The $900M flip of

Joining me is an entrepreneur who bought a public company and sold it just three years later. The company he bought was I want to find out what he did in those three years that allowed him to sell it for over $900M.

My guest today is David Karandish. He is currently the co-founder of Capacity, an enterprise artificial intelligence SaaS company focused on helping teams.

David Karandish

David Karandish

David Karandish is the co-founder of Capacity, an enterprise artificial intelligence SaaS company focused on helping teams.


Full Interview Transcript

Andrew: Before we get started, you’re about to listen to an entrepreneur whose whole life changed because he figured out marketing.

So I want to talk to you about my very important sponsor.

Let’s pretend for a moment that you’re about to launch a campaign, get tested it. Your entire team is happy with it. Everything is going according to plan, except for that one thought in the back of your head, how do I ensure the people that I want to target are going to be in the right mindset to receive my message?

The answer LinkedIn, because when you market on LinkedIn, your message reaches people who are ready to do business, and that means your advertising campaign will work as hard as you can. As soon as you launch it, you already know that over 62 million decision makers are on LinkedIn, including me. And they’re thinking about their business when they’re on that platform.

It’s one of the many reasons why then 78% of B to B marketers rate LinkedIn as the most effective social media platform at helping their organization achieve specific goals, it comes to Mixergy guests. We use LinkedIn because it allows us to not just reach out to anyone who’s an entrepreneur, but somebody who’s an entrepreneur with a business that’s successful.

That’s. We go through our whole checklist and LinkedIn helps us do it. That’s what LinkedIn will help you do. LinkedIn can help you reach your short and longterm business goals. They offer tools for brand building and lead generation. Not only can you target and reach a professional audience down to their job, title, their company name, and their location.

We’re talking about amazing targeting. But you can engage people you already know based on who’s visited your site and who you are contacted in the past. You can even customize your campaign based on the action you want your customers to take, and the objectives that you want to achieve doing business on LinkedIn, the world’s largest professional network can help you reach your marketing goals.

So here’s my call to action. Do business where business gets done, get a $100 advertising credit. That’s right. A $100 advertising credit towards your first LinkedIn campaign. All you have to do is visit

That’s terms and conditions apply.

Let’s get into the program.

hey there, freedom fighters. My name is Andrew Warner. I’m the founder of Mixergy, where I interview entrepreneurs about how they built their businesses. joining me is an entrepreneur who started a business or then bought another one and sold it three years later. And if that sounds confusing and it’s just a one sentence description, I feel like the journey that he took.

Is so different that it’s worth spending a little bit more time on and understand how he did this because it is so different and it could be confusing. Alright, David  is a among other things. The guy who saw and said, this could be so much more, he raised money bought the company was public at the time.

Wasn’t it. Yeah, bought the company instead of starting in a competitor, bought did something to it that I want to find out in this interview, what that was, and then he sold it. And as a followup to that, he said, you know, I think what the world needs is another way to do customer service, to handle customer service, chat, customer service email.

I got to tell you, David, there are tons of SOPs, tons of software now that we could use. And still David comes into the space with a company called, capacity. Got a great domain It’s a support automation platform. David I’m skeptical. When I hear that there’s so many companies that said, we’re going to automate your customer support.

You’re saying that you going to do it, but first let’s talk about answers. Do you remember the day you sold

David: Absolutely  any time you sell your company? There’s a mix of emotions that occurs, both , getting off the rollercoaster of going through any kind of cell process, wondering what the next chapter looks like, the, ups and downs of building a business and then getting it to a great exit.

I can remember one of the few times in my life where I had so many different emotions happening at the same time.

Andrew: What was that day? Was it the day when you signed? Was it the day when you shook hands? What was it?

David: I also learned in my experience that the deal is not done until it’s closed. And so, yeah, I would say that it was the day that the day that the deal actually closed.

Andrew: Signed. And then in your head, what did you think? I did it and Oh man, it’s done.

David: So if you ever watched the movie Slumdog millionaire, And you watch as the protagonist accounts on how all of the experiences he had done up until that point brought him to that place. I felt very much like that. And so I got my start doing custom programming in high school. At that time, if you could build a website for someone you were somewhere between a wizard and a warlock in terms of your magical powers.

Andrew: years are we talking about here?

David: So I graduated in 2001.

Andrew: Okay. How old are you now? Is that inappropriate


David: I’ll be 37 this month.

Andrew: Okay.

David: and then my, my business partner and I  after we started building white pages for people. We went on a bunch of different adventures. We launched a, eCommerce website in the celebrity fashion world. We launched a lead generation website in the financial services space.

It’s got a text to speech plugin for the AOL instant messenger. We tried to build a 3d internet world. So we we’d done a lot of different types of businesses,

Andrew: this was just the two of you coding it up together and bringing it to the world.

David: high school and college. Yep.

Andrew: Why, why didn’t the three D world work? Why didn’t the eCommerce world work? Why, why didn’t those work? And I know that expo group ended up taking off, but why didn’t the others?

David: Yeah. I think one of the things that we learned, the hard way, six businesses in is that so many entrepreneurs fall in love with their idea and the product, but they don’t know how to get it into the hands of consumers.

I give, , I give a good example. one of our earliest projects was a plugin for Dale and some messenger called aim talk. The basic idea was a Microsoft had launched their text to speech library, and this was a very early bot that you could add onto your. Your instant messenger that would speak your, your messages out loud.

And this is pre-Facebook pre Instagram kind of days when AOL instant messenger was the closest thing to a social network that our friends were using. we made this product, we tested it. We actually printed out CDs and went dorm room to dorm room selling.

And we pretty quickly realized that a door to door sales was not in my future.

Because it doesn’t scale well, so we ended up, putting the, putting the And that was our first distribution channel. Now the data we put it on I think we’ve got a few the downloads, but then we were like, well, what next? And so we started learning the ins and outs of search engine optimization.

How do we show up at the top of the search results? And that became our first foray into, Spending as much time on the marketing and customer acquisition side as we would on the product itself.

Andrew: And so what happened with aim? Talk

David: , I think we got a few thousand downloads over time. eventually we got an, a nice trademark dispute with a AOL AOL corporation over the name, aim talk,

Andrew: because aim

is their name and right. They don’t want everyone using it.

David: it’s a good way to know when you’re on the thing. As soon as you have trademark disputes coming up, that’s when that’s, when I tend to find people, people figured something out.

but well, the  project itself was. Was not a huge success. In a financial standpoint, it taught us a lot, taught us about developing custom software. Our taught us about distributing software taught us about the very beginnings of showing up search engines taught us about email marketing. We did some email campaigns early on and ended up working at a local startup called monster commerce.

it was something they were selling an early SAS version of the Shepard software. So if you wanted to sell your widgets online, you could use their, your software to do that. And they had built at that time, their business almost entirely through. SEO. And so my business partner, Chris and I, intern there, he was there first and I, I joined up semester two after that and we just soak it up like sponges.

What does it look like to market online? What does it look like? I’ll just Google work, how to search engines. And we use that to then feed into our retail business and our celebrity celebrity fashion business, and our, Our online lead gen business and what each project that we did, even though it may not have been as financially successful as we would’ve liked, the key learnings were then applied to that next business and that next project and all the way up until our sale.

Andrew: You know, it seems to me, David, like you got really good at that marketing part and that almost took over everything else. So you stop making software consumer software anyway, and started focusing on lead gen. You got good at doing search engine optimization. You understood it because you work with, I guess monster group.

Network solutions. Is that right? You were there and you said this is going to be our model. And this was in the early aughts, 2003, 2006. , when there was a lot of money in lead gen, right?

David: Yeah, I would argue there’s still a lot of money in lead gen. but it’s, but it’s a, it’s a much harder market to get into, at that time, if you were good at optimizing meta-tags, you could get to the top of the search engine. Now there are few more features, that you need to have to make that work.

Andrew: All right. Let me ask you a question. That’s a little too personal, but this is what I’m doing. How much money were you making back then? Two guys early on in their lives.

David: Yeah. So I think when we, when we graduated, we we’ve started college with, $1,500 in the bank. And by the time we graduated, I think we each had 60 grand after taxes.

Andrew: Okay. And so one of the things that I’ve recognized doing interviews here is I keep thinking of this guy, Rob Ross, and he got really good at this type of product of doing SEO, doing lead gen collecting payments. But it was empty and a lot of people just didn’t know to leave it early enough and move on.

What led you to finally say we’re good, but let’s move on to S to look for the next big thing.

David: I distinctly remember. dealing with the, what, what people in the industry would call a Google dance. So every 30 to 60 days, they changed their algorithm and we’d have to figure out how to dance with Google and like a dances in person. You don’t want to step on anyone’s toes. You want to keep to the beat.

You want to be dancing the right dance at the right time. But we w we realized two fundamental. Truths that were, that were in, in opposition to each other. One was that we happen to get in early in an industry that was taking off the online advertising space in the early aughts was so nascent compared to where it is today.

That just, just by being in that market in an early phase, we were able to see things, that we wouldn’t have seen. I always like to say. there are shots that you only see being on the basketball court that you can’t see up in the stands or on the sideline and at the same time. Well, we also battled very difficultly with how do you build a business on top of another organization’s ecosystem, great way to grow very quickly, but you also have the risks of what happens when they, when they changed their mind.

Andrew: Okay. We’ll get into the path of before you, now that I see that you were looking for something different, I read that you had a checklist of things that you wanted. Let me see if I could, do you remember it or do you want me to refresh your memory?

David: I’ve had many checklists, so go, go for it.

Andrew: I heard what you said was I want to have a brand this time, right? Because if you’re doing lead gen, you’re not mastering your own brand. I want it to be in the content space, which I don’t know why you want to be in the content space. You wanted a huge addressable market. That means that it couldn’t be.

Connected to one geographic location. It couldn’t be connected. Even sent one language or even one category. You wanted to have multiple streams of revenue, including some recurring revenue. Right. This was like a checklist that you guys would put up on a whiteboard. Am I right?

David: let’s  jumping our Deloitte and go back in time to the context. So we had a, you know, online retail website where we were selling celebrity fashion. So we were one of the big fond Dutch resellers, as an example, online here were those much trucker hats that the celebrities were wearing for about 15 minutes.


Andrew: Huge in Los Angeles,

David: Huge. Yeah. Big, big in the big, in the skater scene, big in the kind of all left scene. So we ended up, Selling these products online, but we realized the retail model is very difficult, low margin. You end up with a lot of inventory at that time. You end up a lot of inventory. And so we were like the only way we can make this work, because if we can scale out to multiple brands.

So we started contacting additional companies, but we were a bunch of kids that people didn’t want to. Give their product and inventory too. And so what we did is we actually started placing pages up on the site, around those brands before we even had any products. But right at that time, Google released their ad sense product.

And it was the first time in the history of the world where you could take the ads that showed up on Google and place them on your website. And so what we ended up finding out was, wow. The content actually monetize better tend to econ did. And so we ended up saying, well, what, what if we could build a, almost like a comparison shopping type experience where our content is just as valuable or maybe even more valuable than, than the econ itself.

And so that was the original idea that we, we started, Back in 2006 was to go build a comparison shopping website quickly realized though that we didn’t have any content. We couldn’t make the content scale. So we ended up partnering with the likes of the Yahoo shoppings and shopping back, comes to the world to syndicate their listings to us.

Now, as we started getting going, we found that people were searching for things that didn’t match our content sets. They might actually start shore. Auto products that, which we didn’t have, or maybe even services that we didn’t have. And so we, we ended up verticalizing what we were doing across an auto website and a health website and a travel website.

But we had, again, this issue where we didn’t have the content for every single vertical, we want it to be in. And the online advertising market was growing so quickly and so fast at that point that it was just a race to even exist and B be in those markets. And so we said, okay, well, what if we had a Q and a platform that we could place on those websites?

We, our users could go ask questions, get answers back. It’s good value for the advertiser. It’s good value for the user is good value for the search engine. So we built a little Q and a platform and we started applying it to our sites. And I don’t know, maybe we had 2000 questions and answers on it and we said, wow, this is a pretty interesting model.

If there’s someone a little bit ahead of us, that we could partner up with that could help us accelerate what we’re doing. And that’s what led us to ultimately go acquire

Andrew: And this was one developer who built this QA site and the content. Yes. And the content came from where,

David: I think our internal team started out seeding it, and then we would set it up so that people could ask questions

Andrew: so anyone could ask questions and anyone could answer it. Got it. And this was still a company that was bootstrapped or was somewhat involved at that point.

David: Yeah. So we launched the company in 2006 and then summit came in at the end of 2007. And then a couple of months later, the market went South in 2008, but we continued to grow, grow straight through it. And the thing that we realized at that time was that. The market was so inefficient for online advertising that the price for shoes and Nike shoes and black Nike shoes and black Nike Jordan shoes.

You could buy a click at one price and end up displaying advertisements for one of the other levels of depth and end up making a nicer margin on that

Andrew: From arbitrage

David: was a huge arbitrage model. Just like all the other shopping engines for doing.

Andrew: then did summit partners acquire you at that point?

David: So you bought a little more than half the business.

Andrew: Got it. Okay. And so that’s why your LinkedIn profile said that you are a portfolio company of summit partners, but you still owned a considerable amount. And then you said who out there could be a little further ahead of us. So you said we’re going to buy, why buy versus make you guys are so good at making.

David: Yeah. Great question. So we had honed the paid marketing platform at that point, so we could go launch a new vertical, bring the content in advertise to it, but there was a big cost associated with both standing up a new vertical, as well as, Doing the advertising costs. And so we said we would really like to diversify and not only be reliant on paid advertising, we’d like to have a lot of free organic traffic as well.

And so what ended up, what was fascinating about at that time is that was about 20 million in revenue, but all of the content was free. It was all submitted by the community and they had no paid advertising going on. So there was no cost of the traffic. So while we viewed it as a $20 million revenue business, it’s really like a $19.5 million gross profit business.

The way we would think about it because they had no traffic costs and no content costs. And so when we bought the company, sorry, go ahead.

Andrew: what would their only expenses then? What were their expenses at that point?

David: Servers people being a public company. and so we looked at it and we said, wow, this, this company is really to be public in the first place with all the Sox compliance and regulation that had come out prior to here, or since they became public. We also looked at it and we said we could turbocharge and optimize guys those landing pages with all the split testing and AB testing technology we had developed.

And at the same time, every additional lift. All of that gross margin, she fall right back down to the bottom line.

Andrew: So you acquired a, for like five times revenue, $127 million. If I remember right.

David: That’s correct.

Andrew: That’s okay. Still a lot of money for the time. And it’s a publicly traded company, which means by the way, you had other people who said, wait, we should have bought it for more money. They sued you. You guys dealt with it.

That’s why a lot of people wouldn’t buy publicly traded companies. If they’re private and at your size, your eyes are like telling me I’m on right track. How what’d you think get the guts to do this? Where did this come from? The idea let’s go buy this publicly traded company.

David: I always like to think that, you have to, as a CEO, you have to be a couple of chess moves ahead. And so you have to be thinking about how can you expand and not, and be willing to take some risks. We looked at buying another company. we actually looked at buying a. Retail me not for different different use case.

And that one did not, it was a, it wasn’t public, but it was, it was pretty sizable at the time. The deal didn’t end up going through, but we ended up in there interesting scenario where we, we had a business that was producing a fair amount of cash, but we also knew we had some risk associated with. the different search engines we were buying from and the revenue concentrations, we went as fast and as hard as we could to try to build a more diversified and scalable business.

Andrew: . I got to ask you about retail. Me, not in a moment. Let’s let’s come back then and just understand the story of This was a domain that I think bill gross. The guy from idea lab bought back in 1999. He was this guy who said. We could start all kinds of internet companies.

I’ll have like a farm of people who can create my ideas, bring them to life and raise a lot of money. And he did it just as bubble burst. And then he ended up being the poster boy, quite literally, or the failed, version one of the internet. I forget the headline, but I think fortune had him on the cover saying this guy just lost a billion dollars.

Why is he smiling? Something like that. Right. But he was onto something and then other people took it on. Do you remember the journey? Do you, do you know the history of it to give us like a quick nutshell of what happened?

David: I know, I know a little bit. So what was interesting was that. At the time started out as a encyclopedia like aggregator. So originally they were pulling from 10 different encyclopedic content areas. So if you typed in Michael Jordan, they wanted to be the definitive page on that particular topic.

This is pre pre Wikipedia. Well, then most people don’t know there was a, there was a little tiny company and the name is gonna escape me right now. that came out with this idea of doing a Wiki based Q and a, where anyone could ask a question. Anyone can answer the question. So actually bought that company, merged it with their encyclopedia.

Roll up. And then ended up where the quickie Q and a just took off as the community did really well. And, just building, building great content. And so by the time we came along, the, reference answers, I think is what we referred to it as was it about 10% of that business? And the rest was all, this is Wiki concept where people were asking all sorts of different questions.

Andrew: I think the company was called FAC farm, like frequently asked questions, farm that’s it. And then it was rebranded Wiki answers. Okay. And so at that point it had gone public, I think, on America, on the American stock exchange and went to NASDAQ. It was like you said, pretty small company when you acquired it.

I, I went back and I tried to see what you changed. Did you add like a retail? Me not concept to it immediately, which was coupons on the homepage. If you wanted to buy from best buy, you get a discount. And by the way, we get an affiliate commission. Am I understanding that? Right?

David: We did a lot of, a lot of experiments on the site re we added a long, longer form articles. We added a coupon section. We added a celebrity contributors. We added new content formats. We updated the search. Okay. I probably tried 30 different things on the website.

Andrew: So you, you did improve the product a lot more. I think, than I noticed, I thought it was just going to be, let’s just super charged this with our revenue engine and our arbitrage model, but it’s not, it was improve the product, improve the product. What’s one idea you thought would work well, but bombed. And then we’ll talk about one that did really well.

David: Yeah. we. We flirted around with the idea of doing paid answers for highly important topics. So you’ve got a health question or maybe a sense of some kind of set, very sensitive question. There was another company at the time, I believe it was called Justin answer. That was going down that path. And we looked at the size of our community and the size of the audience and thought that we could, we could figure that out.

And we tried really hard and we flopped. Well at that time and maybe they’ve turned it around since then. A lot of what we found out is that a lot of these paid question and answer websites, the only way that they could make the economics work was by doing recurring billing. And some of their recurring billing practices were not, is above board as I would have liked them to be.

Andrew: Meaning I would go in and ask a question, like help me understand the history of the site would charge would say it costs $10. If I read the fine print, I’d realize that it’s $10 a month until I canceled. Got it. You’re nodding. That’s the thing. So. But you know what a lot of people tried that model.

I remember interviewing Jason Calacanis at the time and he looked at me with such a smart look like I went to Korea. I understood that in South Korea, this is the way the model works. I brought it back. It absolutely is going to work. And then it failed. And I never got to ask him why it failed. I heard him do a little bit of introspection on it, but I’d love to hear from you.

Why do you think paying for answers never worked.

David: I think part of the challenge is that we, we had a free community where we were trying to bolt a paid product on top of that was in. And, what’s the word I’m looking for? It was in opposition to the mission of the people who to manage the product, the, the community folks on, on where the lifeblood and are the still the lifeblood of that, of that company.

And. They’re not paid they’re, they’re not in it for the money. They’re not they’re in it for categorizing and organizing knowledge. We’ve had, we had people who met in the community who ended up getting married because they manage different sections of the website and fell in love with each other’s, you know, Roman history curation or, or, or whatever.

And so, You know, we, we learned pretty early on that you can’t, you can’t always just apply a paid product on top of a free product.

Andrew: Okay. What’s one that did work really well for you.

David: the one that did work for us really well was doing split testing and optimization on every pixel on the site.

We found that there were very, very small changes that we could make that would have big dramatic impacts on both. Questions submission rates, monetization rates, any, any major metric. We are trying to move

Andrew: like what? What’s a small change that would have a big impact.

David: anything from the color of the buttons, to the placement of where the submit looks like to the, showing avatars next to the questions, all of these types of types of small, small tweaks.

Andrew: And then I never did ask you, where did the acquisition money come from? Did you go back to summit to raise more money from them? So you could acquire


some summit put in, some amount of subordinated debt. We had some cash on the balance sheet, and then we got, we got another third party to provide some debt. So a hundred million of the 127 was, was through debt and 27 of it was from the balance sheet.

Andrew: Got it. Oh, Alrighty. You didn’t have to get diluted any further.

David: Not in that round. No.

Andrew: Wow. This was, this was huge. Do you remember the day that went through?

David: Oh, I remember, I remember that we were flying back and forth to Jerusalem. We ended up getting 57% of the vote, which is pretty low. but all we needed was 50.1% to complete the transaction. And what was. So crazy about that whole process was that there were, there were another two bidders that came in toward the tail end that tried to offer a higher price, but were, had grammatical errors in their, in their letters of intent.

So it’s hard to know if they were serious or not. And. We had a very bi-modal distribution of responses to that deal. Half the people thought we were crazy for paying $127 million for a company that wasn’t growing that there half the people thought we came out with a, with a great deal.

Andrew: And you did come out with a great deal and the opposition. I didn’t realize the grammatical mistakes, but I saw one of them raised money from some pretty well known people. I think Bain capital was involved going up against you and still you want, and it was a Jerusalem because I guess it started out as guru net in Jerusalem in 1999.

And then they just kept changing. They just kept growing out now growing the name

David: But Bob Rosenshine the CEO there as well as did a fantastic job building up as a sign of utmost respect for him.

Andrew: I’ve been so curious about what he’s up to now. let me see. Wikipedia. Does that tell me anything about him? He is, Nope. Doesn’t say much about him, but he looks happy in the photo there. Alright. So then you sell why.

David: Yeah. So we had looked at trying to go public. By this time we had a second private equity firm in TA associates. And, summit had been in for awhile TA had been in for awhile, but not, not quite as long. And we were looking at saying, we want to go public to either raise the next round of capital or, provide some liquidity to shareholders.

And in the middle of that process, we ended up dual tracking it where as we were working through the  and going through those details, you had a private equity firm come in and ended up buying the company.

Andrew: Wow how much? I didn’t see that was it. Oh, no. I do know that. Well, I let you tell it.

David: So it was $900 million for the main business. And then the part of the original business got spun off for about 60 million. It’s about 960 million in total.

Andrew: How did your life change after that?

David: Yeah. I mean, it was a, it’s a surreal experience. Anytime you sell, you sell your company. both. All the hard work that had gone in trying to figure out what was next, trying to leave it in as best hands as possible, but also realizing that once, once you don’t own it, you don’t can’t call all the shots.

So that was, that was a challenge and difficult. I started my parent company when I was 22 with my college roommate and, and so. it was my career post-college was, it was a, it was a crazy ride. I was like an entire career packed into a 10, 11 year experience.

Andrew: You owned over 10% of that at that point. Did the, did you, while it, sorry, what were you gonna say?

David: I owned about 10% at that point, but I ended up rolling a lot of it in the sale too. So I, I, it wasn’t a big transaction for me, comparatively.

Andrew: Did you buy anything for yourself as a memory? Did you? One of the things that I always liked about wall street, I wanted to go to work on wall street. They would have these little like trophies they’re loose sight thing that they would have of the tombstone from the wall street journal or something.

At least they would have something that said I was here. I did that. I get to hold it in my hands. Did you do anything like that?

David: I do have a little, a little tombstone.


from when, from the

tombstone, the investment bankers. Got it for us. Yeah.


still sits in my office today.

Andrew: Does it matter to you or is it something that’s just around?

David: I think my kids like to play with it. but that’s, that’s about it.

Andrew: All right, then you could have done anything. Why get into the, the tech support customer service, email space.

David: So after leaving, just selling answers, leaving answers, I wanted to figure out what I wanted to do next. And. I started reflecting back on a couple of things in no particular order. One was I had about five months or so in between answers and starting this company. And I realized something I didn’t. I didn’t really know about myself, which is I need to be on a team, played team sports in high school.

I wasn’t good at sports, but I realize I just, I like being on a team. I like leading a team. I like being part of a team and that

Andrew: if you don’t like sports, you just like people.

David: I like competing. I wasn’t, I liked sports. I wasn’t good at them, but I like competing. And I liked, I liked teamwork. I like people working together to accomplish a goal. And I like, Yeah. And so that, that five month period, it was a great family time and great time to relax. And I took my wife and kids to Montana and we saw some moose and took a little break, but I, I, I realized I really need to be on a team.

I like to be on a team, now that didn’t dictate necessarily going in the space that we’re in, but that was the first kind of first order. Second order was that I reflected back on getting into online advertising early. And I thought, wow, if I got into online advertising today, it would be a very different market, different position.

and I felt like I kind of did as much as I could in that space. So I wanted to do something that could leverage my skills, but I also wanted to get in early to a space, where. There was a big learning quotient. I love to learn. I read Wikipedia all the time. CA I’ve got a big stack of books on my, on my desk.

And so I wanted to go into a space where I could both have a little bit of skills and expertise so I could bring in, but also where I had a big learning, learning quotient. And so when I looked at what was happening in the whole AI world, it wasn’t even bigger up into the right graph than the up into the right grass.

I saw on the early thoughts around what was going to happen in terms of internet advertising. I think the market was a couple of billion dollars for AI software at that time. And now it’s projected to be, it was projected to be 60 billion by 2025. Then two years later, they raised that to 126 billion by 20, 25 and a couple of years from now, they’re going to rerock that up to 150 plus billion

Andrew: What was the early artificial intelligence? When was there one that we all are familiar with? Is it like Alexa and Siri?

David: At Christmas of 26 team, the top selling item on Amazon, wasn’t a doll. It wasn’t a video game. It wasn’t a toaster. It was Amazon’s Alexa. It was the first time that the most popular selling product was an AI. And I looked at that happening in the consumer space. And I thought, man, I wish I had an Alexa for my office, or I had Siri for my scheduling or I had, you know, some kind of automated assistant that could help me through throughout my Workday. And so we had just toward the tail end of my time at, at answers, we had just adopted Slack as our instant messaging platform. So I built a little prototype of placing capacity on the, you know, in the, in the Slack interface is a little bot and it was our first prototype of the, of the software.

Andrew: And what did it do? does that first version do?

David: So the first version was very simple. I could ask it a set of basic questions that came back with an answer that was version zero version one was I connected it to my Google calendar. So I could say, you know, schedule a meeting or what, what meetings do you know what what’s on my calendar for the day?

That’s where again, it just started building from there.

Andrew: By the way I fricking love your shirt. They usually company tee shirts are such garbage. That just looks like a superhero shirt. It’s super understated. It’s not in your face saying it’s capacity. I own bitch. No, it’s just, that’ll see. That looks like a comic book. Am I right

David: Thank you. Thank you very much though. We intentionally designed it to be like something you want to wear and street.

Andrew: Yeah, I constantly look at it. I feel like you, the design of your product in general is good, but now I’m getting off target here off topic. So you were just asking questions, building the whole thing yourself.

David: So I built the initial product and I hadn’t told you did not in a long time. So I think there’s still like 1% of. 1% of that code still exists in the system. It’s all been fixed and upgraded by much better and more talented developers than I am. But I built a prototype out where I could start to prove the concept and show it to people and say, Hey, imagine a world where you could ask, ask a bot, these, these internal questions.

Andrew: Okay. And then how did you go from that to customer service?

David: Yeah. So it’s interesting. Initially I had focused on the HR space and I looked at HR as an area where a lot of our questions were repeated. A lot of it was, was, stuff that you’d ask over and over again, like your vacation balance and your maternity leave policies, things like that. But what I found is that most HR teams don’t have authority to implement technology on their own.

They ended up having to go through it to make that happen. So when we started showing the earliest versions of capacity to these it teams, they were like, well, well, David, I’m doing a Salesforce implementation. Can you help me with that? Or we’ve got a big knowledge management product we’re trying to project, we’re trying to get off of SharePoint.

Can you help me with that? And so it kind of crystallized the product. It’s not really department centric. It’s really focused on any part of the org that needs support. HR support it, support customer support, marketing support, sales support. If you have any person in the org that gets lots of emails, phone calls, tickets taps on the shoulder.

That’s a good place for us.

Andrew: Okay. And so am I, am I misunderstanding capacity? Is it, it is for customer service largely, right. It’s replacing Zendesk for people who don’t want to answer every question themselves, or is it going beyond that?

David: So it’s a, we call it a support automation platform. So what that means is that we can help you automatically answer questions internally for your team externally for your customers or both. So you might have an internal, it help desk. That’s powered by capacity. You might have an external customer support website.

That’s powered by capacity. Now the way that, the way that we built out the product, and this is I think a really big differentiator from how some of our other competitors went about it. Is it, first of all, we started with the bot. So we didn’t start with a help desk. And we tried to shoehorn in AI. We started with the AI first, but then very quickly realized that I like to joke around.

Most bots are like those fainting goat videos. My kids like to watch on YouTube, where if they get scared, they just fall over and freeze up. And so w core to our foundational principles and how we built up the software out is that if you ask capacity question and it doesn’t know the answer we knew very early on, we were going to spend a lot of times stickering out.

How do we intelligently route that to the right person on your team so they can follow up with you directly. But when we do that, we actually store that question in the knowledge base. So the next time somebody asks that question, The bot can answer it directly. So the bot basically learning the ins and outs of the org with every question that you throw at it.

Andrew: So if someone on my team needs to know, for example, what do I say? If a guest wants me to edit out the sale price, they could come into capacity. Ask that question. And even if they’re not asking it in as. Well, I didn’t ask it very clearly, but even if they’re not asking, clearly the bot will either know it or if it doesn’t know it elegantly pass it on to me and say, Hey, Ari, Andrew’s going to get back to you in a bit.

David: Exactly.

Andrew: Okay. And then the first

David: So we ended up building a, we ended up building a help desk later to handle the human part. We actually built the AI part first. And so the help desk is really the tool of last resort. It’s the category that we’re in, but we think that the category got the whole thing backwards by starting with the human piece and then tacking on the AI because we launched three years ago, we got to start with the AI piece first and now build out the rest of the human experience.

Andrew: You know, I do find that the customer support email software that I look at very slowly keeps adding a little bit of AI, a little bit of, and it’s more like you wrote this answer before, let me help you save it. And then I’ll intelligently let you bring it up as a template for the next one. And by the way, we might, as somebody type it types in a question in the future.

Suggest this pre-written answer. And that’s the way that they’re going with a little bit and a little bit more AI as, as they continue. Okay. And so you wanted to go directly into artificial intelligence. One of the issues that I have with passing it on to passing questions onto artificial intelligence is it’s the, it’s the microphone question.

I used to get the same question over and over again, Andrew, what Mike do use, and I would email people. The blue snowball, the blue snowball, the blue snowball. And then we said, we’re answering this for so many people. Let’s just create a template with all the equipment that I use. And we sent it out until somebody said, Hey, Andrew, I think a year ago, I saw you switch over to the road podcaster.

And I realized, Oh yeah, I did. But I don’t think to go back to customer support software updated. How do you deal with that?

David: Great question. One of the things we recognized early on is that knowledge basis, which are related to bots, but we’ll just focus on that for a minute. Knowledge bases, most knowledge bases are not designed to. Clean themselves up. So you look at classic SharePoint. So many Internet’s are powered by SharePoint would be roll is SharePoint websites.

They’re all excited about it. Hey, great. We’ve got this intranet then a week goes by and then a month goes by and then a year goes by. Yeah. And it just continues to become stale. So one of our, or other founding principles of how we, how we built out the product is that we added then expiration date should encourage our clients to not.

Let things be evergreen, put an expiration date on all of your knowledge. And in that way, we’ll, we’ll let you know as it’s starting to change, but we can actually give you a proactive prompt and say, Hey, it looks like this information is becoming stale. What should we do about it? The other big problem though, is that most knowledge bases don’t have a good feedback mechanism so that every single response that comes back, you give you a thumbs up, thumbs down.

If you thumb it up, we’re teaching the machine learning to match that a little more often. If you dumb it down, we’ll actually come back to you and bring in that human in the loop to go improve the response.

Andrew: By the way the bottom of your webpage says, capacity was found in 2017. and as part of, what’s and what’s the connection to equity.

David: Equity com is a business incubator that my, my buddy and I started back in 2017. We have a multistage approach where we both launching businesses in the, for profit space, but also nonprofits. So we launched a nonprofit called create a loop. That’s teaching kids, computer science. with a one for one model, we’ve had a couple hundred kids who’ve gone through the program.

and regardless of what your socioeconomic background is, we’ve got kids learning to code and build flappy bird clones, and AWS websites and that sort of thing. And so is our parent company over, overcapacity.

Andrew: All right. Why don’t we close it out then with an example of how capacity has helped the client before we got started, you told me about West community credit. Am I right?

David: Yeah, it’s a great, great example. So West community credit union wanted to help reduce the calls going into their call center in the midst of a season of growth. So over about an 18 month period of time, they doubled their assets under management. By bringing capacity in, we were able to reduce their call center calls by 25% and lift their NPS net promoter scores up by 10 points.

So more revenue, less costs, better experiences all, all at the same time.

Andrew: And so now when I go to their website, I’m on it’s West community, C If I hit the chat bubble, it’s a bot. I called Jane that responds to me and I can ask something like, where’s the news ATM. And even though they’re getting more customers, you guys can come in and answer this and they don’t have to take on more calls.

And I see the thumbs up thumbs down. Yeah. Alright, great use case. Thank you so much for doing this and for not constantly telling us it’s brag like crazy. If I own that, if I own You know, stopping me. I, I appreciate your subtlety with that. And I appreciate you coming in here and telling us the story of and now

Thanks so much.

David: Thank you very much. Appreciate it.

Andrew: Congratulations. Thanks. Bye everyone.

Who should we feature on Mixergy? Let us know who you think would make a great interviewee.