Live SaaStr Annual: 3x - 5x: The New Gold Standard for Early Stage with Scale Venture Partners

Please welcome Scale Venture Partners operating partner Dale Chang, Honeycomb CEO and co-founder Christine Yen, and Dusty Robotics founder and CEO Tessa Lau.

Awesome. So great to see everybody here on this sunny Thursday afternoon in San Mateo. Really excited to be here, joined by a couple of great founders from our portfolio to talk about the growth, the growth within early-stage startups and how, what their secret was around top decile growth. Before we get to introductions, I just want to set the stage a little bit and talk about how this conversation all started. You know, a couple of months ago, you know, like a typical third day at a SaaS conference, I was doomscrolling through Twitter and came across this post. 1 million ARR growth is, to 3 million in 12 months is no longer top decile. It's now 1 to 5. And as you can imagine, what happened on Twitter was the typical Twitter arguments. First, there were people that just didn't believe. They were calling out doubt as to whether or not this was happening. Creative definitions of ARR, whether or not it was actually possible. There were the party poopers, the efficiency gurus that were saying, well, how efficient was this growth? Growth, which is actually fairly decent conversation to have today. But, you know, I eventually started thinking, hey, we actually have this data in Scale Studio, which is our benchmarking platform that we've built. And Eddie, who's an associate of mine, pulled this out, which actually looked at time series of growth rates of $1 million startups and how fast they were growing. And what you saw, what we saw was pretty striking. And it actually proved the point where it was, you know, companies were tripling year on year. in the early stages all the way up until 2020 and all the way into 2021. All of a sudden that just spiked up. Um, and so what we've done is we've asked a couple of founders from our portfolio to start talking about this. In our prep call, we thought about that there were 5 things that we would talk about today. One, which is tips for defining your north star. What is it, uh, that the company that you want to build? Um, the second one, which is focus. How do you actually drive focus to drive growth within the company? Company. Third is metrics and investing in metrics and automation early. So you actually have a visibility into how you are growing. Fourth is thinking about hiring reps early because never being behind the 8-ball on that. And then lastly, having conviction around what you're doing. So I thought that we would start with a little bit of— everyone loves an origin story. So maybe we start with the origin story of Honeycomb and how you define what your north star— what is your north star and how you got there.

Sure.

For anyone who isn't familiar, Honeycomb is an observability platform. We help engineering teams understand what their code is doing in production, especially when it's different from what they thought it should be doing. When Honeycomb started, observability wasn't a word that we used to describe this. There were logging tools, there were monitoring tools, and there were APM tools. And my co-founder and I— my co-founder came from an ops background. I came from a sort of developer background, both engineers, and we had grown up in this mindset of logs versus metrics, logs and monitoring. And we had the good fortune of being exposed to an internal tool at Facebook where a number of our assumptions of what these tools should look like were blown up. And as we, as we overcame our skepticism and started using it, we realized how much it transformed our workflows in a way that we could not have imagined thinking about logs, monitoring, and APM. And so when we started working on Honeycomb, we were able to sort of follow this certainty that if we tried to align ourselves with just what people were used to, the words that everyone had been using for decades by that point, we would be similarly bogged down by the baggage associated with these existing markets. And it was really something where we had to hold true to ourselves. And, and, you know, our early days were not full of, you know, 10x growth in ARR because we were laying a lot of the groundwork that would support that in later years. And a lot of the, that conviction was able to be expressed through thought leadership and evangelism and really shaking up our industry to say, hey, there's something new coming and we've got it. Come check it out.

Yeah. So hard return to a different type of company. Tessa.

Yeah.

Why don't you give us the origin story and the North Star as to where you are?

Sure.

Where you are today.

Sure.

So I'm also a technical founder. I got my PhD in computer science. And then more recently, about 10 years ago, I got into robotics. Started my first robotics startup. Honestly, it didn't go so well. And about 4 years ago, I decided I wanted to try again, start another company, but wasn't sure what I would do. And at the time, I was actually remodeling my house. It just finished, thank God. And that was 4 years ago, and I had construction people coming to my door, crawling around on their hands and knees, actually building, rebuilding the interior of my house. And I thought, well, you know, if that's still the way things are done in 2018, there's got to be an opportunity for robotics to make a difference. And so I bought a hard hat, I bought some steel-toed boots, I started walking around on construction sites just trying to understand this process. I didn't, I didn't know anything about construction. It was completely foreign to me. But I knew that with my background in robotics and in software automation, there had to be something that we could do to make the industry better. And the more I walked around on construction sites and talked to the people in the industry, I saw all of these manual workflows. They're done by people making phone calls in order to get work done, or people using measuring tape in order to dimension out the building. And I realized that all of these things are places where inefficiencies happen. And not so surprisingly, the construction industry is the second least automated industry on the planet. The only other industry less automated than construction is hunting. And so that tells you how bad this industry is. And yet we all live in products of the construction industry, right? You live in it, you work in it, you play in it. And so this industry touches every single person on the planet. And so I saw this huge opportunity to leverage my background in robotics and automation and create tools and, and better solutions for this industry.

Yeah. And how do you think then that helps drive kind of focus around accelerating growth. And you could be all things to all people, or you can figure out what it is that you really want to do. How does that— how does the definition of your North Star in the early days and holding true to that drive the focus that you need? And maybe even an example of that today.

So focus is super important. And I think the reason we're both here is because we focus. When you're building a hardware product, you by definition have to focus, because if you don't, it's going to take you forever to come to market. Hardware has really long lead times compared to SaaS, and you have to be very clear about what is it that you're building, and even more importantly, what is it that you're not building. So one really good example, at the beginning of Dusty, we were prototyping our first hardware product, building it out of parts that we bought off the shelf, hot gluing it together, creating something that kind of looked pretty janky, but it would would do the job. And every customer that we would show this to at that time, we were iterating really fast. Everyone wanted a slightly different thing. And that's where the North Star really helps because you have to ask the question, is this really going to help you increase automation and help make customers' lives better? How is it going to make customers' lives better? Is that feature a nice to have or are they going to pay for it? Right. And using that as the lens to decide which hardware features are in and which ones are out really helps focus the team. And that's what allows you to bring something to market faster, which allows you to start scaling faster.

So at home, I feel like with my kids, I'm the VP of no. Like, I can say no to them quite a bit. But, you know, maybe in the professional world, building robots and customers asking for 1,000 different things, and you're an early-stage company, you want to sign those customers. When have you had to say no and how has that helped you?

So in my early days of customer discovery at Dusty, I'll tell you that every— I talked to 10 customers and I got 12 different ideas of the robot they wanted us to build. And most of them were not feasible. They were like humanoid automation or something that was just impossible to build with today's technology. But even within the robot that we are currently building today, there's a lot of asks for features. People want sensors on it. People want different capabilities. And you have to be willing to actually say, "No, I'm not going to do that for you." And the reason I'm okay doing that is because you have to also have a mindset of abundance. You have to believe that the thing that you're building is going to be useful for enough people, that it's going to solve a big enough problem, that you don't need to say yes to all of these things. Everyone wants a drone paired with this robot. Everyone wants it to do, you know— we print with one color at a time. Everyone wants multiple colors, right? The fascinating thing is how do you solve those needs without actually defocusing your team to build hardware that's not necessary?

Right. So we had in our prep call talked about the spectrum between bootstrapped and soft-backed. And, you know, maybe one of the things that, you know, Christine, you and I talked about was really being resource constrained. Help drive focus. And so from an early days, super lean engineering team that took you a really long way. I mean, maybe talk about that and like how in the earliest days did you decide on where to place your bets on investment? And then how do you think about it today in terms of making investments as you end up having more capital?

Yeah.

Yeah. One of the things about our space is that it's extremely crowded. Anyone who has a software team, I can almost guarantee you're using either Honeycomb or one of Honeycomb's competitors, or multiple of them. And so the— there's a lot of noise constantly of what we should build to try and solve this problem, often because, oh, they used this one tool that one time that has a feature that they liked. When we started building out our early engineering team, we wanted to find folks who we didn't have to explain, we could, where we could operate in a shorthand almost in what, what was actually important, what was not. And this led to bringing together a much smaller group of seasoned engineers who had been around the block and, and sort of intuitively were able to understand the problem space. Obviously observability being what it is, our engineering team was also able to rely on our own product to constantly be, you know, very in tune with how our customers were actually using the product. And this led really to always being able to question, okay, we built this thing. What was its impact? Why did it matter? Do we want to do it? We want to solve this problem. Who would be impacted? And even thinking about saying no to customers, You know, the phrase I— in 2 years, 2 or so years into Honeycomb's existence, we came to this realization that we had, you know, in a world where everyone was asking for this faster horse, we had built a car, and no one was going to use it because everyone wanted the horse, and so we needed to paint the car to look like a horse. And these were things that if, you know, if we had brought in folks if we'd built out the team really large and brought in folks who weren't as familiar with our space, it would've been harder to turn the ship, harder to explain what that meant, harder to do the painting of the, of the car. But ensuring that our, our team was always hyperfocused on our North Star, hyperfocused on how to sift through the noise, I think really helped us accomplish, build a great product and continue building great product with a team that's almost always like— people are always surprised at how small we are based on what we've been able to build and deliver.

So there's a lot of early— oh, sorry.

Can I build on that too? Because I just thought about another anecdote. We also get the same comment. Like everyone's amazed at how far we've been able to come with such a small team. I'll give you a what not to do from my previous company. So my previous company, Savioke, another robotics startup, we were building robots for the hospitality industry. Think 3-foot-tall R2-D2 robots that would deliver room service. One of the mistakes that I made there as CTO was not investing enough in people and tools. So realize— so the thinking back then was, okay, we're trying to be economical, we're trying not to spend We need to focus on the things that are really important to us. And there are all these tools out there like Honeycomb and other developer tools that we could use, but no, we're trying to be cheap. We're trying to save money. You know, we need to do it ourselves. And in hindsight, that was absolutely the wrong decision. So what we're doing at Dusty is kind of the inverse, right? I realized that our people are the most valuable thing we've got. And their time is the most valuable thing that we have as a company. So why not spend the money to buy the tools that actually make those people more productive? And it's totally worth it because that allows you to focus— focus on the things that actually matter to your business rather than inventing something that already exists.

There's a funny theme here between trusting in your intuition and the North Star and then grounding it in data and— or your tooling or outsourcing it. And it's, you know, that difference maybe is one that falls on the intuition side, but it's— That's something that is so easy to get wrong, and thinking about our customers, a challenge can often be trying to outsource too much to a tool or trying to build your product to promise too much magic, when instead it's the, you know, tools should augment your people. Tools should make your people more— better at being the intuitive, clever, context-full humans that they are.

Totally.

So actually, I'm going to skip around because I think that you brought up an interesting point around conviction. And someone had once said to me that there's a difference between thinking and knowing. And thinking, like, you're running— oftentimes, you have a lot of data at your fingertips. You're running analysis. And you think, based upon that, you're led to think that you have the right answer. Other times, something's just deeply ingrained in that you, that you know that, that like your gut is telling you this is the right way to do something. Yeah, I would love to hear just like when do you listen to which voice? It's like that's at the, at the heart of it I think is like how do you, how do you decide between like in absence of data, how do you make a decision as to this is the right direction to go? You gave the hardware example before. You're locking yourselves in for 2 years. How did you know that that was the right way to—

you never know. Yeah, okay, but you have to guess. Um, so I mean, the problem is you never have enough data, right? You always want more data. You want to make data-driven decisions. You want it to be very quantitative, but you never have the luxury of collecting enough data because by the time you have it, it's too late. And you know, the challenge of this environment is that you have to constantly grow. You have to constantly act with not enough without enough data, and that's where the gut comes in. That's where the instinct comes in. One example of where we made a decision recently without enough data is which customer do we go after? We've segmented our market. So we sell to construction companies. If you're familiar with construction, you know there's a general contractor and they hire subcontractors to actually do the work. At various times in our lifecycle, we have sold to GCs and to subs, right? And we've kind of, switched around which ones we go after at different times in our history. And just recently we decided to shift that again because sales were struggling with one particular customer for a particular reason. And so you have to— but when you make that decision, right, of, for example, which, which is your customer segment, what's, what's the ideal customer profile you're going after, you're never going to have enough data. We did— we We looked through our existing revenue and which customers were they coming from, and you compile a spreadsheet and you try to analyze the heck out of it and find standard deviation certainty, but you don't get that because as an early-stage startup, by definition, you don't have enough data. And so it really comes down to what is your intuition and what does your heart tell you about what is the right thing for your business. And what I think happens is that if you don't make that decision, you don't get the focus. And if you don't make that decision, no decision is a decision. And if you don't make that decision, you end up selling to everyone, and then you don't have the focus. You build a product that doesn't do the perfect thing for any one particular segment, and you end up not winning.

Right.

And maybe it's unfair to ask an observability company around making decisions without data, because that's really what the software is meant to try to do. But, You know, maybe that helps with developing and being innovative with software, but like internally, like—

It's true. You know, when I think about the different— the times I have had to weigh what I knew versus what I thought, the know is always the things that you have experienced yourself, right? This is why there's so much advice about, oh, go do the thing that you, like, aren't sure about, and then develop a hypothesis, and then harden it, and then hire someone in to do it. I think that you're never going to know, but this is where, while you're in think mode, gathering as much data as you can to then, to then, you know, be able to point back at a paper trail, learning to hear what that gut feeling is. Like, even if you're doing something, you know, I was very new to sales when we started, when I started trying to do sales with Honeycomb. And the developing the sense, going from think to know that, like, there's something here, it takes time. There's no shortcut. I will say sometimes it's true you will never have enough data. Sometimes you have too much data.

[Speaker:JASON M. LEMKIN] Yes.

[Speaker:JESSICA YU] And you need something to help you cut through it. Is what— where I always go back is who has the— when you're looking at the data, what is the context that isn't captured in the data? What is the lens you're trying to view it through? Is that the right lens? And if so, and no one else has that lens, then how do you capture that and make it repeatable so that someone else can interpret, interpret something or pick up on the same signals that you are?

The other thing about data is that I'm realizing that it's not so much about quantitative metrics. It's more about the stories that you tell with the data. And to be honest, it's probably the case for it, for all of us and any good salesperson to be able to tell any story you want based on whatever data you've got. Right. So I think the important thing for our business, and hopefully this is true for other people as well, is in order to get the company focused on the right path, what you think is the right path. You look at whatever data you have, you make the decision about which— what are you doing with it, what is your company strategy based on that data, and you go after just that. And then you retrofit the narrative on top of the data and on top of your strategy. But, you know, I— if there's someone out there who's like, do it in the opposite order, I'm very impressed with that. But that's what's working for us.

So we talk— one of the things that we've talked a lot about here is data. And, you know, I'd love to get your take on just looking at advice that you would give around, you know, defining your metrics early on. So, you know, North Star metrics. So you define what the company is that you want to be, the focus that you have, you know, and I think back to some of the days when we were working, you know, one of our portfolio companies, DocuSign, which was, hey, listen, we're measuring a whole bunch of things, but the one thing that matters to us is the number of digital envelopes that are being sent. How many signatures are out there? Because that indicates how much we are driving the adoption of e-signature. I would love to hear, are there any type of metrics that y'all have?

I'm gonna let you go first 'cause you're nodding excitedly and I will feed in my more nuance.

I was hoping I could think while you were talking. It's okay, I'll go. All right, so robots are these little data collection machines. And we have the luxury of building a product that is inherently instrumented. Because every moment that they're operational, they're generating huge amounts of information. The challenge is always which bits of that, or what, how do you take that deluge of data and turn it into actionable insights and metrics? That you can use to track. One of the challenges that we're facing right now is of all of the information that our robots are collecting while they're operational on construction sites, what is most impactful to our customers in order to figure out how do we get that DocuSign-like metric of utility, right? Is it time spent? Is it amount printed? Is it numbers of projects worked on, right? And these are questions that we're asking But we wouldn't be able to ask them if we hadn't invested earlier on in saying this is going to be important. And the challenge with instrumentation is that you have to build it without knowing how it's going to be used. You have to have a belief that this is going to be useful for you eventually in the future. Because I have a team, and if we have, like, you know, one robot operating, they're like, why do I need to instrument it? We know what it's doing. And then that one robot becomes 5 robots, that becomes 50 robots, it's going to become 100 robots. 100 robots and 200 robots soon, right? At that point, you wish you had the automation, but that's too late to start investing in it. And so, you know, having that belief early that you're going to use this data and you have to build it before you start to scale is really important.

I'm going to have two answers to this. One from an engineering perspective, because it's something we see a lot in our customers and it's been relevant for us ourselves. And then one from a more business perspective. On the engineering side, again, this goes back to that tiny engineering team I mentioned. We instrumented the crap out of all of our software early on because we knew that we were going to be hammered with sort of all of this unpredictable traffic from our customers and wanted to be able to respond quickly and be able to sift apart kind of what unexpected thing happened this day. So this is— that piece, you know, metrics and automation, especially in your software, the piece the part of your organization that is going to be sort of weathering the most, weather, the most chaos, that, that is the core of the Honeycomb hypothesis and has been really good for us as we've grown and our engineering team has scaled and we've been able to talk about things like, you know, being able to use engineering concepts like a service level objective in company OKRs. Because we can believe in that data being mature enough that we can really understand that at a company level. On the business side, honestly, when I think about automation and data that helps us scale this, I think about sort of these like early customer conversations and customer personas. Right. And this also comes from, comes out of the focus conversation. When we first started talking to customers in our first couple of years, there was a lot of, oh, hey, I know you. I sort of know your problems. Let's figure out a bespoke rollout plan, and I'll totally customize my pitch based on what I know about you. And I wish we had been more, you know, if I could go back time and, like, set up the form fields for the discovery notes, like, holy cow, that'd be so much more valuable now because all of, all of the things that we now know actually impact how someone adopts the product or the biases that they carry into a POC. We thought about them early on, but we didn't know them because we didn't know them. We hadn't had time to sort of settle into them from a sales perspective. Perspective. Those are my two.

So then controversial today is like hiring reps, just hire more sales reps. You know, one of the things that we looked across our portfolio is this concept of tooth to tail, which is the number of sales reps as a percentage of your total headcount. And like the best performing companies were somewhere between 15 and 20%. So for every, you know, you know, every 5 people that you hire, 1 should be a salesperson is what the data was saying. But, you know, I'd love to kind of hear like, you know, from your perspective. I mean, it depends on company too. And like, you're selling Motion.

I will, I will say there is a period where I think at every board meeting, our Scale board member asked us why we weren't hiring more salespeople. Yeah. And we have almost historically always been under, under-assigned. We have, because we are a technical product and we knew fairly early on that it was gonna be a Challenger sale, we had a very specific persona type for the sales rep that we felt could be successful at Honeycomb. And we knew that the sort of organization we wanted to build was not one that would feel head to tail or else we wouldn't be able to attract the reps that we would need to make us successful. And so I think it came from— and our ability to push back came from being able to describe why and who we were going after and why volume play, I think, was not right for us. I'm curious about your world.

So we're— according to that metric, we're failing. But we're trying to do better. We are actually hiring sales, if anyone has any recommendations. So I think the way I would characterize how we think about building up our sales team is that we've had a number of hypotheses over the past couple years about how this go-to-market motion would work. And in particular, where would sales come from, what kind of profile would make a good salesperson, a good salesperson, when is the appropriate time to hire that kind of person into the company. We're actively growing now. We're following my latest hypothesis, which is that, okay, I need a couple enterprise AEs in order to actually sell this. But we've sold at different price points into the industry and, you know, gotten to this point with essentially a very junior team, a small number of very junior people, and demonstrated that there's a lot of a lot of success to be had that way. And now we're moving upmarket, moving to the larger enterprise deals. And I've learned recently that you need different reps to actually do that kind of growth. And so as we've been learning this, I've been constantly like rebalancing what do I think is the optimal size and composition of a sales team. One of the things I'm really glad we didn't do is, is overhire early because before you know what you really you really need. And so I would say that this advice, I would put it hand in hand with understand your go-to-market motion and what kind of skill sets do you really need in order to sell at your price point to your customers.

Awesome. One last— sorry, go.

I'm just sorry if we, uh, we're countering your slide.

No, that's okay. Um, one last question for me, and then if you all open it, if there are any questions from the audience, we'd be happy to take them. Um, not that younger you, less experienced you, let's put it that way. What advice would you give yourself?

Uh, the flip answer is therapy and coaching are great. Uh, you know, invest in yourself. Um, more, more seriously, uh, I like your number 5 on this slide. Um, we spent a lot of time speaking to our audience in the early days. Even before we knew what our, you know, marketing message was, before we had our sales motion in place, we found ways to hit a nerve. We, and we, we, you know, developed early relationships that where even if they didn't convert, they were fruitful. They taught us what words did resonate, and many of them did convert later on. And there's a, there's a lot of pressure, especially now, to do things that immediately result in 5x, 10x growth. And sometimes the things that you will drive your company the fastest in later years is laying the groundwork and really building up, you know, in our case, a great engineering brand, and often being known for being credible in a certain domain, which doesn't play well on a balance sheet or investor deck, but was so critical for our growth today. Yeah.

I definitely second everything you said, including therapy and coaching. Highly recommend. Um, let me talk about conviction. So we have, we have a very high bar for hiring, and throughout the entire life cycle of our company, we've always gotten pressure to hire, hire, hire, grow, grow, grow, hire faster. And I'll be very honest, the hiring environment pre-2022 was crap. It was really, really hard to find people, and it's still hard to find people, but it's getting easier, thank goodness. And so we've made a couple mishires that came about because I was bypassing my conviction in, in holding to a really high bar. And inevitably those result in disaster, and they set you back, and you lose time that you would have otherwise not wasted if you hadn't made that mistake. So I would apply have conviction to If I were to give one piece of advice to myself, it's believe in, believe in what you believe to be true about the people that you want to hire and don't let anyone tell you otherwise.

Great. Are there any questions in the audience? Seeing none on a Thursday afternoon on the third day of a conference. So I just want to thank you all very much. Thank you, Christine. And thank you, Tessa, for joining me on the stage today. Enjoy the rest of your Thursday.