Exactly, a good place to vacation.
Good afternoon. We'll go ahead and kick things off here. My name is Brett Bracelin. I'm the co-head of tech research here at Piper Sandler. We have Ed Hallen, the Co-founder and Chief Product Officer of Klaviyo. Ed, thank you for joining us in Nashville.
Excited to be here.
Absolutely. So listen, as we think about Klaviyo and opportunity around Klaviyo, three core tenets in the thesis that we have: differentiated, custom-specific database architecture, this ELG or ecosystem-led model that's helping drive efficient growth, and this idea that you're a must-have product-
... with a very clear ROI. Before we drill down into those three tenets, maybe let's start with the origin story. Let's start with when you first met AB and what led to the start of this idea around Klaviyo.
Yeah. So, before Klaviyo, we were both early employees at a business intelligence software company. And what the company fundamentally did was to help large-scale enterprises, really the Fortune 500, customers like Starbucks, Walmart, Bank of America. And ultimately, that business sold to Mastercard, right before we went public, as it was thinking about going public. But what Andrew and I spent a bunch of our time on was solving the data problems that those businesses ran into.
So what we'd see is that if you're Starbucks and you're thinking about your pricing strategy, and you wanna, you know, figure out how to optimize that, the system that understood what your sales were was different than the system into your inventory, and then was definitely different than systems like your time clock data, that knew whether the store was short-staffed that day, or even the real estate system, which you could look up and figure out if it rained for a given set of hours during the day. So we'd spent a lot of time dealing with very large data sets, unifying that data, and then making that data available with very low latency. So, coming off that, one of our first...
When we paired up and decided that we really wanted to launch something, one of our first theses was really saying, you know, businesses largely don't have a singular customer source of data source of record. And so if they wanna figure out, hey, who are the customers who bought X, Y, and Z, who were on the website or not on the website in a given period of time, who then returned things, who bought things during a discount? All that data lived in different systems and invariably involved engineers, and invariably involved SQL. It was very tough to get access to.
So we took the problem that we'd spent years solving previously and said, "Okay, we will build this customer source of record for businesses." We were bootstrapping, so pretty early on, we realized, okay, no one's gonna buy this from two guys effectively sitting in a basement. So how can we drive a business outcome with what we're selling? And that'll make the sales process much easier.
So that's when we said, "Okay, if we stick email atop this customer source of record, we can actually drive business outcomes, and then we can also measure whether or not the actions that someone took worked and the size of those outcomes." And so for our first, you know, 10, 20 customers, that was a complete mix of some e-commerce, some hospitality, some SaaS companies, and we then leaned in and said, "Okay, we've got to pick one. We've got to focus," and focused on e-commerce as the place where we were gonna build and focus within that.
Really, we were always started with email largely 'cause it was the channel that most of our customers wanted to use, but we were always channel agnostic in terms of how we're gonna reach out to customers, but all powered by this customer source of record.
So you start out with this customer source of record. What, why focus on retail? What was the unique maybe about retail? Was it just your exposure, and that's some of the customers you worked with before in your prior kind of era? Why, why retail?
Yeah. So, to start off, so we, you know, of those 20 customers, just a handful were retail. The reason we started, you know, really built the business around one set of customers, is we knew we had to focus. We knew that if we weren't gonna raise tons of money, we had to focus on a similar, you know, one set of customers. And we also knew that the retail business was built around platforms, both in terms of the platforms like people like Shopify, who have, you know, data on consumers, but also there was an ecosystem that existed around those businesses.
And so we focused really on not just retail, but really on e-commerce within that, as a beachhead to say, okay, if we can build a profitable channel to find customers in one area first, then we would eventually expand into the other industries from there. And so the product was always built to serve all different types of customers, but really the go-to-market was architected to be as capital efficient as possible around one singular market.
So this customer-centric, retail-centric data store, very specific schema around the customers that you built. Why marketing? As you think about the data platform, you can go to all sorts of different applications. How did you land on marketing as the first one?
Yeah. So the fundamental belief was, it is easiest to sell something when you are directly driving revenue. And so we knew that an advantage we had was, having lived through a company that focused entirely on intelligence, we were always beholden to going out and somebody else taking actions to drive income and driving business growth. With Klaviyo, if we could stick email atop it and directly drive the revenue, it'd be much easier to sell. And so in our early sales meetings, we could have someone sign up and say: Look, you will use Klaviyo. You will drive, you know, X dollars for every dollar you spend with us. You'll be able to measure that in Klaviyo and see exactly how many dollars that's delivering.
We knew that that'd be a very powerful sales proposition, and so marketing was the best, one of the best places to start, because there was a deep desire to do that. I think we also looked at the set of incumbents, who largely had focused purely on the, at the time, tough problem of delivering email. They'd all built their own email infrastructure, had really focused on just how do I get one message from a company to a whole bunch of people, but not on the personalization side. Coming from the data side, we knew there was a huge opportunity to optimize those activities to actually drive a ton of revenue.
So marketing made a ton of sense, and then with e- commerce, we quickly saw that because of the economics of spending on ads, it was one of the most effective ways to spend money and drive store and merchant growth.
Historically, marketing as a category has been viewed discretionary, and this is back to that third tenet of a must-have versus a nice-to-have product.
When tough times, we've seen those marketing budgets and ad budgets contract, and when people want to expand again and the economy's healthier, we've seen those budgets expand. Your business and growth has been a little more resilient.
As you think about one of those attributes, being a data-driven model, how important is that clear ROI driving revenue for the business? Obviously, it's different than others. So two parts: How important is it, driving that clear ROI and KAV? And then two, why aren't other email marketing platforms doing that?
-unique?
Yeah. So, and we think it's fundamental. Like it. You know, where marketing spend is discretionary, it's like the closer you get to brand and the closer you get to not being able to quantify. But where you can truly measure an impact and understand what that is, it's incredibly valuable for the marketer, but also as part of our pitch. You know, I think we've seen a lot of, you know, now other marketing platforms incorporate this in some way. I think that the key thing we've seen is that they're built. They're either built entirely separately from data, so they still exist in a world where they can send messages but not understand the impact that message actually drove, or they're doing it in a very superficial way.
So they have a slim slice of data, or they're, you know, cutting down the amount of data that's available, and you really need access to all the data to understand the impact that you're driving. And so, because we came from solving the database side and the data problem first, it was natural for us to then apply this immediately to understand the impact. Versus if you're coming from the marketing side and you lack the fundamental data infrastructure, it's a much harder problem to solve or a much harder problem to solve credibly. So there's a tendency to say, "Yes, we're doing this attribution," but quite frankly, it doesn't tie out to, in a way that is believable to the merchant.
So our estimates, we have you approaching a billion-dollar business exiting this year, largely in this marketing category. Within retail, you have these customer-specific data points, this data-driven model. How difficult would it be to add another app optimized for retail beyond marketing?
Yeah, I mean, we've always believed that most software should exist with this customer source of record underlying it. And already, if you look at just where, you know, what systems customer information come from, it's not just marketing. And so having both unifying the communications across different tools, as well as then having that deep understanding of who a customer is, have the potential to just be much better software and to drive much, much better outcomes. And so, it's, you know, really, from day one, it was something we spent a lot of time talking about and something we, you know, architected the data infrastructure around.
Yeah. How would it be hard to do?
No, fundamentally not. So, the difficult problem is fundamentally, how do I understand all the customer data over all time and make that available very, very quickly, make that usable, let that power actions? But whether those actions are marketing, whether those actions are conversations back and forth with a customer, you know, it's not a fundamentally heavy lift for the product side of the platform.
So is it a like a data architecture kinda issue, or would it be a you know a schema, how you do that?
No, I think. I mean, really, it's, you know, the data. You know, we think of the platform as the app built atop, so marketing built atop the data infrastructure, and it's really just adding another set of activities atop the existing data infrastructure, so really not a big lift for us.
Okay. Interesting. Let's talk a little about Shopify, right?
As you think about that opportunity, this ecosystem-led growth model is really driving very efficient growth, even below a billion-dollar run rate, yet your 10+% op margin model. Walk us through how that Shopify ELG relationship evolved?
So it very much grew out of kind of an early focus on bootstrapping and early, you know, lack of capital. And so we knew that we needed to find the most efficient marketing channels to drive leads for us and to drive our initial growth. And so in doing so, we headed, you know, the three motions we initially built were agency partnerships, so working with the other people who worked with e-commerce stores, think marketing agencies, a strong PLG motion of people coming in the door via word of mouth. And then the third was a very intentional focus on platform partnerships, and specifically like Shopify.
And so our original theory was, as we saw, you know, a handful of initial Shopify stores, was how can we align our incentives with any platform, but in this case, Shopify, such that, well, you know, the actions we would take would be beneficial to them, and therefore, we could, you know, effectively ride, you know, grow more quickly without needing to spend resources ourselves. So the way we did this was, one, because, you know, our fundamental focus, like as someone uses Klaviyo, they're seeing the dollars of growth they're driving. We knew that every dollar of growth we drove for a store was revenue that Shopify received-
... as well. Second, as Shopify launched new functionality, we would quickly build into their functionality such that in Shopify sales process, if they were going to upsell a customer on, say, Shopify Plus, the best way to use that functionality and upsell them was actually to also talk about Klaviyo, because we could help drive that source growth, so we very intentionally would move in sync with Shopify, and then build closer and closer relationships on their sales team, on their success teams, largely driven by our customers, so getting to know the people that they work with Shopify, and then over time, with Shopify writ large, but it was always driven by a belief that if we could align common incentives, and help Shopify grow faster, that we knew we could grow faster.
And it's how we approached every platform at the same time, and still how we approach platforms.
Can you talk about maybe the Shopify partnership opportunity? What are you doing today? Are there more things you can do with Shopify? And then conversely, I actually get questions around what happens if Shopify goes away?
So kind of two parts of it: opportunity and then the risk.
So on the opportunity side, I mean, I think we still think, you know, there's a huge, it's a great partnership. We've worked very closely. We still work very closely. There's a lot of opportunity to grow together. We've continued to see this play out in different market segments and different geographies, that, you know, if we can deliver on this promise to fundamentally help their customers grow faster, it makes for a very symbiotic partnership, where, you know, we can grow into the same markets together, and as they grow, we grow at the same time. On the flip side, you know, if, you know, what happens if Shopify goes away? We very much, you know, have always built the model to stand on its own.
So, you know, the vast majority of our customers, they may come over at the same time as they're, you know, re-platforming of Shopify, or they may come via a totally different channel. And so what we see is that, yes, Shopify is a wonderful partner who we expect to keep growing with, but, you know, fundamentally, most of our growth is not being driven by Shopify as the channel partner.
Hard to have a discussion without talking about AI. I think we went a good 18 minutes without talking about AI. But let's touch on AI. Obviously, you're still involved in product pretty heavily. As you think about the role of AI in sales and marketing, walk us through your thoughts around how it's rolled out.
Yeah, so we think about AI in two fundamental ways, so the first, which seems to get a lot of the airtime lately, is generative AI and how do you produce content, and so, you know, we see a lot of, you know, and typically, this plays out for us in things like automatically generating SMS, automatically generating text and email, or generating imagery in email, the ability to, you know, send a bunch of different variations of a given marketing message to someone, and then to individually tailor, you know, one piece of content to one person, one to a different customer segment, and have all that happen under the hood, so it's very impactful.
We tend to see. You know, it does, you know, fundamentally drive KAV, but in terms of usage, what we tend to see is actually, we thought, hey, what we might see is that the individual marketer would spend less time, you know, drafting an email, drafting an SMS, doing you know, setting up something in Klaviyo. But what we actually see is that it's the same amount of time, but more time spent on the high-value activities. So starting with AI, prompting a whole bunch of different pieces of content and then picking the best one. So I think that's one that we see, both for ourselves and many other marketing platforms, being important.
We're most focused on the second category, though, which we really see as using AI to drive higher KAV by effectively increasing the ability to kind of tell our customers what's the next best thing they can do, whether that's automatically setting up automated emails and campaigns and flows, or it's analyzing the customer base to identify pockets of opportunity to figure out what's going on, and that's, you know, fundamentally driven by having access to that full life cycle of all customer data over all time, and then being able to understand, okay, when you take action X, Y, and Z, this is what's likely to happen.
Here are the actions you haven't taken, here's the type of business you are, and then shortening the cycle for a marketer to actually get that set up and implemented and start measuring the results of whether or not that worked. So I think overall, I think we think both are fundamentally important, but it's really the more that you can build, well, do what we call, you know, Klaviyo playing the game of Klaviyo, the more that we can kind of give marketers superpowers to understand all the different revenue opportunities that exist and then help them get to those faster, the more powerful Klaviyo can be for our customers.
If successful-
... in driving more revenue for customers, how do you think about monetization? One of the things we think about is maybe a move in the SaaS world after twenty years of this transition to seat-based pricing, or think about outcome, you know, based-
on pricing and these add-ons. Like, how do you think about in that world where you can drive recommendations and automate campaigns, how do you think about monetization of that?
Yeah, I mean, we fundamentally... Yeah, we think all software eventually go to be highly outcome-focused.
It will replace the human activities that people don't want to do and leave humans to focus on the activities you do want to do. If you think about, this is, you know, what makes TurboTax nice, but really could go even further, is like, you want to spend as little time as possible on your taxes. You want to just provide the high leverage where you actually can. We think about, you know, you know, Klaviyo as a revenue engine that effectively should be the outcome engine that drives revenue, where we replace as many human activity as we can and then lets humans focus on the places they can add a ton of value. And at the end of the day, this is what makes it very, this is what keeps it from being discretionary-
is that you are directly measuring that. And over time, because you're driving revenue, you're measuring that revenue, we think it we don't charge in form of take rate, but we think about it more in terms of, like, people are willing to pay for the revenue you generate, and that's how they should pay for software. They should pay in terms of the impact you have. And so we think that will, you know, become even more true in the future.
I like that. The revenue outcome engine. That's a good tagline for what Klaviyo actually is at the core. As you think about evolution of the business, the fastest growing part of your business is large enterprises. A little surprising, full disclosure, you know, but walk us through why these large customers are now kind of coming to Klaviyo, when historically the base of the business was smaller customers.
Yeah. I mean, it's really, the pitch on the small side is the same as the pitch on the large side, which is fundamentally having this customer source of record, compile the first- party data, understand everything they've done, lets marketers drive more revenue. And so, you know, we've seen over time that both some percentage of our tiniest customers have grown up to be the big brands of today, but also then more and more of big customers like Samsonite are coming to us first. So, we see it's fundamentally the same pitch, it's fundamentally the same things. The competitive set is different, but for marketers, it's really about being able to drive more revenue.
I'll take any questions from the audience, if there's any. If not, I'll have one more question here, and we'll try to get us back on time. Any audience questions? Perfect. Let's end with what are you most excited about for next year? It could be the product. You can't preannounce stuff, but.
We've got people monitoring, Jack. But, what are you most excited about for next year?
Yeah, I think, you know, we continue to be incredibly excited about just the things we're doing and the ways that effectively that you know we can keep growing what we call Klaviyo Attributed Value but how we keep being that revenue engine. And so when we talk about AI, it continues to be you know what's the next thing we can actually do? Take this full customer source of record to do to actually drive more dollars for our customers, and at the end of the day, that flows directly back to us. It's incredibly valuable for our customers, but it's also you know what you know helps drive Klaviyo's growth. So I think we're super bullish on just the more ways, the tailwind that we have is leveraging this complete customer view to drive even more growth for our customers.
Ed, thank you so much for taking the time here. Welcome back to your roots in Tennessee.
Thanks.
Thank you.