The Secrets Behind Creating Scalable Products with Box CEO, Aaron Levie and SaaStr CEO, Jason Lemkin

Well, everybody. Well, let's welcome back Aaron Levy, who maybe not intentionally, has had a great journey with Saster. Aaron was the first speaker at the 2015 annual One Week after the the Box IPO. You might not remember, but I remember.

Okay.

Yeah. You were tired.

I'm always tired. That makes sense.

Yeah. And. And I remember you remember, but I.

Don'T think I was like Covid tired, though. Like, this is like pandemic tired is a whole different stage.

I think it was tired of talking to 100 analysts of whom three understood what recurring revenue was. Right.

Makes sense.

Yep.

And NRR and all of these crazy metrics.

So we'll fast forward to today. The title. Can you see it maybe down on the screen here? Zero down. We were, we were asked to talk about behind scalable products, so we'll at least come close. I do want to talk about multiple products at first. And so what I wanted to do is I grabbed just a few slides from Box's public company stuff and our. And our five interesting learning series. So maybe just to ground everyone knows what Box is, but not everyone might know your metrics, where you are today. That's possible, right?

Yeah.

Yes. So we're coming up on that billion dollar in revenue we talked about on TechCrunch a few years ago, right?

Yep. So FY21 is last year for us, so we'll do about 855, 60 million in revenue. I forget the last GUID range. But. But in that kind of rough scale.

Yeah. And if you had. And we'll get into it, but if you had to summarize how Box is different on the other side of a billion.

Yeah.

What's. What's. What is that other side of a billion?

I think actually it is. You know, it's so cliche because we all like everybody here and we wanted to always build a platform and. But like that literally is the answer. You have to build a platform. And we. But I think the way you build a platform that now is hopefully doctrine at this point is you build it through initial use cases that actually find your product market fit and then you're able to evolve those use cases until you have a full comprehensive platform. And so that was kind of our journey. We started as a. I don't. I mean, probably nobody remembers like USB drives and FTP accounts and just like everyone remember. So maybe a couple people like back in like 2005, we were on Blackberries and you're on Internet Explorer and you had USB thumb drives. And so we were just like, let's build a better way to share files and access data from anywhere. We started Box to have that vision. We did not know the word platform. There was no sort of conceptualization of what that would mean and that element of the vision. But literally within about a year people started saying, hey, can we integrate Box into other applications? And they were like, okay, well we'll create APIs. So then we had APIs. And then eventually customers came and said, hey, we need advanced data governance functionality. And so we said, okay, we'll build a governance module. And then eventually we stopped getting sort of like totally pulled into these reactively. We decided that was actually going to be the strategy. So we would build out a multi product suite and then eventually a full platform approach to go and address these use cases of how companies manage all of their content. And so I think at the other side of a billion, Box looks much more like a platform and a suite of multiple sort of categories of products all built on one architecture that can hopefully get procured by the customer in one simple package and then they deploy us across their organization. That's kind of the journey we've been on. So it was one product that was very, I think most people would have just said, hey, we're a feature. And actually we saw that as a badge of honor initially because end users don't want to buy platforms, they want to buy features. And so we got in through the feature and then we were able to ultimately turn that into a platform. And that kind of brings us to today.

Well, it is interesting. Everyone needs a wedge, a 10x feature, right? And so that, that, that ism. That tritism of a feature versus solution is pretty stupid if you've been out in the field, isn't it?

It's, it is because the, the, you know, your, your especially incumbent competition is always trying to deposition you as a feature and, but, but the, what they don't realize is, you know, hopefully what you're building behind the scenes. And I think the thing that, that I look back on is as maybe one of the biggest lessons of this, of that journey is architecture matters more than almost anything else you can imagine in that journey. Because if your feature is then completely disconnected to the next feature and then thus the next feature, then you really aren't on your way to building a platform. And so we made a bunch of now in retrospect kind of lucky, fortuitous decisions in our platform architecture that were not intentional. We just lucked out that we made them the right way. And then eventually they started to become Much more intentional because we knew that we were building a platform. But we. One of those small things that you really. Actually, it's worth paying attention to even when you're just three people in a garage.

Well, that's interesting. I hadn't thought about talking about that. One reason I think you tweet a lot about this, one reason I think startups can always beat folks that come behind them is their platforms are always at least 10 years old.

Yep.

I mean by the time, by the time, even with the pace we grow today, everything that you might want and so there's always a place where that architecture or solution is just almost too dated that they're not even going to bother. Right. What. And box originally is. We're at whatever we want to call it, SAS 1.2.0, we're in web three right now. But, but we started so early. What, what were the, what was something that you actually got right that you're talking about? Because it was so easy to get things wrong in the early days from scalability. What, what's an example of what you're referring to there?

There's a litany of things. So we, I mean first of all, now again, it's kind of ridiculous to even talk about this now, but, but just we were multi tenant from day one and we made no exceptions to that. So when customers said, hey, can I go run this inside my data center? We said no, you can't. And we lost. I could probably add up maybe 100 million in revenue, I would bet in the.

Easily maybe even today still, right?

Yeah, yeah, absolutely. There's governments globally that we're not serving because of that requirement. But you can imagine when you only have like 15 sales reps and you know, three of them are losing major deals because you're not on prem. That's very painful. And like that, that becomes a major topic of challenge for the business. So we basically rejected all forms of any modification to the software. Any chance of having sort of fragmented architecture. It was a sort of, you know, the Henry Ford. You can have any color as long as it's black. And like we have one piece of software and it's going to be in our cloud and this is how you're going to deploy it. And we had a lot of people say, well that's not how SharePoint works or Documentum works or some of our, you know, kind of incumbent competition. And that point of view and principled approach meant that we, you know, filtered down to the customers that actually wanted exactly what we had. We could learn from them instead of being fractured in what we were doing. And that just continues to compound over time. So one multi tenant architecture all on the same data model, which ended up being another sort of important aspect of the, of the business, which is, and this is just esoteric inside baseball for what we do, which is you store one asset once and then you can do anything against that single asset as opposed to again, there's lots of different places where your files might be and where your content might go. So that single data model, single multi tenant architecture was really important. And then because that was all abstracted from infrastructure, this was the key point. We never sustainably fell behind from what you could actually do with computing, with software. Because even though we started in our own data centers, because believe it or not, we literally launched the company before AWS existed. So we started in our own data centers and then we got to leverage the public cloud and then we did even more in our data centers, but then now we're doing even more in the public cloud. All the while, customers never had to do a single thing or even worry about it once. So that's all happening in a way that's fully abstracted from the customer. And now at this point, we're moving even more into the public cloud, leveraging every possible available service. And again, the magic is your customer is just going to get all of those tailwinds benefiting their deployment. And so I'm very religious. I probably only have one religion. And it's anti on prem you have to be cloud. And it's because of this velocity of innovation that you get by being able to have a single architecture have you.

That makes complete sense to me given where the world has gone. No temptation to do anything that's hybrid cloud or hybrid something. Even now you have the scale, you're coming up on a billion. Right. You could probably put together a team now. There's just so much money. If you're willing to do it, the money's more than ever, isn't it?

There is.

And you get $100 million contract if you're willing to do some of these hybrid offerings.

There is, and I mean clearly there's absolutely a way now at this point to not deal with the baggage that we had 15 years ago where if you deploy something on prem, you know, almost instantaneously, you're on a wrong version. And we can solve it out of date forever. Yeah, we can now solve that. I would still posit, like if you really just think about in practice though, think about every time that Amazon or Azure or others have launched these sort of on prem versions of their clouds. These are all relatively subscale versus just the pure public cloud infrastructure benefits that customers are getting. So we're still staying the course on our strategy, I guess never say never, but in this case we're big believers.

One follow up question that I want to make sure we have time for product, but I guess it is related you platform, right. And I think one of the things you that we've learned over time is as you approach 100 million, right, which is now an order, an order of magnitude less than where you're going to be, you want to start being multi product potentially, which we'll talk about in a minute, and you want to start being a platform for others. And Box was early, but it's interesting that your vision, it's still evolving now for this platform, right. Are there one or two things and we'll actually look at the platform on the next slide and maybe I'll flip to it, except I can't go back, so I'm cautious. But are there one or two things you wish you'd done earlier on your platform, looking back, that you would have pulled them down earlier?

Yeah, I think there was. I think, yeah. I mean every, I just say as a general statement, every feature that today is or product that today is performing well, I wish I had done earlier. So there's like unlimited, it's a given, right? Yeah. So like I wish I could do a lot of that faster. Sometimes you have to find your way of which products are working. And I'd say every maybe year we have two or three bets going on in the business of new areas of innovation. And some years all of them work. Some years you look back and you say, okay, only one of those worked or maybe we even skipped a year of actually kind of that type of value that we want to drive. But in general, you're always planting a few seeds and then while you're working and incubating the ones that you planted the prior year. And so today data security is a big area. Workflow automation is a big area. We're now, thanks to your inventions, we're now doing a lot more in E Signature. So we appreciate all the wars that you had to fight before us on that front. So we're going to have multiple pillars of new capabilities all on one platform and then integrated together and then offered as APIs to anybody else that wants to build on Box as well. So that's kind of the vision. But I wish I did all of the stuff that's working sooner.

Yep. And is how much of your time now is co. Do you get to spend on these two to three big bets? What's, what's your. How much of your 60 hours a week, how is it split up at scale?

Certainly when we're not in a proxy contest, I can spend more.

I wasn't even going to bring that up, but congratulations. I think you beat that by a higher margin than anyone at all time.

Right. About seven people were paying attention to that. So thank you for, for your vote of confidence.

I was there. I mean, literally, I think you beat the contest by more than like anybody else of any proxy.

We're happy about the win. Yeah. But ultimately the proxy contest is we just believe shareholders agreeing with our strategy. So it's sort of not even that much more maybe celebratory than that. But ultimately I spend as much time as I can on our new bets. But as again, everybody here knows you got to equally balance that with customers and investors and other dynamics that we all face as founders.

What about today? Customers versus prospects. How. What is, how much do they trick the CEO into spending his time with? How do you balance the two?

It's actually more that, yeah, I trick myself into it. So I am, I am willing to talk to anybody interested, remotely interested in moving their content to the cloud. So I like, you know, 90% of the time I'll leave some zoom call with a prospect. I'm like, that deal is going to close next week. We got this. It's like a year and a half later. It's like, okay, finally we're closing it. So there's a sort of optimism that, that refreshes on a daily basis as a, as a founder, but so a lot of time spent with customers. But, you know, we've definitely shifted. So we have 100,000 companies on the platform. So that means, you know, almost by definition, the vast majority of our bookings and revenue are going to be from our existing installers.

That's why I asked the question. The majority of your new revenue will hit it is from the base. So a CEO, do you align your time with that or do you not align your time with that?

You align your time with where ultimately the dollars are going to come from. And so in the base, we often will have lots of customers that have significant expansion opportunity and that can look as important or be as important as a net new logo that we have to go get. Yeah.

So we'll get to that in a minute. One just niche question. When I see this. Just curious. So 96% subscription revenue as you go, more solution and even more enterprise. Have you thought about pushing that number lower, doing more services for these big deals?

It's actually crazy. I've, I've been trying to push that number lower for years. Lower. I can't. And, and the reason I can't is because there is this sort of dichotomy where the simpler the software is and the easier you make it to implement that number is that recurring side is going to be higher. And so there's this inherent tension which is hopefully if we're building good software for our customers, that the amount of extra services they need should be less. And the more you become a platform, probably the more that system integrators and other partners do it for you. So I actually think it'll probably remain this. Even though we do become more solution oriented, it's because we'll have a bigger partner ecosystem. We still want to build very simple software that's easy to implement. But I think there's no magic number. There's plenty of amazing SaaS, companies that have 15% of their revenue as consulting revenue and those companies are doing unbelievably well.

So even though Qualtrics is 20% services, you probably could do it if you wanted to. Right.

But I've tried like to, I've tried to get it to like eight and I can't. So I think every company finds their natural spot for their product, for the market that they're in. Maybe this will, maybe, maybe the number will get higher over time. But I'd say right now, you know, if I were talking to an investor, I would say it's, you know, expected to hover around this range. Cool.

We could spend all our time on this slide. Let me make sure, just a quickie. And I don't want to spend all of our time on the future of work that we talked about. Right. But while we have you, this is, this chart you have on NRR is interesting. Right. So there was a little bit of a deceleration during, during the global pandemic. And we're back. So I think we know that firing on all cylinders. What more, maybe more tactically or specifically what. What's coming next in the next two years, right. Beyond the world being pulled ahead two years and, and E signatures that the light going off. But what, what, what really is going to be different in the next two years?

Well, in, I think, you know, generally for everybody in this, in this room and beyond, what's going to be different is that, is that every company is going to be a hybrid work organization. And even for the most sort of ardent fans of offices, every IT organization on the planet is going to have to implement hybrid work. So you got basically the singular largest megatrend in enterprise software history is playing out right now.

And is it early? Is it just starting?

Really though, it is, because if you literally look at people's vendor stack and if you were to go interview any CIO of any top 10,000 enterprise, the majority of their dollars are going into brand names that existed a decade ago. So that kind of tells you we're still pretty early. Because it can't possibly be the case that for all of the world's largest organizations, all the software that they're using is still the Same software as 10 years ago when we didn't have hybrid work as the main fixture of how we operate. So something still is yet to completely go and actually change and make these markets continue to evolve. So you got hybrid work. Everybody knows everything's going to go digital. There's literally not going to be anything other than a digital transaction. So think about all the software that's to get written to go and help every company in every industry and every size organization all around the world be able to process everything digitally. We're still in the very early innings of that. And then you've got the mega tailwind and just tectonic shift of security, privacy, global data privacy challenges that every organization is facing. So you compound those three things and there just have never been more tailwinds in enterprise software than right now for this type of transformation. When you and I were in the early days at best, we were like iPhone3 and AWS were like the tailwinds and that was it, that was the disruption. And today you have entire ways that we're shifting all of society and how we work, plus every possible technology tailwind that you could imagine. So that's the macro view for everybody. And then for us, we think the biggest impact for our business is that that just means there's going to be a wider generation of unstructured content and data. Every interaction that you do with a client, a partner, a supplier, there's going to be some amount of data and content that you're going to work with. Financial data, contracts, invoices, marketing assets, media research, data, knowledge management, all of that has to be stored somewhere, managed workflows around it and then integrated across applications. So that's what we're going to do.

So one interesting thing I want to follow up on because maybe it's obvious, but I hadn't thought about it too brought it up is what's going to happen in the CIO's office, not now, but in the next couple years. Because when Covid hit there were certain categories. Zoom's the most obvious. But it's not just Zoom, right. It's call center. Right. I mean, good God, that space exploded E signatures that we didn't quite get. And that happened because it had to happen. Right, but now you're talking to CIOs and they've got 10 year old software that they're going to replace because of these changes. But it's not all happening this week. Right. So is this, is this post Covid buying cycle going to take three to five years actually to play out? It's not done. Right.

Well that was the thing. So, so you know, every, every company that, that, that was public certainly than anybody else sort of had this line last year of we saw, you know, 10 years of acceleration in the past year and that was true. But it was usually with existing customers that were seeing bursts of new use cases.

Maybe Zoom got all pulled in, dramatic.

All got pulled in. But it was because just literally the demand of whatever the indirect relationship was, the B2B2C element was the pull in. So Twilio obviously exploded because we're all buying doordash all day long. And so by definition Twilio explodes. Zoom obviously explodes because of the obvious reason. So but that was like, that was like the urgent sort of triage, sort of keep businesses afloat. That was not the we're going to now automate our business processes. That was not the sort of we're going to reimagine, you know, how we serve our customers over the long run. That was not the, you know, we're going to change the actual way that we work and collaborate and ideate and you know, engineer things. That's, that's, those are things that do take two years and three years and five years because there's software to get built, there is culture to change, there's organizational behavior dynamics that have to evolve. And that's I think the era that we're entering now. So last year was the sort of like you could see rapid expansion because everybody was in this sort of crisis mode. And so maybe the actual, you know, sort of rate of change is going to be slower, but the size of the shift is as big as it ever has been.

Yeah, it's interesting the we might have hit this slide expired for the next.

Few days of growth. I think, I think all your, you know, you're just finding my most inane tweets. So I probably not all tweets. I promise I tweet other things than just like I'm so fired up right now. Like. Okay, okay. I just, I'm so pumped. Hold on, let me skip to another one.

Talk just a little bit. And maybe this is not, not any more interesting than data, but this is. Talk a little bit about going up market and what you've learned and what's interesting. We have a later slide. Maybe we'll get to. We won't. I think about how you. Everyone characterizes the lines between categories differently. But I think today you say you're about to half enterprise. 50 something ish. Right. Maybe 8% still self service. Right. It's not gone forever. Did you say 8%? I think. Do I have that wrong?

I got a maybe on revenue. I don't know the latest.

Not allowed to go back. Right. I would get it. And then 30 or 40% is sort of mid market SMB. Right. And what have you learned here in terms of. But, but the. You've got two big growth factors quantitatively. One is selling more and more to the base.

Right.

Driving up acv, I think. I'm not sure. Right. Yep. But then more to bigger customers. Outsized growth. Right. So what's changed at box as you go? More Enterprise at 800 million.

Yeah. So you know, it's interesting. I'm. I've. I've decided that like I don't need to get any points for the thing I'm about to say. But I do want to be remembered as, as still having this as a core part of the strategy. But I just don't care to talk about it that much. A lot of founders will come to me and they'll be like how did you do enterprise sales at the beginning? And I was like, I didn't. We didn't. Like we have always been a freemium product led growth company. We just stopped talking about it because it's just like, it's just not that much to talk about at some point. So these numbers are super important, but they don't come just because we do these super big enterprise deals all the time. They come because we have instrumented every element of our product to be self serve, freemium and viral. And then we have an enterprise sales team that goes and works with customers to turn those deployments into much more strategic relationships. And that was the balance that we lucked into early that we made the bet on and it just, that's a flywheel that if you can get humming almost just takes care of itself. So our model is most of our upsells of customers are the customer has already deployed the software, they're already using it and it's just spreading and then we go in and then we help that become a much more strategic deal. I'm the biggest believer you'll ever find in product led growth, freemium viral adoption. If you can nail that for whatever category you're in, that is magic that makes SaaS so much more compelling than the prior era of on prem software.

So I should know this, but I guess I didn't. So the majority, if you don't exactly know, we don't do. But the majority of these 100k even million dollar deals is the sales team hunting existing accounts and thoughtfully getting, presumably getting better and better over time at migrating them to bigger and bigger deals.

Yeah. And the way it works is literally it's, it's idiosyncratic to every company in wherever they are in that journey. So sometimes it's going to a line of business and finding a couple of use cases and then rolling it to the CIO.

Sometimes it's going straight even 2021 for box at box to scale. That's still the predominant way the bigger deals. Right?

Yes.

Is silos and Landon expands and grows.

I don't know of another model that you can do in cloud. So and I'm sure somebody out there, maybe, maybe Palantir or something is doing a completely different approach. But like for all of us mere mortals like you just you get adoption, you build a, hopefully a great product, you get lines of business and it to sort of connect the dots at some point and then hopefully you drive bigger and bigger deals.

No, it's. Well, it's just interesting to me that I mean every, every business is hard. Right. It's not any easier being a CEO today than it was one year or two years pre ipo. Right. But the, but the brand gets bigger. Right. And I would think in 2021 you'll have CIOs that have deployed box now three or four times over their career. You would think they would come in and say look Aaron, I want to do a 500k deal. I want to do one because I've already done this. I don't want to spend five years, I'm ready to roll it out across my new gig. Right.

And that does happen. But last quarter we did, I don't know, 74 deals over $100,000. So it's. That's a fraction of those, but it's.

Still a small fraction.

And you still have so many more that are just, they've adopted more of the product and we have an incredible, you know, account executive that goes in and connects the dots. And that's the majority of our revenue still.

So are you. Do you think product LED growth is nothing new?

Well, I think there's ways that they've got. It's gotten way better at it. Like, you know, now we'll have a Slack channel that will automatically ping the sales rep that says, hey, there's some usage pattern here that you should go and connect the dots for the client. So the instrumentation we get, the automation we get is 10 times better. But if Martin Mikos was here, we'd be talking about MySQL you'll be here later. I mean, he was doing product led growth in 2003. So literally you download MySQL and then they sell you an enterprise version. So we haven't really changed that much in 18 years.

Collectively, the instrumentation maybe is a good summary.

That's. You get a lot more telemetry now.

Yeah. So this slide. Let's at least spend a couple minutes on this. This is so interesting. Right. So from 2020, if I have this right, to 2022, attach rates from 10%, 73% on multi product. Right.

What is this? I didn't, I didn't actually realize this is gonna be an investor conference. This is great. So, you know, okay, I need to alert my IR team about what I'm doing here.

I take this stuff and I find five things that are interesting for founders, not for Wall Street. I don't really care whether your stock price are up or down.

Right.

Yeah, I.

Today is not a great day for anybody's stock price. Yeah.

So we're all going long, aren't we?

I hope so.

We're all going long. But this is pretty. I mean, I know this is a way. You're also sharing this with Wall street, but this is, this is a big jump from 10 to 73%.

Yeah. So this is. So this is maybe the ultimate financial maybe proxy for what's going on. Which is. Which is we have a multi product platform. We've bundled it now that I would argue we took too long to do. So we had this sort of dynamic where you had to go to the customer and sell one product at a time that was, I mean, it's like, duh, that's obviously a bad idea at some point. So High calorie, highly inefficient sale. And then the customer gets this invoice of like nine line items and then they're just like, well, I don't want that line item for that many people. And then you're like, no, no. But you got that discount because you're buying the other stuff and it just gets confusing. So we did now these suite bundles and this stat is our 100k plus deals that have a suite or a multi product plan. And so that's been fortunately a big driver of our growth.

Well, that was interesting. What I didn't realize, a slightly older version of this I had on the SAST website had like three lines of all your different products. Right. And the new one has one. And that's because you learned selling three different products to the same base was not worth it, right?

Yeah, it's a lot, it's a lot better to just sell the whole thing.

Yeah, it's interesting. And then this, this, this last one on this one that I grabbed because this is telling 125% NRR with multi product, 90% with no add on. That's like not a small delta, is it?

No. And this is, I don't know that, I don't know the date in which this stat was shared, but so I.

Want to like beginning of the year. So let's consider it not inaccurate today.

Okay. Well, I'll just say whenever it was, whenever it was published, it was accurate. And so what I'll say is that if you can get customers to be more strategic, think about your platform more strategically. In our case, that can be approximated by customers that adopt multiple solutions, then they stay with you longer, they grow faster and that's the result we see in the net retention rate improvement.

Would you have done more of this multi product stuff, lean in more on security compliance and everything earlier if you'd realize where the world would be today?

Yeah, totally.

Yeah, I think that's everyone's lesson. And one last one. Maybe we have that one. We'll ignore this one just quickly. Interesting. Now that you're all this multi product, everyone's TAMs are crazy. Today everyone has a $70 billion TAM or 120 billion now that you've been through this. And actually the initial box category baby is mature. Right. The very early folder on the web is immature. But is there any ceiling in tam? Does you even worry about this or think about this? Do great teams find an infinity tam.

Yeah, we've always been believers in bottom up and top down. And I don't know that that's, you know, industry, an industry shared term or not. But we, we will look at top down, ie, where's all the, where are the dollars going in the categories we serve.

Yep.

And so this is an example of some of the categories that we, we serve. And now you can add E signature as an example. But, but so top down globally people are spending $55 billion a year on things about storing data and files and working with those files and securing those files. Okay. So that's, that's sort of like the TAM you're gonna go disrupt. And I'm sure half of this room has read Blue Ocean Strategy and the other half has read Innovators dilemma. I'm massive fans of both those books because the more you can be taking existing spend existing categories, but just changing the variables a bit so they favor your strategy. So what we do is we go to customers and we say what if you had one solution that did all of these things as opposed to buying individual products that do parts of those things? That's our disruption play. So that's your top down analysis and then your bottom up analysis. And this one, you know, you're now seeing in VC slides more and more. And so it's, it's obviously losing its credibility as you say. Okay, well there's a billion end users on the planet that need, that need some higher end content management collaboration functionality. And so you could do, do various ways of discounting and squinting your eyes and saying, okay, well if everybody paid X amount or if you got this percentage, you'd also be able to get to a very similar TAM number. So we've always tried to triangulate using those two metrics and I'm actually still a believer that, that that is a fair way to do it.

Cool. Well, Aaron, I got, I wanted more, I wanted to ask you, but it'll be the next Sessor Annual. Thanks for your time.

Sorry. Thank you. Bye. Sa.