a first salesperson and like all of a sudden all these things break and like you don't know why. And it's probably because whoever you hired is like not thinking holistically and just thinking about it like, you know,

applying best practices they maybe had at a bigger company. So actually, I want to also spend time on how your sales team works. I think I have a very distant view of like how it works at say Google as an analog company. I'm sure it's going to be different there. There's like account teams and all that kind of stuff. And then also,

how you are potentially using AI in your sales process or using AI to like rethink how a sales team might work. I'm not sure if you're all going that direction at all. We only have four people on the team who are mandated as, you know, salespeople. You know, we call them enterprise growth leads. The way you have four sellers and 4,500 clients is inbound, right? And so for us, the unique channel we have is like I was saying,

at pretty much any company we'd want to work with, whether it be on the distribution side or having them be a customer for Plexi Enterprise Pro, there's already somebody there that's using

perplexity in a personal capacity right and so how do we turn that into our advocate um but but actually i mean a great uh example of how we use ai is you know for any outbound person like if you're going to be using ai and perplexity in particular for for anything it's this it's just before you have a meeting ask about the company the person right and like get and if you just think about

how much prep time you can save. That I think is like the major unlock. And you can just make that a lot of the discovery questions you would have asked in a meeting. You can actually, you know, maybe 80% of the question. Yeah, the answer is ahead of time. Wow. I have like goosebumps right now on this one. I didn't think about this in preparation for the first sales meeting. Replicate how you're going to run discovery.

That's so brilliant. Like my mind is racing right now. I got to keep playing with this. I haven't even slept on it yet. But just like where we're headed is like, why should this company buy my product? What's the upside to it? What's the downside to it? How should they think about whatever my category is? How should they think about solving whatever the problem is that I solve? I can't believe this is the first time it's come up.

Well, anyway, I hope that gives you like a ton of ideas about how to prepare for meetings, how to train your reps to prepare for meetings, how to do preparation with reps on discovery. I don't even know. Am I comfortable going as far as Dimitri says where I'm doing this on the call with a buyer? There's a lot we can experiment with here.

All right, let's get back to him. You still may want to ask those same questions in the meeting. Never assume. But you basically now have a much smarter way of framing them because you already kind of think you know the answer and you're kind of can then, you know, nudge the conversation, you know, towards fruitful grounds for where you want to take it. Perplexity has this feature called spaces where you can, you know,

basically either connect internal files, like say a Google Drive folder, or you can upload them. And then you can prompt against that. And you can either prompt to it by attaching a new file. And so it's like, hey, fill out this file I just attached. Give me the answers to all these questions in the same way that I answered it in the other security questionnaires. And for the ones where there was no...

No similar question. Look at our underlying. I also upload all of just our documentation around security and be like, just fill in the answer based off of what's in the documentation. And again, Perplexity is able to do that. You know, I'd say like save 90% of the manual time. In my experience, that's the sort of like

kind of grind work that sellers like will put off because it's like you know kind of cumbersome and like you know you're just like you're never excited about filling out a security questionnaire yeah the same way you are about like yeah like oh just got a juicy new like prospect like let's let's like learn about them uh and so i think making it really easy to get to that 90 draft it means you know people aren't procrastinating and you're not you know creating um slash

slow down. Using Perplexi to prepare for meetings. I genuinely do. I am pitching, I'm raising our next fund. I go in anytime I have an institution, I always ask about the institution. I always ask about the individual I'm pitching. I always ask about venture capital funds that they've invested in in the last 12 to 18 months. It's amazing what Perplexi can teach you. And I love your comment on just think about

the discovery questions you might ask in a first meeting with this individual and ask perplexity just to prepare. Are there any other creative or interesting prompts you've come across as you've done this for nine plus months that is now part of your regular routine in prompting perplexity to prepare? Often what I like to do is I will then at the start of a meeting, I'll share my screen and

And I will rerun some of the same queries I just ran before. And just like in this like transparent and my favorite one is be like, hey, you know, I'm meeting with, you know, so-and-so at so-and-so. I am, you know, the CBO at Perplexity. What would be good things for us to talk about? And Perplexity is quite good at like, I mean, basically, you know,

it captures usually 80% of what I would want to cover on the agenda anyways. And it just creates a really interesting kind of frame up for, so, so, you know, what I encourage people is like, use this stuff in real time, almost as kind of like a neutral arbiter of like, you know, you know, ideas are like, Hey, like, um, when I, when I speak with journalists, I, I always tell them, uh, you know, you should have used perplexity to come up with questions to ask me. Um,

And about 15% of the time they say, I actually did, which is awesome, but...

playing with the fact that it is probabilistic, not deterministic, and like, you know, like sharing in the delight of what it might say. Well, let's do HubSpot. I was just going there. This is intriguing. Yeah, so it's like, you know, what do customers, you know, who've been using HubSpot for a decade say about it, right? And why is HubSpot considered the leading, you know, so-and-so, so-and-so by so-and-so?

And then you get an answer that's like, again, not a commercial, not written by HubSpot, but it's perplexity, like grounding that, that is very authoritative, right? It just becomes like a testimonial in the room that, you know, you couldn't have faked that that's kind of working in your favor. Yeah, it's, well, I was going somewhere even more extreme, Demetri, which is probably not advisable, but it's kind of like, let's say HubSpot's pitching perplexity and it's just like,

Which CRM should Perplexity use and why?

Should perplexity be a HubSpot customer? Why, what are the reasons they should not? You know what I mean? That's pretty extreme, but you know. - Or like, you know, how can perplexity, you know, triple revenue by becoming a HubSpot customer? It will push back on the premise of like, you know, why is HubSpot the best company to have ever existed? But if you kind of frame it as like, you know, how could you do this with that? It will, you know, attempt to answer the question.

Well, Dimitri, I really appreciate you making time for the podcast here. You know, I just can't imagine where we're going to be next quarter and the quarter after. And thank you so much for being a big part of that ecosystem shift. Thanks so much, Mark. Hope to talk soon. All right. That does it for today, folks. Our episode was written and produced by my favorite producer, Matthew Brown. Editing comes from Patrick Edwards.

The science of scaling is a proud part of HubSpot Media. And if you like what you heard today, make sure you follow or subscribe to us wherever you're a fan of the show. And if you're a founder ready to scale, check out my VC firm, Stage 2 Capital. We are backed by over 800 CROs, CMOs, CCOs from the best companies in tech, and we're ready to help with your scaling journey. Okay, see you on the next episode.

Today's episode is brought to you by HubSpot for Startups. Right now with HubSpot for Startups, eligible startups can save 30% to 75% on the AI-powered and connected CRM that 248,000 plus customers, including some of the fastest growing startups, trust to increase leads, accelerate sales, and streamline customer service.

grow better, and scale smarter. Visit HubSpot.com slash HubSpot for startups, or click the link in the episode description.