David Cancel puts the “serial” in serial entrepreneur. David has founded a total of five software companies over the years, which he says make him “certifiable”. The list includes Performable, which was acquired by Hubspot, where David subsequently spent three years as Chief Product Officer.
In 2015, David left Hubspot to start Drift, a Boston-based conversational AI platform for marketing and sales. The company has grown very rapidly and now has a whopping 50,000 customers. Drift has raised a total of $107M from a number of venture firms including Sequoia, General Catalyst and CRV. The company has also been recognized as a Forbes Cloud 100 company.
David also has built a very strong presence and brand in the entrepreneurial community. He writes a popular newsletter, ‘The One Thing’ and hosts a long-running podcast, ‘Seeking Wisdom’. He’s very involved in a number of startups as advisor and angel investor. He’s also an Entrepreneur-in-Residence at Harvard Business School.
David and I had a really interesting, wide-encompassing conversation at our most recent Data Driven NYC event, where we covered a range of topics including:
- Building a global SaaS brand with 50,000 customers in an astonishingly short amount of time
- How Drift was founded to take advantage of a fundamental paradigm shift
- Creating a new type of CRM, driven by conversational data, with automation at the core
As always, Data Driven NYC is a team effort – many thanks to Jack Cohen for co-organizing, Diego Guttierez for the video work and to Karissa Domondon for the transcript!
FULL TRANSCRIPT (lightly edited for brevity and clarity)
[Matt Turck] I’d love to start this conversation with your personal journey, David. You are a five-time founder and a two-time CEO. So I would love to hear about what you’ve done prior to Drift and maybe starting as far back as college, because if I understand correctly you, like many other great entrepreneurs, actually dropped out of college before starting your career. Is that right?
[David Cancel] That’s correct. I will say dropped out is too active a term. I think I just faded away from college. It wasn’t that clear cut.
What happened next? What were the early years of your career leading into the first company that you started?
It’s a long winded story, so I will give you the short path. I will say that, no one used the term entrepreneur or no one that I knew used that term back then. I think entrepreneur to me back then meant that you were, it was before the culture of entrepreneurship. You did infomercials or you had a franchise or something like that. It was not something that I inspired to.
I was obsessed with the early commercial internet and I started to code and do things. That time when I was doing that, it was so long ago that it felt like we were a bunch of pirates figuring things out. There were no books, there was obviously no Google, there was no information and so it was just a really interesting time where we could just make things.
Those things led to companies. I joined a couple early start-ups a long, long time ago. Then I started my own company in 2000, post-bubble 2007, I had a knack for starting things in bad times. I’ve started five, including that first one. I always say that if you start one company it’s excusable, because you’re naive, two, it’s questionable, and three or more, which is my camp, certifiable. So, I’m certifiable.
Your background is as a software engineer? So you’re coming at entrepreneurship from a product and technical standpoint?
Yeah, I studied software engineering in school, but I was frankly bored by it in school because again, this was pre-commercial internet. So we were mainly building drivers and server-based software, desktop software. Now I can look back at it and say, the problem was that I didn’t have a feedback loop where we would create products, never keow if anyone actually used them. Maybe someone would use them years later, they were put on disks, floppy disks and then CD-ROMs and things like that.
So I had no connection to the customer, and it was really building things on the internet and having that kind of feedback loop cycle that I became kind of addicted to. That’s what led me to do this stuff.
I rose from being a software engineer – well not rose, in some ways that was a lot more fun – but software engineering to starting companies to being a CTO, to leading product teams and then did that at various companies. My own company – sold one, my fourth company to a company called HubSpot when we were like a 200 person or so company at that point, about 30 million in ARR.
That was Performable, right? What did the company do?
Yeah, that company was Performable. We were a marketing automation product. Basically the insight that we had was that we were going to a world of kind of omni-channel presence, which is kind of obvious now, but this is 2009. We were taking signals from everything and we were one of the first that I can remember in the marketing and sales world to actually have something called a marketing world, I would say, to actually have user based profiles, right? Because that was the only way you could unify across omni-channel. I remember everyone that I could remember in that world saying it was the dumbest thing ever because marketers never looked at one person, they only looked at segments and then obviously everything has a face in it, everything has a profile in it today.
What was the insight that led to Drift? Why now, when you started it?
Why now is probably the most important question.
If I think about my journey across five companies in 20 years, I would say that I started from a very engineering standpoint, which is, it’s all about the product, it’s all about building the thing, it’s all about scratching the edge, it’s all about eating your own dog food.
Then I went all the way to the other end when I started Drift and said, what are the undeniable trends that are happening in the world? Right? How are those changing user behavior because I figured out doing it this way that user behavior was almost impossible to change, right? So, what were the changes externally that were affecting user behavior and changing our behavior?
Then I started to look at those and realize that most of the stuff that we had built, even the stuff we had built at the most recent company which was Performable and HubSpot, didn’t really make sense anymore because we believed that we were undergoing this paradigm shift where we were moving from a world where the company controlled the buying process largely in the B2B space to a world where the buyer had all the control and that was driven because of mass commoditization in SaaS and a whole bunch of trends that we were looking at.
That was the why, now we were starting the company. And why now is the question that we ask ourselves basically every day and everything that we built at Drift.
I’d love to dive into the product part of this. I think a lot of people are familiar with Drift and for anybody that doesn’t know, it will sound very familiar. You get on a website that could be Snowflake, that could be InVision, that could be GrubHub, that could be GitHub, and then you’re going to have that little window that’s going to appear, it’s going to pop up at the bottom right corner is going to say, “Can I help you?” Or, “Check out our next webinar.” All that is often powered by Drift.
But there’s actually a number of products which I personally hadn’t realized before diving into this. There’s the live chat product, there’s a chat bot product, there’s an email product, there’s automation. Maybe let’s take all of those and actually spend a good amount of time explaining what they do. Because it’s not just a chat, right?
Yeah, sure. Happy to talk through all that and talk through the thought process on it.
Basically what we’ve been trying to do from the beginning, which is invisible to everyone, is that we are trying to build a new type of CRM. We believe that we’re doing that on a different type of data model, which is largely driven by conversational data, right. But no one sees that, you have to start with a wedge in the market and then you have to expand from there. Our wedge was and is, chat. Website chat was not new, website chat existed for my entire career. But something had changed externally. And that thing that had changed was that when we were using website chat a million years ago, it was not normal. It was just a bunch of us, geeks like myself, who thought this was normal to have a conversation with someone on a website. Because of that, it was largely relegated to a support use case. What was different when we started Drift, and back to the why now, was that from a messaging standpoint – we’ve all seen all the charts, whether it’s Mary Meeker’s or just the adoption rate of messaging had gone larger, maybe because the smartphone had gone from millions of us geeks or very few of us to basically everyone in the world, right? You, your parents, your grandparents, your kids, everyone defaulting to messaging. The reason that’s important is not the technology. The reason that it’s important is back to the user behavior change. Behavior change had happened. We had gone from a world where we had defaulted to in-person and phone calls and things like that, to a place where we all default now to some sort of messaging first, before getting to those other types of communications.
So, that was the ‘aha’ for us. We said, okay, if that’s true, we think we can now finally use live chat, and website messaging in an actual sales use case where it had never been used successfully before. Things had been hacked together, but no one had really done it because the sales use case was very different, right? If you’re a salesperson or a sales organization, the last thing you want to deal with is a support issue. If you ever get a support issue, if you know anything about salespeople, if they get one support issue, they will never use this channel again, right? Because they will say it’s noise and they will never use it. So the qualification part, and the targeting part, and the routing part, and all of these kinds of things that live chat never answered were really difficult things to figure out.
Those are the things that we set out to build. That is the thing that’s hidden underneath the iceberg of our live chat bot. From the surface it kind of looks similar to other things but when you get into global routing rules and you get into targeting, whether it’s ABM or other forms of targeting, whether you get into the qualification and what needs to happen before something gets surfaced as a sales-ready lead, or a conversation, there’s a lot that has to happen there. And so that’s really what we worked on in that context. And you’re right, we have continued to grow into other communication channels. Video is one of them. Not video like we’re doing now with Zoom, but asynchronous video.
So maybe on the live chat, the ABM part, so what does that mean? So ABM is, for anybody that is not familiar with Account-Based Marketing, that means that you have a targeted list of accounts. If you say FirstMark is one account that you’re targeting and I come on the website, then you’re going to recognize me? How does that work?
Yeah so, ABM is just a fancy word for what you just said, which is a target account list, which anyone who’s been in enterprise sales, many of my companies have done enterprise sales for 20 plus years – ABM is a new acronym wrapped around this kind of idea. But if you sell in the enterprise, most of the time, there’s a very limited set of customers for you in the world. So those are the only customers that you want to talk to. You don’t want to talk to Joe the plumber, you want to talk to a Fortune 100 company only. So what we do is, for those companies, that subset of customer that we have that has that kind of target customer list, we make sure either through qualification from a conversational standpoint or all sorts of targeting whether it is linking into emails, linking into various cookies, linking into IP-based rules, there’s a million different ways that we do this thing. We make sure that their people only get surfaced, companies that are in their target account list. It’s very important. If they get anything outside of that, then again, the whole thing fails, right? The whole thing doesn’t work in their view.
And then, so you’ll recognize me and then you route me to the right reps. So if you’re the account rep in charge of FirstMark the chat would go to them?
Yes, so that’s the next hard thing. So one is the recognizing you, that’s a hard thing. The second thing is okay, who owns this account? It might be me. I might own them. And where in the world am I? And am I available right now? And if I’m not available, how do I get to you? And then how do I resurface that thing in case I wasn’t available? And then how do I make that experience feel like a great seamless experience for the buyer, right? So how do I make that all happen? So that’s the complicated thing that we have to figure out and we have to do all the time and most systems again, when we think about all those systems, you could put together all the sales forces you want and all the add-ons and the 50 to 100 tools that people put on top of those existing systems today, none of them were built for a real time world.
Once you dig in from a technical standpoint, they were built for a world where real time didn’t matter, right? Because again, the paradigm was different. The company controlled the paradigm, they could get back to you whenever. That led to technical choices that are like batch based systems, very old school systems that take a long time to surface those things. None of that stuff works. We had to build our own enterprise level routing system that has to work in real time and make decisions that’s closer to, from a technical standpoint, it’s closer to an ad server than it is to what exists today in that world, right? Only out of need because that doesn’t exist today.
And then you have a chatbot product, right? Which is an automated version of this. What does that do? How does that work?
Sure. So, that was something that we brought in pretty early and we feel like we were probably the first ones that really did that from a website standpoint, which is like the sales qualification bot. To really figure out if you are qualified or who you should talk to. When we did it, we weren’t sure if it was going to work, if the timing was right yet, if anyone wanted to have a conversation with a bot and the truth is, the buyer doesn’t care if it’s a bot or not a bot or whatever, the buyer is selfish just like all of us, and we want an answer to our question immediately. That’s why it works. It doesn’t work because it’s a bot or not a bot. We’re kind of indifferent as buyers. If the option is I have a conversation with… as buyers, we just… If the option is I have a conversation with this thing, whatever this thing is, and I get an answer in real time, versus I fill out some form and I wait two weeks for someone to get back to me, I’m going to take this path.
That’s the way it works. But we started with a basic decision tree-based rule bot, which we still offer to some customers. But then we have a very sophisticated, truly machine learning-based bot, which works by ingesting the conversations that you’re having, whether those conversations are across email, website messaging, video. We ingest all that conversation data, and we continue to do that over time. Then we build an infinite model that actually takes into account the best performing reps and the best performing conversations and the context of all of that to build that out so that the company, our customer, doesn’t have to build any rules. They don’t build any rules. We ingest data, and we create all of that rule-based system for them.
Yeah. Since we are on the topic, there’s the concept of AI designer, right?
Yep. So, yep.
So, in very practical terms, but what does that mean? That means you say, “Hey, I want the Drift-powered chatbot,” and then it’s going to be somebody from the Drift team that’s going to just go through the history of conversations, for example, and then identify the composition of the most interesting?
Sure, sure. We have this idea of a system augmented by humans, which are these AI designers. And again, the reason that we do that is because our customer, and more importantly the buyer doesn’t care. The buyer doesn’t care about, just like when you search on Google or you do something on Uber, you do something on Facebook, I’m just using random examples, or Amazon, you just want the best result possible. Again, you don’t care if that was AI, hard-coded algorithm, AI plus 4,000 people that they have in India who are hand building a model. That stuff doesn’t matter. We’re too lost in that kind of thinking, I think, in the enterprise tech world of caring about software. Look at the way that we live our lives today. We use software all the time. Guess what? Nobody cares about software.
We’re still talking about things we did 10 years or 15 years ago, about hand building PCs. We care about PCs and specs and chips. And this, nobody cares about this. They care about an outcome. We buy outcomes now. We buy experiences. We don’t buy software. So that’s a whole another tangent I can go on. But the way the system works, to answer your question is that we take the AI designer, which is a human team, basically does not go through the conversations and sit there and say, “This one’s good and this one’s bad.” They deal with exceptions. The infinite thing in all data processing is that there’s infinite exceptions that are going to happen over time, they take those exceptions then, and then they make sure that we have the ability to train the next time an exception like that happens. And then that collectively, across the entire customer base, continues to get better and smarter over time. So that’s what they do. There’s a very limited window and amount of time that those AI designers actually spend in dealing with this. It’s very, very limited. It’s a very small team.
Interesting. And there’s that concept of learning across the network of customers, of data networks as well where the AI becomes smarter with each new customer that’s added to the network.
Yeah. And that’s more you would expect from a grammar and a context standpoint, not from any type of sensitive information.
That’s a really interesting topic because there’s always some debate about it, but that’s something that you’ve seen actually an improvement of the AI-
Oh yeah, 100%, 1,000%. And then we have to do that across many language sets. So we don’t only do that across English, but we have to do that across, and any of us who speak multiple languages understand there’s a lot of fixes that we have to do from a language standpoint when it comes to understanding other grammars and other slang.
Yeah. How does the interplay between the two products work? Because to some extent, they almost-
Compete. They almost compete with each other. Yeah. I think for the most successful, our biggest and most successful customers will use the automated product over time. I think it depends on the size of the company and the kind of the readiness that they have and how weary they are to go all the way to that end. And so a lot of our customers start with a decision tree-based approach, and then they go pretty quickly to that other approach because decision trees don’t work after a time. So anyone who’s used a marketing tool or sales-based tool or any business productivity tool at some point is dealing with these decision trees and trying to hard code and guess at what the next step will be. In a conversation, it’s almost impossible to actually guess those things. So they move pretty quickly to that end, but it’s obviously more expensive. It’s in our enterprise packages. So people have to graduate or self-elect into that place. But I think it’s the place that we want everyone to go in and I think over time, for sure it is the place where our entire product will look more like that than it does and get rid of all the decision trees pretty soon.
I’m sorry, to try to make sense, the future is automation first as opposed to-
100%. Across every product that we do, I think it goes back to my mini-rant that I just had there, I don’t think people care about software anymore. I think the reality that we’re in, especially in our context where we are in sales and marketing software, in the world of Salesforce and all those kinds of tools, is the idea that you need 50 to 100 tools, which is not an exaggeration. If you think it’s an exaggeration, you should spend time and audit the actual expenses of your teams, no matter how small or big. 50 to 100 tools that are loosely coupled together, that are passing data, or sort of trying to pass data, using Zapier, MuleSoft, et cetera, et cetera, Tray.io, a million other tools, and an entire business operations team that is dealing with that thing.
If you think that is the future of software, I think you’re mistaken and I disagree with you. I don’t think that’s the way the world is going. So that is where we are now and that’s just an artifact of time. I think we are moving to a totally different world. It doesn’t take much imagination because it looks like a lot of the software or the tools or the outcomes that we seek in our personal lives. Nothing works like that if you look at it in your personal life. Nothing works the way that we expect things to work in the B2B world in your personal life.
Speaking of which, how do you think about integrations with the rest of those tools in the enterprise? I mean, ultimately, are you trying, I mean, I’m sure you’re in this to win and to build the platform, but is the end game being the system of record for all conversations? Or are you partnering? Are you building? How do you think about it?
Yeah, I think it depends when. I think we have a very long view on what we’re trying to do. And that path is, from the very beginning, has been working with existing tools and systems today. Because I think the idea that you’re just going to rip something out and just going to replace something is not going to work. Those tools are going to exist for a very long time. And so we integrate into those things. We write back to those things. We are partners with all of those systems. But we think what we’re trying to do is create that layer or that interface for the people that are on your team and the automation, the bots let’s say, that actually have conversations with your customers.
So think about all the people that have conversations in the company with your customers or your prospects. That starts at sales, but that can be support. That can be your account management and your CX teams, et cetera, et cetera, your technical service teams. All of those people are having conversations with your customer today. We want to be the interface for those people. And then we will continue to write back and deal with all the systems underneath, but they shouldn’t have to worry about those systems.
Again, this idea that we have today of a CRM or a centralized place to keep all this information is a pretty insane system. This idea that humans enter things into one big ledger, and they have to enter it in a certain format. And all those humans actually have no incentive whatsoever to ever enter anything into this system. They have the opposite of incentive to put it in there. It’s the craziest system. So we like to do a lot of things, we like to abstract out and think of things and invert and think of things from just a human storytelling standpoint. And if you were to tell someone from a storytelling standpoint what a CRM today is, it would be the most insane idea ever. This idea that people enter stuff in and they type stuff in, and if it’s not in the perfect format and the perfect field, then the whole thing doesn’t work. It’s the most crazy thing ever, but we have accepted this reality, and we’re stuck in this reality. And I don’t think that’s the future.
Do you want to talk briefly maybe about the other products, the email product? What does that do?
Sure. So basically the idea that we continue to march down is, if there is a conversation that’s happening with the customer/the buyer, the buyer doesn’t only want to have a conversation with one channel. The buyer is kind of like what we did at Performable, is across all these channels, is this omni-channel thing. And during your buying journey and your customer journey, you’re going to bounce through all those things. You’re going to go from website chat, you’ll go through email. Sometimes you go back to video, asynchronous, synchronous video, synchronous being like Zoom that we’re using now. Sometimes you have a phone call, sometimes all of those things. And then if you get into a big company where you have lots of people involved with this product or this company, there’s lots of those conversations happening.
So for us, it was like, how do we start to get in the middle of those conversations? Because it’s important for us to capture that context and the conversation data. And more importantly, how do we provide a better experience for the buyer? Because again, fundamentally we believe that this paradigm shift has already happened, company control to buyer control. So, everything we do starts with inverting the existing model and saying like, “How do we make the buyer experience better?” So email, for example, we say, “Okay, email…” Again, if you think about it from a simple standpoint, it’s kind of a crazy system. We have marketing automation systems, and now we have sales engagement systems. That is an artifact of time. Back in the day, an email was a one-to-one communication tool only. Then we messed it up in some degree, from a marketing standpoint, and turned it into a one-to-many system.
But now we have a one-to-many system that is communicating with your customers that has no ability for those customers to ever reply back in a two way communication channel. So you get all these emails, you can’t reply back. Try replying back to a marketing email, it’ll bounce back and say, “No reply at, and don’t contact me.” Crazy idea. So we said like, “Okay, how do we take those marketing automation systems and sales engagement systems and make them a two way communication thing?” So if I was a company and I was marketing to you, I could say like, “Hey, do you want to come to the event?” And you could just say yes, not click on a link, open a landing page, fill out a form, re-engage the whole thing, get re-synched into 15 different systems.
The whole thing is insane. And so our email bot product basically does that. It allows those marketing automation systems. And again, we play with Marketo and Eloqua and all those systems you use today, but we turn it into a two way communication channel. More importantly, so it’s a better buyer experience, but more importantly, if they reply, that then goes back to the global routing and gets routed back to a salesperson so a salesperson has context, and they say, “Hey, Jack just replied to a marketing email, said he’s going to an event.” And they know about that event. That doesn’t exist in the world today. So that’s what we brought.
I read this whole sort of cool stuff as well where it captures data as well. So in your sort of signature block, if there’s your phone number, the system will automatically capture that in the CRM.
Absolutely. All of our systems have that in common. Whether it’s video, whether it’s chat, or in this case, email, it is constantly updating the system of record. Because again, the idea that a human is going to do that by hand is a crazy idea.
Very cool. So maybe switching gears a little bit away from product into go-to-markets, maybe talk about how you guys have been able to sort of differentiate in terms of marketing positioning, and especially in a category that’s pretty noisy. It looks like you’ve been working on creating categories, and I’m curious how you thought about it.
Sure. Thought a lot about it. So, I will first start by saying that I have a different view on commoditized markets versus not. I think when I was at HubSpot, we would religiously read blue ocean strategy and talk about blue oceans versus red oceans. I think everyone got obsessed with this idea of finding these perfect blue oceans where there’s zero competition and they create these things. My argument even back then was most of the time a market that has no competition is not a market. So therefore, you are cannibalizing from something, there is a red ocean somewhere. So my view, which kind of came from that perspective when I started Drift was, look, I want to go into the most commoditized market possible. Really, we’re trying to go into the CRM market. I want to go into that red ocean. But what I want to understand going in is understand that it’s a red ocean, understand that we need a fundamentally different paradigm and technical shift, so that we can re-segment and re-categorize this market so that we have an opportunity to win. That’s a very different approach than saying, “It’s a red ocean. I’m just going to build a me too product, and somehow I’m going to win because it’s nicer.” That’s not going to win. So, we had a very deliberate view. Part of that view was go into this commoditized market, have a wedge approach and then figure out, how do we expand that wedge over time? Can we create a category? What is that category? We created a category called conversational marketing, a name I hated because it was so long that I didn’t think anyone could spell it.
But then we woke up one day, we did that in 2016-’17, and then we woke up one day and then Gartner recognized it, Forrester, SiriusDecisions, all those people. Wiley asked us to write a book. We wrote a book on it in 2019. And now, everyone from Salesforce to whoever calls themself conversational marketing in some way. That’s nice. But for us, that had little to do with us, and it had to do with a shift that was happening in the world. And we didn’t know when other people would recognize that shift, but we thought if we were at all right in our belief, that all these companies would have to move into this space. We would have to go from a world where we were the only ones saying we were a conversational marketing tool to now we have at least 100 companies, a lot of them public and very large, who use this category definition.
And so that’s more validation of the market and the readiness, less about Drift for sure. So anyway, category creation is an interesting thing. Spent a lot of time thinking that. And because of where we are from a SaaS commoditization standpoint, at the very beginning of Drift, and we continued to do this, my view was we have to build a global brand. We have to build a global brand. I don’t want to build a tech brand. I don’t want to build a software brand. I don’t want to build a Boston brand or a San Francisco brand or a New York brand. I don’t want any of that. We have to fundamentally build a global brand because we are in a different part of the cycle right now where we are closer to Proctor & Gamble and selling Tide soap than we are in building Homebrew Computers and building Apple and early Microsoft. It’s just a different part of the cycle. And so you have to understand what part of the cycle you are in in your market, and then market or go to market appropriately.
How do you sell today? I mean, 50,000 customers is a huge number of customers, and so presumably you have a bottoms-up type of approach. How does that work?
Yeah. Lots of ways. I think we went into market in 2017. That was the first time we sold anything, which was a total touchless approach. Freemium, so everyone knows that. Touchless, we had no salespeople. That was it. And we got a lot of customers early on from that approach. And then in 2018, we added and we started to experiment with this idea of, can we add in some salespeople, and I had done this at my last company, Performable, I didn’t know it was going to work. I thought of it more as kind of an AB test to understand, what could they do versus this?
And we approached that way very much not that we were going to change our model, but that we had this ability to bring on this team – let’s see what they can do. Worst case is that they could sell nothing, or it was a highly inefficient model, but at least we would capture data and we would capture conversation. We would force conversations to happen every single day that weren’t happening. We could capture that from a data standpoint, and learn and adapt. So 2018, we did that and all of a sudden, wow, that started to really work. That model looked like a high velocity inside sales model, super high velocity model where, we would call when one chat closes, and they would basically close on one conversation kind of thing.
Those are typically more your SMB low, mid-market type business. Then over time, what has happened is we started getting more and more adoption from what we consider more enterprise type companies. Again, our definition, everyone has different segment definitions, but for us, they were in a classical sense, they’d be low to mid-enterprise type companies. Those companies started to adopt us largely in enterprise tech. Although now, it’s across all different verticals. And we started to see success.
Then what we realized, it’s all obvious in hindsight, was that in some ways we were a way better value proposition for those types of companies than all the rest, because they really had big problems in terms of all the stuff we talked about, global routing and all these different segments, different markets that we’re in, and different product lines and all this kind of stuff. The complexity of their product and the complexity of their workforce, meaning all of these salespeople, account managers, blah, blah, blah, etcetera, really made our product really valuable and really work there in a way that we hadn’t seen even in the other pockets where it works amazingly well. But in these pockets, it was amazing. Then from a payback standpoint, it was a very easy proposition. Because if you’re selling a 300,000, million dollar, $50 million kind of ACV-type product, it doesn’t take many conversions to pay for us from a product standpoint.
I would say those things are happening today. I think one of the biggest things that we saw that we didn’t think would happen this quickly is because of this pandemic that we’ve gone through, the forced digital transformation of the high enterprise is pretty insane for us. And so all these people had thought, right or wrong, that you can never close a $300,000 deal, a million dollar deal, $20 million deal without getting on a plane and steak dinners and this. And guess what? 100% of them across the world have to do it right now. And so we believe that we’ve seen a ten year acceleration in digital transformation in our market easily, and we see it every day.
Amazing. All right. That’s fascinating. I’ve got so many other questions, but want to give some time for the Q&A from folks on the chat. Jack, if you have some questions.
Thank you, David and Matt, for that fantastic conversation. So, let’s start with Nicholas’s question, David. What’s next on the road between the current chat environment and the new type of CRM that you propose?
I wish I knew. Nicholas, if you have the answer, please tell me. It’s an iterative process. I think for us, we look at it two ways. We look at it from fundamentally, how do we change the buyer experience for the better? And so we’re obsessed with that. That’s number one, that’s our guiding light. We’re constantly looking at, how do we make it a little bit better? Then we measure that by how much, from a pipeline standpoint and then a revenue standpoint, we’re able to affect our customers. We can track that throughout the entire cycle, even post-sale. So we track that, and that’s how we can tell that we’re having more of an effect.
From a product standpoint, I think what we look at is how do we get more of those people who are having conversations with your prospect buyer and then customer to actually spend more time in Drift, and that’s a little bit easier to measure. Our product roadmap looks like us spending a lot of time trying to answer that question. How do they spend more time? How do we automate more? How do we make the buyer experience simpler? If we can do those things, the outcome will be this kind of future CRM that we think of.
Knowing the importance of the feedback loop, especially with either other companies that you started, how did you change the way you ingrained that in the Drift culture early on?
Yeah. I love that question because it’s so near and dear to how I think and my heart. Because I would say that is fundamentally the most important thing at Drift. It was kind of a happy accident that I had at Performable in 2009 of forcing everyone in the company, because we were so small, to do support. And I had the aha moment that when we did that, when engineers finally did that and they did a kicking and screaming, that all of a sudden amazing things happened for our customers. Things that were fundamentally impossible before, when we were asking for them internally, magically were done overnight. All these things happened, and it was basically the discovery of the feedback loop. You had put in people who actually can change things as close to the customer as possible.
Having them feel that pain, have that empathy and then have the ability to change is way better than a few people at the top trying to dictate that throughout your company. So at Drift, it’s been fundamental. Basically, every single team in the company is measured in some way around how much time they actually spend with customers and how much time they’re forced into that uncomfortable situation of having to deal with customer and then buyer pain. We have our internal customers, but also our buyers. So that’s how we fundamentally think about things. We have all sorts of guardrails and systems and ways of thinking of including, accepting and then talking about all the time that we believe that every idea that we have is fundamentally wrong. It’s 100% wrong. Our job is to get out there, get feedback and figure out how off they are. It is 1% off, or is it 100% off? There is no idea that is perfect at the get go, no matter how many movies you watch.
That’s great. Thank you. So this one from David, and I know you touched briefly at the end of the discussion on COVID impact, but would just love to hear takeaways from COVID, how you’ve seen conversations change or anything you can share about that would be really interesting.
Sure. So at first, like everyone else, we didn’t know what was going on. We saw a hit in our kind of VSB, so our very small business and SMB segment, which then recovered in kind of May, June. Then mid-market was kind of a little affected. Then enterprise, we were always surprised from the beginning, was unaffected. That was in the early days. Then what we saw was a fast acceleration in the enterprise segments and somewhat in the mid-market. From a conversation standpoint, I think what we saw was kind of what I touched on before, which was all of a sudden, a lot more conversations were happening.
Because we can track revenue and kind of pipeline, a lot more conversations were happening with your higher dollar ACV-type products or services. All of a sudden, those big decision makers and those big sales that companies were making, more and more of those were happening online, versus offline, whether that was through email or on the website or through video. We’ve seen a massive change there happen. I think that behavior change won’t go back. Some of it will go back, but from an efficiency standpoint, it’s too efficient and the incentives are set up for companies to continue to go that way.
This one’s a little bit more straightforward, but how quickly are you able to stand up a Drift implementation in a standard B2B setting? Is it hours, days?
Hours, hours. Then it goes all the way on the other end of a massive enterprise customer who could take, a month I would say is probably the longest, but that has to do with humans and not technology. I would say is probably the longest, but that has to do with humans and not technology at all. That is change management 101.
Then outside of traditional B2B and B2C, are there any unique use cases you’re seeing? For example, local, state, federal government?
We haven’t. We haven’t branched off into those areas. I think we try to stay focused on… We think our market is considered purchases, right? There are two types of purchases online which are transactional. We do those every day, Amazon, etcetera. Then there are considered purchases. Then we narrowed that even more in the beginning of just to say B2B, but really there are all sorts of considered purchases. One of the ones that we love the most is a customer called Peloton, which all of us know, right? That’s a B2C thing, but that’s a highly considered purchase. They sell somewhere between 40 and 50% of their bikes over website messages. That would be the bikes, that would be descriptions, that would be the care involved in that. That is a massive amount of transactions that are happening over the system. So we see a lot of systems that going on right now in the B2C world. So it’s interesting.
Awesome. All right. Well, this is a, hopefully a softball for you, but Andrew asked what book do you find yourself recommending the most right now?
Oh, my goodness. For me, if you know me, I’m so obsessive around books and podcasts, so much about books that that’s a hard one. I think the one that I would say that I give away the most and that I re-read the most is Peter Drucker’s Managing Oneself, which is a $5 pamphlet, basically, that you can buy. Everyone should buy that. I think it’s important to re-read that book every three to five years, because every three or five years, you’re in an entirely different context and will take something away from it. It is tiny and you’ll read it maybe in an hour, and it’s worth your $5.
Very interesting. By the way, for anyone that’s not familiar with that part of David’s work, David has a site, he has a newsletter and a podcast on life lessons, which is really, really interesting. Where do they find that, David?
It’s my podcast called Seeking Wisdom, but you can just go to drift.com and you’ll see it easily.
Very cool. Well, on that note, thank you so much. This was really, really interesting. Really enjoyed it. So really appreciate your coming on today to share this with the group. Thank you so much, David.
Thank you so much for having me. Take care.