I think I’ve scared the bejesus out of the marketing department (aka Brenden and Kelly, who sit next to me) with the decision that the CEO wants to write some long form content on why I bought back my last startup, Snapsort. We’re going to hit send on this, look at some heatmaps and realize 80% of people bailed at the first paragraph (right… here.)
Or not. We’ll see. The thing is I have no clue about whether anyone is reading long form content, but I’m personally enjoying it (waitbutwhy!) and if I had to write some kind of click bait bullshit, I’d puke. “He bought his business back - you won’t believe what happened next!”. So I thought I’d try and write something interesting to read. I hope it’s worthy of a latte and the soft place you curl up on a Saturday.
Speaking of no clue, a lot of my time as a founder, I’m flying around with no idea what I’m doing. I think that’s normal though, because if you’re actually hitting the 10% growth a week goal so many of us aspire to, the amount of newness in your day is insane. Growing 15-20% per year is an exercise in efficiencies and incremental improvements. It’s tractable, you work through it, there are people who are good at that. But 10% per week is like voluntarily strapping yourself to the front of the Millennium Falcon while Han Solo takes you through an asteroid field, drunk, asking you to direct him because, well, you’re blocking his view. You’ve got to think fast and wing it. Presumably, you have a great team, you don’t scare easily, and you work your collective asses off to make it happen.
What does that have to do with buying a business back? Good question. Well, probably not much other than buying a business back was just another really interesting, net-new challenge I was faced with. I essentially winged it the whole way through and thought it might be an interesting story for other founders to read, because I learned a lot in the process.
Buying your business back sits outside of the items the world typically puts on your to-do list as a founder:
- Start company with huge world-changing vision
- Release product and get real market traction
- Get written up by the tech press
- Get millions of users
- Raise money from someone who seems more important than you
- Hit revenue targets
- Sell company
I was fortunate enough to have hit all the milestones and sold when we were 7 people and around two and a half years old. In retrospect, the sale made sense. I’ll never know if it was the optimal outcome, but it was a hedged bet and a de-risking strategy. Inside the echo chamber that is startup news, it’s almost impossible to know what normal should look like, but it still felt a bit like selling out. That is part of the reason I bought my company back. To understand why we sold, it’s important to understand where I came from. I’m not a 22 year old entrepreneur - when I was that age and ran my first startups, I was operating in a vacuum, and it’s that experience that helped craft my perceptions of risk and reward when building a business. My prior two tech ventures were not huge successes, especially not according to the press’ definition of success, but for me, in retrospect, they were critical, incredible and important in many ways.
My First Ventures: Booksoft and BrightBlocks
BrightBlocks was a consulting company we started in school (2000). We were basically code monkeys: two engineering students, two computers and no clue what we were doing. I was the closest thing to the business founder so I got us clients, which left Alex Black as the technical founder, and off we went. This was all after the tech bubble had popped. Our clients were mostly horrible, soul-sucking mega-corps. Our most colorful client would leave our employees in tears if we let him get too close, so I learned a lot about how business founders shield the team from bullshit. It was 2000, and at the time, Waterloo was an ecosystem devoid of mentorship, funding, accelerators or anything useful to startups. You could count the number of software startups on your hands - it was nothing like it is today. It was lonely, uninteresting, and stressful. I grew up a lot in those years and had my fair share of dealings with assholes but it gave me a backbone (and a bald head - still not sure if I liked that trade).
The first product company we started was called Booksoft (2000-2005), which was focused on providing digital yearbooks to schools. We had hundreds of schools using it, distribution deals with all the major yearbook companies, and a sales team earning real revenue. The best part, though, was a free online community for students to keep in touch, share photos with friends, direct message each other, maintain a friend list, and write on what we called a mini-blog. Students loved it and we could see why: they used it to creep each other. The problem was no yearbook meant no access to the community - we didn’t understand the power of giving it away or of network effect, so it never reached critical mass. But we built something we were incredibly proud of.
We funded Booksoft out of consulting revenue, investing close to a half million in that business as poor twentysomethings. In the end, we also managed to wrangle about $700k out of the consulting business before we shut the whole thing down. There was no “fuck you” money and caviar. Alex and I shut things down, said fond farewells to our employees, and had an “expensive” dinner at the best restaurant in town before we set off on our own paths.
During the time I fell in love with building products, Booksoft was magical for me. The consulting business, on the other hand, was the devil. As the business founder, I dealt with everything untoward and in the end it was the asshole client, with the $800k bill they wouldn’t pay because “what are a pair of kids going to do about it”, that pushed me over the edge. It’s what caused us to shut the business down, and it took months of my time to get it paid (while everyone moved on, which is another excellent story).
After all that, I said “Fuck it. I’m taking a break from tech”, and decided to use my hands to reno properties. I did two by myself, learned plumbing, electrical, drywall - you name it. For the next seven, I hired contractors because, yes, there are people who are a LOT better than me at tradework. It was a good time to have bought a lot of real estate, and I did okay in a non-unicorn way from those investments.
Two years later, I sold all the properties, we had our first child, my wife left her job, we moved into a tiny house, and I decided that being a tech entrepreneur is pretty much all I wanted to do. So with no income and a newborn it was time for the next big idea: Snapsort.
Snapsort: The Company I’d Eventually Buy Back
Fast forward to 2009 (after the second tech crash). By this point, the Waterloo tech community was not insignificant: there were accelerator centres, VCs were starting to pop their heads into town, angels were actively investing, mentors were actively helping, and there were plenty of other startups. The wounds of yesteryear had faded away, and Alex and I once again joined forces to, this time, “to build a real internet startup” - none of this BS we had done before with consulting and selling CDs to schools. We wanted B2C, millions of users and something that was a big deal on a global scale. We were fully sold on our ability to change the world and build the next Twitter (at that point in time, Twitter looked like it could take over the world).
Purchasing decisions are hard, and we theorized that if we could take the collective knowledge of the web and let you query it, we could make buying things easy. Not search results but actual answers. Metaweb was an inspiration, and the slew of crappy product comparison engines fueled our desire to fix things. We set out to crawl the web and build a perfect, ontologically correct view of what the world thought about anything you could buy. Armed with that data, we could generate new content that’s super helpful. For example, if you read all of Yelp’s reviews, you could build what is essentially a new review on the best sashimi in Lower Manhattan.
Long story short, we built the tech and started launching vertical-focused sites using it. We raised money, got coverage multiple times by all the big guys, grew to over a million users, started making money, and best of all were beating all our US brethren: Chris Dixon’s Hunch, Oren Etzioni’s Decide.com, Kevin O’Connor’s FindTheBest - and we were doing it with a tiny group of really smart engineers, a PR person, and an order of magnitude less capital. Then the acquisition offer came. At the same time we were doing a bad job of raising a second round. By bad, I mean I didn’t know how to run a proper process, and I certainly wasn’t getting in front of a tenth as many VCs as I should have been. In the end, we did have some investment options, but we weren’t even being offered the cash portion of the buyout offer, let alone the optionality on some significant upside. The deal seemed decent: derisk, in exchange for some good cash now and a chance to roll the dice on proper beluga caviar money. Part of me wonders how being in the Valley could have rearranged the outcome. The money our counterparts were raising to build products with fewer users and less technical sophistication baffled us. All but FindTheBest ended up being acqui-hired and shut down. But we had two good options in front of us, and the acquisition offer was just better in every respect.
The company that bought us, Rebellion Media, was a US investor backed firm looking to do a roll up to IPO. They had raised a decent war chest and had begun picking up a wide array of internet companies. The basic tenet was simple: a lot of companies will never get revenue scale to IPO, and a lot of those companies shareholders would welcome a liquidity event. If you could piece enough of them together in a cohesive way, you could list the company on the public markets and create a lot of upside for all the founders and investors. Mostly Rebellion needed revenue scale, but what we provided was sex appeal: the ability to launch and monetize an endless array of high-quality content on almost any topic. We were the engineering, tech-driven, sexy icing on a big revenue cake.
Inside the Belly of the Beast
Rebellion and Snapsort was, in many respects, a match made in heaven. You hear horror stories of borg-like assimilations, products getting shut down, legal tying up productive work for 12 months, teams split up and jettisoned founders. There was none of that. We got there, they told us to do whatever we thought made sense and they gave us the capital to do it. We went through millions of dollars of their capital, building building building. The team grew from 7 to 27, and we launched more domains, services and apps off of our engine. I didn’t even have golden handcuffs; technically, I could have left on day one. The trust they placed in me and my team was incredible, that was a strong management lesson for me on building relationships that work not because they are forced, but because you have the right people and aligned interests.
Unfortunately, a lot of good stories come to an end and that was the case here as well. While we reached 6M monthly users for our various recommendation services, a notable achievement, yes, but we also realized why B2C marketing and demand generation is the hardest thing in the world. We had no data to support a clear path to 100M MAU. We had built a good business but not one that could be great without serious pivoting and reinvention.
That would have been okay as long as the rest of our mates on the SS Rebellion pulled through and gave us the revenue path to IPO. To make a long story short the thesis didn’t play out, there were a lot of acquisitions made and some went bad. The ones that went bad, went bad enough that it made the existing revenue cake look sufficiently less attractive. This blocked new acquisitions, and without new acquisitions, Rebellion couldn’t reach revenue scale. No critical revenue scale meant no IPO, and no IPO meant that that investors wouldn’t be getting their money back quickly or perhaps at all.
Investors are in it to make money and in matters like this, their marching orders are understandably clear. They made the decision to ensure Rebellion focused on paying them back. They were in first position, and everyone else wasn’t. That meant that things that made money were cut back a lot, and things that didn’t make money were cut back to the bone. I remember standing in the office of the CFO when he asked me a simple question “Chris, what is the minimum number of people you need to operate Snapsort?” I answered “3”, and knew what that meant. A few weeks later, my entire team sat crammed into the large boardroom. Christmas was less than one week away and everyone was tense. It’s hard to hide all the extra meetings, the feelings of remorse, disengagement and distress. People knew something was happening, just not what. Firing 23 amazing people was, as expected, awful. After that, I worked hard to ensure everyone got a job. I guess it was good for Google, as they picked up a lot of incredible engineers. In the end, as shitty as it was, everyone landed in a good spot at a great company.
I managed to keep 5 of us, including myself. It was a team, while much smaller, that could get some serious shit done and I had no intention just sitting around. My technical co-founder, Alex, left for the west coast to start something new, and so began the journey out of the beast. The common shareholders had a mountain of pref and debt sitting on top of them - the likelihood that investors would get paid off with anything left for the commons was low to nil. There was really no reason for good engineers to stay while things unwound, so I needed a plan.
The Desire to Buy it Back
Snapsort had always been a standalone business within Rebellion: development, marketing, sales were all part of our fiefdom, but it was more than that: we had our own culture, style, events and focus that meshed well with our mission. We came back in the new year and it was just us, a cavernous office, and some vague ideas about what to do next. A lot of that culture and feel just disappeared after the exodus. What remained was a mandate to help Rebellion’s investors get paid back over the next 4 years. It was clear early on that it wasn’t a workable solution and so I began, almost out of nature, to plan the exit.
Being an eternal optimist meant I got excited about the future despite the drab present. I knew we could do something big, I knew we were an incredible team, and I believed strongly in the ideal of getting it done. The problem was those thoughts and feelings are very different for an optimist. I can generate those out of very abstract and disconnected ideas, sort of a gut feel, before I can generate a solid business case. So without even knowing what I would do, I knew what I had to do, and I had to do it while the rest of team felt very differently than I did.
It seemed a damn shame to just let the Snapsort assets rot away. Snapsort was a B2C-focused publishing business built 100% on organic users. Our sites were pretty sticky and growing. Getting them to that point was incredibly hard, but with most of the investment behind us and the founding engineering team remaining, we were in a great position. Since our sites essentially ran themselves, we had the opportunity to cash flow through a pivot of the business.
Turns out the romantic ideas of buying back your business are easier fathomed than accomplished. It took 10 long months by the time we wrapped it up, having enumerated through 20+ options. Perpetual negotiations with the CEO, CFO and Rebellion’s investors, exploring investor interest in a buy-back, and developing a net new idea in parallel was a taxing marathon.
The hardest thing I experienced was managing expectations and keeping morale up, which went through my mind every single day as I balanced being transparent with shielding bullshit. Rebellion was winding down during that time, most of the remaining staff were being slowly let go, innovation had left the building and not much remained. In the end, we lost two of the team members. Despite being a small team, and despite knowing in my own mind that we could do great things, the weight of the environment and situation were crushingly depressing. When the buyback was done, those two team members left their current jobs and rejoined as part of the founding team.
To make things worse, getting a deal done was a moving target. As I built interest and showed the ability to pay, the price moved. It was a giant game theory mindfuck trying to a) be a reasonable human without having the other side move the target and b) negotiate something reasonable for the team and investors. In the end, we got a deal that worked; that is to say no one was super happy about it, but we could agree to close. That process included a few sharp elbows, raised tempers and irrational outbursts. I didn’t always keep my composure, but for the most part we all behaved well. I have nothing but immense respect for Ted Hastings and Jeff Collins, the CEO and CFO of Rebellion, respectively, whom most of my negotiations went through. I’m also thankful that Rebellion’s investors took the time to get a deal done despite having much larger things to worry about.
An Idea Worthy of a Buyback
Because I felt the economics of buying the business back made sense, it would have been easy to deceive myself as to the fit and attractiveness of a net-new idea. Confirmation bias is horribly powerful and hard to resist, even when one can identify the overconfidence in real-time. This was a risk to the team, myself, and the investors who would ultimately back us.
Our big new idea was tangential to and based on all the learning we’d had as publishers. It was based in real pains we’d experienced, and those of our peers. We tested it out thoroughly while we were inside Rebellion - for most of that 10 month stretch, we spent time testing and measuring our thesis. We told Rebellion about it and shared the results with them along the way, some good, some bad.
But ideas take time to form, sales take time to build, and product market fit is almost never instant. As we built confidence in our new idea, we also made progress on the buyback. Even by the end, it was more well-thought-out thesis than slam dunk. In the end, the timing worked out well and that’s likely more due to luck than a master stroke of planning by us.
Economics of a Buyback
There was always the question of buying the business back versus bailing and raising new funds and a lot of people would ask me, “Why buy the business back when you have a new idea? Just go raise money.”
We focused on the buyback, mainly for economic reasons. Common advice is to raise for 18 months of burn, assuming you’ll have very little net revenue produced. It’s now been twelve months since we bought the business back and we’ve burned through $1.2M. By 18 months, it will likely be double that (growing company and all). Assuming 18 months of burn, we’d have needed to raise a seed round of $2.5M. That’s a big seed round by any standard, and also one that I likely would not have raised, especially for the idea we had (ad tech-focused, and thus the least attractive investment category on earth).
The alternative was to raise money and buy the business back, using the cash flow to fund the new company. We were confident enough that if we put some additional effort into the current publishing business to buttress its growth and monetization, we could support an aggressive spend on scaling the new business. The disadvantage of this approach is if the cash flow dries up then you’ve now lost two businesses, with no cash on hand to recover. That volatility scared the crap out of me but if anyone could make it work, and understood the risks, it was us - we had built and run the thing for five years.
Lastly, the buyout made sense because if investors believe they’ll be getting a new sexy high growth business and a cash flow cow then the valuation should actually be higher, it’s a de-risked strategy for them, so bonus: higher valuation. Also, if the cash cow holds steady there should be less need for future capital raises, so bonus: less dilution in subsequent rounds.
The end result of all of this is less dilution, and as a founder, that was important to me. I’m not crazy having huge liquidation prefs sitting on top of me and I liked the idea that this small seed could fund a billion dollar business without much more required.
The last question is why would someone sell? Consider that Rebellion was more or less winding down the assets it had bought at that point. Snapsort was a rounding error for the investors and there were a million lines of code with a team maintaining it that would never stay (even if I’d begged them to). The right answer was to find an agreeable outcome. The CEO and CFO were reasonable and while negotiating the right price wasn’t easy, and I wish we could have paid less, good deals are often the uncomfortable ones for all parties.
Raising the Round
Convincing investors that you’re going to buy a business back, de-prioritize it, then pour all the profits into a new unproven business is one thing. When that new business is also the most hated category by VCs, it’s another thing completely. This is where angels can be so much better than VCs: they’re willing to listen, can act quickly, there’s less pressure to be a “me-too” investor, they don’t mind contrarian stances, and there are no LPs, fund lifetime or thesis mandates they’re beholden to.
Raising those funds was pretty scary. I had only raised $500k for Snapsort (which we ultimately paid back - good story there on what to do with investors that are a bad fit) and this was a pretty weird transaction I was trying to close. I had no idea how people would react to the idea but sales is a numbers game, which meant lots of meetings. I built a spreadsheet and set out to systematically work through 40 or so angels. Guessing who will invest before they invest can be a real test of hubris and an opportunity for crushing disappointment. While there were some surprises in the end, there were not a lot - the people who invested knew me, and most who did not know me declined.
I had been building relationships with investors and angels throughout the years. Ultimately they invested because they knew and trusted me. There were no pitches or decks, just me talking about what I thought was possible and why. My investors listened and chatted for half an hour, looked me in the eye and asked me how much I wanted. I’m still not used to that level of trust. It’s humbling and an honor . It was an incredibly exciting time as the momentum built. It’s a fantastic feeling when people believe in you. That affirmation gave me an incredible amount of energy to build something amazing. Even a year on, when I send out monthly investor updates, I think about how thankful I am for their backing.
A Year In
We closed the buyout to little fanfare, choosing to keep quiet and focus on the business. We moved into the Velocity Garage (a startup incubator provided by the University of Waterloo) to help us reboot the culture and be around other small startups and set to it. We’ve been in a sprint for 12 months now, grown from 3 to 20 people since the rebirth, with plans to hire another 15 to 20 engineers and grow total headcount to 50 over the next 12 months (If you know any developers who are up for a challenge send them our way :-). One of us runs the original business, which also continues to grow and the new business generates more gross profit then the original one ever did. We’re growing scary fast, profitable and have no real need to raise. We’ve built an incredible and fun team and we have big mission to transform a broken industry. Last month we moved into our own kick ass office and culture that’s exciting to see in action. I don’t know what the future holds but I’m as excited about it as I’ve ever been and really thankful I went down the path of buying my old business back and launching a new one. A lot of people had doubts, a lot didn’t, but what really mattered was our focus and dedication as a team to constant innovation and to solving hard problems that matter to our customers.
Corollary: What is the New Business
Snapsort was a collection of vertically-focused review sites generated by crawling the web, organizing information about products ontologically and using that structured data to generate net new, highly useful content. That business as you can guess was built by engineers, not editors/writers. While building that business we applied the same engineering mindset to monetization and learned a lot. As we grew and matured as publishers we realized that the department most responsible for a publishers revenue was its ad operations group. We also realized that ad operations and revenue management is currently a horribly manual and broken process. We thought we could solve that. Our new business, Sortable, takes automation and machine learning techniques to mediate all the complexity in revenue / ad-operations. Instead of working with 100s of partners and technologies to optimize revenue, publishers just use us, we handle everything, they save time and make more money. Learn more about Sortable and how we make ads suck less.