Ep. 183 - What $100M SaaS Companies Do Differently

Guest: David Karandish, Founder & CEO of Capacity  

--  

Most SaaS companies don’t fail because the technology is bad. They fail because they build point solutions, chase the wrong markets, and struggle to turn AI into real, scalable value.

In this episode, David Karandish, Founder & CEO of Capacity, joins host Ken Lempit to share how his team scaled past 20,000 customers and toward $100M+ in revenue by evolving from an AI point solution into a full SaaS platform for support and contact centers.

David breaks down the pivots behind that growth, why mid-market SaaS often stalls, and how the compound startup model is reshaping modern SaaS — not by doing more, but by integrating smarter.

Key takeaways from this episode:

• Why many AI SaaS products fail before reaching enterprise scale
• The difference between “salad” vs. “brownie” AI projects
• How platform consolidation creates GTM and pricing leverage
• Why GTM motion must align with deal size
• How integration becomes the true SaaS moat

If you’re a B2B SaaS founder, CRO, or CMO navigating AI adoption, platform strategy, or the leap from mid-market to enterprise, this episode offers a grounded playbook for building durable SaaS growth—without the hype.

---

Not Getting Enough Demos?

Your messaging could be turning buyers away before you even get a chance to pitch.

🔗 Get a Free Messaging & Conversion Review

We’ll analyze your website and content through the eyes of your buyers to uncover what’s stopping them from booking a demo. Then, we’ll give you a personalized report with practical recommendations to help you turn more visitors into sales conversations.

And the best part?

💡 It’s completely free.

No commitments, no pressure—just actionable advice to help you book more demos.

Your next demo is just a click away—claim your free review now.

Transcript

Ken Lempit: Welcome to SaaS backwards, a podcast that helps SaaS CEOs and go to market leaders accelerate growth and enhance profitability. Our guest today is David Karandish, founder and CEO at Capacity, an AI powered support automation platform that reduces support costs and improves knowledge access so teams can do their best work.

Hey, before we jump into the episode, David, please tell us a little bit about yourself and your company. 

David Karandish: Yeah, so Ken, thanks for having me here today. David Karandish, CEO and co-founder of Capacity. I've been in the tech space my whole career, building technology companies, websites, endeavors, entrepreneurial outcomes since before I could drive. Started this company Capacity back in 2017.

Our mission, as you mentioned, is we wanna help teams do their best work by using AI to automate support that could be support for your team, whether that's a HR or IT, or it could be support for your customers in your, contact center, your help desk, anywhere where you're answering questions on behalf of your customers.

Over the last few years, we've grown from an idea, about this crazy AI world to now we've got over 20,000 customers. We're a global company and excited to take the world by storm. 

Ken Lempit: That's quite a lot of success already. It's really impressive. And, uh, we appreciate having you on the episode.

Thanks for having me. So, but let's go, let's go all the way to the beginning of the story, the beginning of the story here. So, take us, take us back, after you exited answers.com and you hit what you called an existential crisis. And what happened during that period that shaped this idea, this learning obsessed idea that became Capacity?

David Karandish: Yeah. So after selling my last company and exiting there, I had my long laundry list of different things I wanted to do. And some of them were similar stuff I'd done before. Some of 'em were brand new, some were in the tech space, some were out of the tech space. And I heard an interview from Jeff Bezos.

Where he talked about how when he was first getting started, they wanted to know like, dude, why did you start Amazon? And he was like, well, he looked at the space and said, if e-commerce succeeds and he starts, Amazon, thinks he's got a real shot at, you know, big company. If e-commerce fails and he starts Amazon, he will have learned a lot in a, in a space that he thought was gonna grow.

If e-commerce doesn't take off and he never starts Amazon, it's kinda the null case. But if his e-commerce thing took off like he thought it would, and he didn't start Amazon, he was gonna regret it the rest of his life. And so I actually, I can't remember if I heard it or I read it, but I, I remember consuming this from, from Jeff Bezos doing an interview.

And the big thing that struck me is this was the exact same dilemma I was going through between starting Capacity, not starting Capacity, AI succeeding, and AI failing. And it's very easy today to look back and say, oh, it was a, it was a fait accompli in terms of clearly AI was gonna take off, but when it was 2016 when we were just dating the idea.

People were talking about Blockchain and AR and VR and all sorts of other technologies, that would be the next, the next big thing. And AI has come in and dwarfed them all. And so I knew if I didn't start Capacity and this AI market takes off, like I, I thought it would, I would regret it the rest of my life.

Ken Lempit: And so, I, I guess you have to give us a little visibility into where the idea came from. you, you knew you had an interest in AI and applying it, but you know, what's the, what's the founding idea around Capacity and how did it come to you? 

David Karandish: So I had a couple different threads that were going, both experiences, I had trends I saw and then this just rock in my shoe.

I couldn't let go of. So I had grown my previous company answers. At our peak, I think we had over 650 people all across the globe, multiple time zones. And I felt the frustration of what it's like to get basic things done internally, communicating internally, figuring out your benefits. You need a password reset.

Like it just got, it got to be so painful to do these things that aren't even, core functions, they're just the underlying items that keep you moving as a team member. And so I, I've always had this burning desire to see people use their gifts and their talents to go kind of as Steve Jobs put a, put a dent in the universe.

And I wanted to figure out a way to, that we could help express that. And I know that logging into a bunch of key systems and sending a bunch of emails back and forth and not being able to access what you needed to like, that seemed pretty painful. So the original, original idea was, I like the SaaS business model.

I'd been doing Q and A for years from our time at answers.com, and I saw AI as a way to take the question and answer background I had but bring it internally for teams so that people could focus on the things that people are good at. They love to do that they're excited about. Like I said, they're God-given talents and then people could leave the things out that frustrate them all the back and forth, all the administrivia.

Now, when we got into it, we learned pretty quickly that our initial idea, our initial thesis was right, that there is a huge problem there. But to get customers to do the unnatural act of opening up their wallets and paying us for this service is a different story. Specifically when we started, we were building the initial company for the middle market and the challenges a lot of mid-market companies don't have.

A ton of budgets for HR and IT. So we were running up against budget constraints at the mid-market, and then we were also running up against a lot of enterprise readiness on the larger side of the market. So we'd go in and talk with these big organizations at that time, and we didn't have role-based access controls and SOC 2 and all the,

three and four and five letter acronyms, people need to be enterprise friendly. And so we had this, this great idea of AI coming in and transforming the workplace, helping teams do their best work. And we found our very first twinkle of a product market fit. So the, the industry we first focused on was in the mortgage space.

We still have a lot of great mortgage clients today. But the thing that was interesting there is they had both an internal use case problem. So people looking up guidelines and loan officers trying to figure out if a loan's gonna close on time, but they also had customer facing problems or borrower facing problems.

So, you're a first time home buyer and you want to know. You know, how does an adjustable rate mortgage work? Or today should I do a 50 year mortgage? I have thoughts on that. We can talk about that later. But they have all these, they have all these questions. So we kind of looked at it and we're like, well, what if we could be one platform that both helps your internal team and it helps your customers?

And so that was kind of the very first pivot that we went through from being just a single internal focused HR IT to going after this vertical where we cover kind of both sides of the house. Now, Ken, as we got into it, we ran into a classic market cycle where the mortgage industry is a fantastic industry to focus on when rates are free.

When rates became decidedly not free, 

Ken Lempit: when they're writing, they're, they're happy. 

David Karandish: Correct? Correct. When rates became decidedly not free, we had, we had customers that really were on the phone with them and they're getting fired. In the middle of our calls. And so we're like uh, okay, we're gonna need to do another, probably do, do another pivot.

Again, not a 180, but we're gonna need to, we're gonna need to adjust, adjust the rudder here a little bit. Specifically we're like, okay, where's this technology gonna have the biggest impact? And we looked at the contact center and we said, okay, this could be the next kind of place for us to focus because.

Contact centers are both a horizontal, so they cover across multiple categories, but they're also vertical, so they have specific technologies they use and integrations and the like. So we started sitting down and meeting with these contact center people and you know, I, I met with hundreds of them and one of the things that just kept coming up is people are like, David, we are so tired of having to duct tape together these various solutions.

And so, they're like, we, we don't wanna go to one vendor for our help desk, another vendor for our live chat, one vendor for our workflows and a totally different vendor for our call scoring. So we looked at this and said, what if we could go put what used to be a bunch of these disparate areas together into a platform and go bring that to market?

So, first thing we did is we mapped out the 24 steps to the customer experience journey, and we tried to come up with a name for this thing and by the way, for the founders out there who are trying to do this, this kind of pivot, expect to not be well received when you first go down this path.

My team thought I was crazy uh, that we were gonna go and and do this came across a guy by the name of Parker Conrad. Parker was the CEO of Zenefits I think CEO and co-founder of Zenefits. He's now the CEO and co-founder of Rippling, and he came up with this term called Compound Startup. The basic idea is that 2025, there's a point solution for just about everything in this SaaS space.

Pick your, pick your category or vertical. The next unicorns and maybe even the next 'decacorns' are gonna be the companies that can go find the individual point solutions. Combine them together, lower the cost of the SKU. Maximize the price of the bundle, and then, and this is very crucial, go be integration as the product.

And so we said, okay, we are gonna run with this strategy, this compound startup playbook. He's applying it to the HR space, we're applying it more to the customer support contact center space. So that's what we've been up to over the last couple years. And it's been great. 

Ken Lempit: you said that your team was a little resistant to second pivot.

How about the board? How about your investors? What did they think of your plan to move forward? 

David Karandish: Exact same. Partly because you are enacting a compound startup strategy, which by the way, I don't recommend for everyone, it's not the right move for everyone. If you're going after that strategy, you're gonna need to cover a lot of surface area.

And so in order to cover a lot of surface area it means you have to get good at a couple of meta skills. One, you have to move your team from being great problem describers and problem solvers to being phenomenal problem selectors. I could pick any one product within our product line. I could go spend the next year just focused on that to figure out how to select problems across the surface area, what we cover.

then second, you have to figure out what that build partner portion of the strategy looks like. So we've got some parts of our platform that we built out from the ground up. We've got some parts we partner in. So like we've got a payments portion of the platform. We've got a partnership with Stripe for that.

I don't need to build the payments processor in 2025. And then we've got some parts of the platform where we've gone out and found a company that has gone really deep in that area. We acquire them. We bring their founders on almost, in almost every case, the found founding team has stayed. And then we go run for the upside.

We do the integrating and the duct taping. Customers don't have to do that. We bring that out to market. We do a bunch of cross-sell upsells, and you got a lot of good stuff that happens on the other end of that. So that's, those are some of the, the, the meta skills you gotta work on for this, for this playbook.

It got to the point to where the board started to see the progress, even in a way, even before the team did. And then once the executives started to see it, then the next level of leadership started to see it. And now we got everybody singing from the same hymnal. 

Ken Lempit: That's pretty exciting stuff.

Aligning, you know, investors, execs, and everybody else probably key to success and such a complex go to market. Right. There's a lot of moving parts now. 

David Karandish: Yeah. The other thing is. I had to get up and say what we are doing, so many times to the point I feel I felt like, man, am I just being repetitive up here?

It's like even, even just recently we did our company huddle. We brought people in from all over the globe, descended upon St. Louis and I got up and it was kind of a David's greatest hits to slides and a bunch of slides. We presented to a lot of people over a lot of time and still there people are like.

It finally clicked this time, David. I find I, I get what we're trying to do. I understand how these pieces are starting to fit together. And so, another piece of advice I'd give to any founder who's running this compound startup playbook is you are gonna have to over communicate. And when you think you're over communicating, you're probably halfway there.

Ken Lempit: Where can people find out about the playbook?

David Karandish: TechCrunch did a nice article on this. He's done a bunch of, a bunch of interviews on it.

Ken Lempit: So, we didn't talk about this, but I, I've gotta ask you tell me about being in the middle of the country as a tech-based AI startup. We're not really so much a startup anymore, but really scale up, right? So what does that, what does that look like? What are the benefits of being in St. Louis middle of the country versus being coastal?

where most people associate startups as being. 

David Karandish: So I'm gonna answer this question pre and post COVID. So pre COVID, the big advantage we had being in St. Louis is we'd have longer tenure people. People get picked off a lot less to other competitors. And we had everybody in one place with a lower cost of operating compared to living in Silicon Valley or New York or Denver, even Austin, Austin's gotten real expensive, right? 

Ken Lempit: Totally. 

David Karandish: So, so there's some, there's some structural advantages from a cost basis. My last company, we started in St. Louis and we opened a Silicon Valley office, and I like to say. We got 10% more talent in our Silicon Valley office and 90% more drama.

And so I would take 90% of the talent at 10% of the drama any day of week. So pre, pre COVID it was very much getting people in again, a big fish, small pond environment all in the same room. Lower cost of operating, keep people for high tenure. After COVID, when everyone started working from home, all of a sudden a lot of big companies started hiring people from places like St. Louis. 'cause if you're in, if you're at Amazon and you wanna hire some great talent, man, you'd love to pay them St. Louis wages instead of Seattle wages. And so we realized pretty quickly that. This was going to become a global workforce where we're gonna end up with people all across the world. And that does, that doesn't just mean the US, that means we've got folks in Honduras and South Africa and Manila and London and

Barcelona, all, all across the globe, wherever we can find the best talent. And so that involved us doing a big shift to a lot more asynchronous communication, which a lot of people did going into COVID and then they did a big push back to office. We still have an office here. We still have people that come in.

But most of our talent is now distributed. And so I would say at this point, anywhere where you geographically restrict yourself, you are probably missing out on either high quality, low cost, or the combination of the two. High quality and low cost talent. 

Ken Lempit: So the location is much less important for young companies.

Makes a lot of sense. Now, certainly our business, same pattern, you know, six years ago we had one office, everybody came in. No matter how grueling the commute, now we have no office. Everybody works from home and they're much happier. So, that's awesome. I, I wanna shift gears a little bit 'cause you used some analogies.

I, I love some of the analogies that you bring into these conversations. And when we were talking, you, you started talking about this idea of salad projects. Versus Brownie projects, and I happen to love salads and brownies. So I was with you. But this was in the context of understanding which AI bets a, a software company should make.

And I'd love you to walk us through the analogy and, help other SaaS leaders understand which AI initiatives are really valuable from the ones where maybe it's not as good a fit. 

David Karandish: So I'm not a great cook, but I have a couple dishes that I'm, I'm known for at my house. And one of them is I make a homemade Caesar style salad.

I actually call it the Cyrus salad 'cause my dad's Persian, or my dad's from Iran. So, you know, you gotta, kinda pay a little homash. But the salad is, it's, it's phenomenal. Ken, if you ever come to St. Louis, come over to my house, I'll make you the salad. It's got the anchovies, it's got the nice good olive oil I make, I can make my own croutons.

I don't eat eggs, so I have this like stone ground, mustard replacement that's just as good. Fresh garlic, black, all the things. It's such a good salad that if you removed one ingredient. It's still gonna be pretty good. You get like 80 or 90% of the value outta that salad. Maybe with the exception of the lettuce.

You gotta have lettuce to have a salad. But it's one of those things where like there's so many different ingredients, so good that even if you deliver an 80 or 90% solution, you're gonna get 80 or 90% of the value out of it. Conversely, I've got four kids at home and they like to make brownies. If they miss an ingredient in those brownies, it's not 80% is good or 90% is good.

Those brownies are in the trash. It's literally inedible. So the reason I tell this ridiculous story is this is the story of AI projects. They are AI projects that are very much salad projects where you don't have to get all the way down to the last mile to still deliver a lot of great value. And they're AI projects that are very much Brownie projects.

So when I got into support, part of the reason why I made the bet that automating support would be a good use case for AI is that even if you could automate 30 or 40 or 50% early on, as long as you have a good escalation path. It's a great experience to both cut costs, create a better UX, and you don't have to automate every single question on day one.

Conversely, there are companies out there trying to replace truck drivers with self-driving technology, and by the way, this is coming. This doesn't mean it's not gonna happen. You can't have a truck that drives really well 90% of the time when you're hauling an 18 wheeler down Highway 40 here.

So I look at this and it's like replacing truck drivers with autonomous driving technology is very much a brownie problem. Automating support is very much a solid problem. Now I'm not putting a value judgment on one or the other. It's just that the salad projects are gonna be the ones that you can get off the ground up and running pretty easily, and the branding projects are gonna take a while.

I saw an interview with Andre Carpathy, who was one of the early , AI pioneers at Tesla. He was talking about how he got in his first self-driving car in 2012. He's like, 10 years from now, we're all gonna be in self-driving cars. All nobody's have a steering wheel, right? And you go out and you look and like I took a Waymo and I was in Austin last week, said, yes, this technology is here, but it is not evenly distributed and it's gonna take a while.

It is very much a brownie problem to have steering wheel less. Self-driving cars as the primary mode of of transportation. So when, when you are figuring out as a founder what AI solution you're bringing to market, you need to be able to distinguish between is this a salad project or is this a brownie project?

Ken Lempit: It makes a lot of sense. And you know, for those of us who've experienced, you know, AI bots in a support environment, they can be better than having to do it with a person. Right? They're, they're always there and the answers they do have are usually correct. So I, I think that's. This is really useful analogy.

And also I think it's good for the people that work for these companies, right? They're the easy questions they don't have to deal with, right? So you take a lot off their plate kind of early on. That must make adoption a little simpler, a little, a little more satisfactory as well. 

You touched on this, but I want to go a little deeper into the customer experience journey that's sort of guiding your product development. And specifically like what was the thinking that made you realize that the unicorn or decacorn would, would come out of a platform like you're building, you know, what was like, what were the signs of that?

So like as we're building. Our companies, you know, what should we be looking for that says, hey, you know, maybe the one or two things we're doing aren't enough to get us where we might want to go with our investors and, you know, our shareholders. 

David Karandish: Yeah. So one, one clue that you can double click on is what is happening and what should happen when your product fails.

So in our case, when we started, if you asked our AI 100 questions, we could respond to 55 of them. 10 of those would get thumbed down in the feedback loop. And so my joke was 45% of the time it worked every time. So the question was, well, David, what do you do with the other 45? That couldn't answer in 10, that the, the user said we we're not great answers.

And initially we were trying to go escalate up to these external help desks and realize that you have to be able to close the loop. So you have to be able to take the unanswered questions, get them somewhere, and feed those right back into the AI if you want it to be able to learn and grow. And so what we ended up doing is we said, you know, we're building this AI company, this now, you know, term virtual agents, but we have to have an integrated help desk along the way.

Because I don't care if it's, you know, now we're at somewhere between 93 to 94% of questions get answered automatically by the virtual agent. You still need something, someplace to send the other 6 or 7%. And so figuring out that, that those two shouldn't be two separate products, but they should be intricately linked.

That was the first spark to take us down this platform path. 

Ken Lempit: So it wasn't like, Hey, I'm gonna go build a platform. It was, Hey, I'm not doing a good enough job. 

David Karandish: It was like, oh no, this technology works half the time. What in the world are we gonna do the other half the time? So that was the first thing.

And then the second thing, like I said we knew that our clients would want the technology to be omnichannel. Not just be web, but get an email on it. Get into voice and get into SMS. So we knew that, but we didn't recognize how many other steps there are of the customer experience journey that these companies keep trying to tie together.

And so that was the unarticulated need that we uncovered when we started really diving in, particularly with our contact center and customers. 

Ken Lempit: I wanna move on to the last thing we prepped on, which was you know, the company's grown pretty large, 20,000 customers. That's pretty remarkable.

And it's eight years old. Is that right? Yeah. Do you talk about your 

David Karandish: revenue? We'll be crossing a hundred million soon. 

Ken Lempit: I'll just say that. So that, that makes you, you know, really one in some small fraction, right? One in 500 companies at that age hitting that revenue. 

David Karandish: I actually did the math on this for a different research thing I was looking at.

For every company that gets started in the United States. If you took 25,000 people to start a company, put 'em all in a stadium, and you said, how many of you entrepreneurs are going to reach a hundred million of revenue, you'll leave one person in the stadium out of 25,000? 

Ken Lempit: Pretty terrifying odds. And that goes, I guess the, the numbers that you've achieved go beyond what I've read.

Like Kyle Poer just put out a piece on his growth on Hinge Substack about,

you know, the, the smaller companies and the, the grueling, the grueling future they face. So, Wow! Good for you. Thank you. That's really awesome. But the context for this really was that you have a mixture of PLG and enterprise and, when we prepped on this, we talked about the different needs for speed of adoption versus the deep integration. And you mentioned some of the security standards, sign-on standards data protection, but there's also other integrations that I'm sure, you know, drive a lot of the enterprise motion.

So I wanna talk a little bit about the young company versus enterprise, like the small solution versus enterprise solution. But then also I want to kind of land on the go to market difference. So why don't, why don't we start a little bit about how, like in product management and marketing, you know, you're managing the difference between the PLG motion and the enterprise sale.

David Karandish: Yeah, so we think about it in terms of how we go to market in two, two buckets. On the enterprise side, we'll go direct where we have code of carrying salespeople out. Doing LinkedIn campaigns, dialing for dollars, sending emails, going up to events, pushing for referrals. We work through channels as well on the enterprise side.

So in the contact center space, we work with a lot of the TSPs and TSDs technology solutions, brokerages and distributors who help get us connected with consultants. And then we will sell through. Other third party software channel providers, whether that's other contact center customers or through vertical, specific systems of record that we integrate with.

So that's on the direct side. Again, call it mid-market up to enterprise. When we go mid-market down to to the to s or small businesses. In that case, we're either doing inbound, SEO PLG, get in. Try something out, or we are still selling through channel where we are just an add-on on top of the SKUs that those customers already already sell, and we're deeply integrated for that.

You can't expect to do a lot of PLG on the Tip Top of the enterprise, and you can't expect to be able to pay a direct sales force for the, you know, lower priced products. So you have to figure out how to align your, your go-to market motions with the price point of the product you sell. There's a, there's a blog, I think it's Christoph Janz who put this out about what, what animal are you hunting? And so, again, no animals are actually hurt in making of this podcast today, but he's got a whole little graphic on how, you know, you can hunt whales and you can go, you wanna build a hundred million dollar business, you can get a hundred million dollar whales you can go get I think it's a thousand 100K elephants.

You can go get ten thousand 10K a year deer. You can go get a hundred thousand, $1000 rabbits. $100, I think it's mice after that, something like that. And it's kind of a silly thing, but you look at it and you're like, okay, in the real world you would go hunt a whale.

Not that I recommend hunting whales. You would go hunt a whale very differently than you would go hunt a mouse. You're not gonna leave traps for the whale with a piece of cheese on it. You're gonna need a boat for the whale, right? You'd also probably be overkill to bring a harpoon to go hunt to mouse, right?

And you could go extend that analogy to everything in between. And so his whole point is, a lot of times entrepreneurs, they get this great idea and they're gonna launch this software, but they don't know which animal they're hunting. Or they think they're hunting one animal, but they're aligning their go to market for totally different type of animal.

And so figuring out how to get that aligned is, is very, very important. And you can hunt multiple animals, but you, that means you have to have multiple go to market motions at that same time. 

Ken Lempit: Yeah. First of all, I love the analogies and I might have to rip you off her Who, who was the guy you mentioned who actually, I believe this was Christoph Janz

So, so we'll have to, we'll have to credit him, but it's a great analogy. And maybe a great place to land the episode. The number one problem we see with companies where they're not achieving their goals has to do with their sales, sales process, the messaging that marketing is putting out there to help drive opportunity.

And a lot of it is. You know, are you saying the right thing to the right people and bringing the right tools to try and make the sale? So that's, I'm not 

David Karandish: gonna catch any whales with cheese. 

Ken Lempit: You're absolutely right. Probably not. David, if people wanna learn more about Capacity or reach out to you, how can they 

David Karandish: do that?

You can check us out. You can go to capacity.com or you can just ping me, david@capacity.com. 

Ken Lempit: That's wonderful. If people wanna reach me on LinkedIn at linkedin/in/kenlempit my demand generation and advertising agency for SaaS companies is Austin Lawrence Group. We're at austinlawrence.com.

If David's episode didn't convince you to subscribe to SaaS backwards, I don't know what to say. David, thanks so much for joining us. 

David Karandish: It was a great, yeah, thanks for having me, man. 

Ken Lempit: Lot of fun.