The Mistakes I Made; Reflections on Nine Years...
The Mistakes I Made; Reflections on Nine Years...
21-Oct-2019This article is reposted with permission. The original is available here.
It’s been about 8 weeks now since I have transitioned out of Cvent, the company that acquired DoubleDutch in June. I’ve had some time to sleep, to camp, to read scifi (hit me up for recommendations), to ride bikes and play hoops, and to spend amazing quality time with my wife and daughter.
So much good stuff has already happened; that mysterious, excruciating pain in my shoulders that had been nagging me for the past year has magically dissipated. My sleep metrics, daily steps, and heart rate variability (a proxy for relaxation) are hitting all time highs. And my daughter and family is thriving - presumably happy to have me around more. And finally, Cvent - a company that we competed ferociously with over the years - has consistently done the right thing in their treatment and onboarding of our employees and customers. Thank you @reggie / Cvent / Vista for your grace.
I’ve also had a chance to reflect on the last ~9 years of leading DoubleDutch; the amazing people and talent that I worked with (!), the things we are proud of, and the many, many mistakes that I made, both big and small. Some of those mistakes almost certainly impacted our outcome; these are the ones that will keep me up forever, and the ones I’ve tried to focus on in this post.
So here we go, let’s get some stuff down on paper.
First, the good stuff.
The People, The Talent
Oh man, this will no doubt end up being one of the great professional joys of my life. We were blessed with incredible, remarkable people during each chapter of DoubleDutch. In the early days, we were like a cult / band / gang, choose your metaphor. Our first 20 hires were perfect. Our next ~400 were mostly fabulous; smart, interesting, committed, and surprisingly comparable in talent and heart across our various offices in San Francisco, Phoenix, Amsterdam, Portland, Hong Kong, Philippines. And a special shout out to the ~90 or so that were with us for the final chapter - tough as nails, heads held high, stubborn as can be, fighting to win until the very end.
Why did the people part work? It’s tough to say for sure, but I have some theories. First, we did the work in the earliest days of the company to define and document our values, and were very serious about applying those values as a lens on hiring decisions. I personally interviewed our first ~250 or so hires (yes, sometimes a bottleneck, but ensured some consistency). I also think the quality and charisma of our earliest employees and leaders was able to snowball and feed itself; good people want to work with good people. Man I will cherish this team.
We Went for It
My startup experience prior to DoubleDutch was mostly bootstrapping; a drawer full of credit cards, paying payroll out of my own pocket in bad months, watching every dollar and scraping by.
DoubleDutch was not that. Once we hit product market fit and hyper-growth in early 2013, we really went for it. Our cost of capital was cheap, and we were not shy about putting $$ to work. We truly believed we were building an epic company for the ages, and this ambition informed every decision we made.
Build a giant event app company? Not big enough. We were going to build Google AdWords for conference advertisers on the back of our event footprint (alas, it didn’t work). Set up some demos for our sales people? Not us - we built one of the most formidable SDR programs in Silicon Valley - 65 SDRs strong at its peak, all making 70 calls a day, many starting dialing at 6AM. Win North America and then look abroad? Nope, we opened London, Amsterdam, and Hong Kong within 18 months of hitting product market fit. And so forth.
Even our homerun seeking investors would often tell us to take a breath, and slow down. We didn’t listen, certainly at times to our detriment.
At our worst, we were betting on everything, taking so many big swings we got in our own way. But at our best, we were unstoppable, and determined to pull the thousand+ year old industry of live events, kicking and screaming, into the future.
I’m mostly proud of the shot we took, mistakes and all.
Our Board / Investors
We didn’t bat 100% here, but all things considered we were close enough. This was not an easy board to be on, as we ran pretty hot, and in the hyper growth that we saw (sprint to $1M ARR, triple, triple, double) many, many things start to break. But we mostly stayed together, and I leave the experience believing that our investors had ethics and heart. A couple of quick call outs:
Ted Oberwager / KKR. Talk about an unreal work ethic. Ted and KKR could not have worked any harder for us, or cared any more about our success. Things I admire about KKR; the horsepower, the reach, the ethics / integrity, the cool heads under pressure. Ted, thanks for being in the trenches with me.
Mike Maples / Floodgate. The tougher things got, the more Mike rolled up his sleeves and got to work. He spent hours and hours with me trying to help us win, even as the likelihood of a homerun exit decreased. I’ve heard him called the best VC of his generation, and I’m not going to argue that.
Kristina Shen / Bessemer. Bessemer, the fund, stood by us through thick and thin, even as partners swapped out. A special shout out to Kristina Shen who showed her character when the pressure was the highest, and went for just about the full ride with us. Kristina, thanks for everything.
Duncan Davidson / Bullpen. Bullpen is the rare fund that could care less about social proof, and is not afraid to make their own calls. They bet on us very early, and were a rock solid partner for us throughout. Duncan is extremely sharp and still thinks like an operator. I valued having him at the table.
Allison Watson. Allison is an executive at Microsoft and joined our board as an independent pretty late in our development. My only regret is not adding her sooner.
Eric Di Benedetto. If you don’t know who he is, you better ask someone. He avoids the spotlight, but his record as an angel investor is unmatched. His business acumen is off the charts, and his directness and often brutal honesty has saved my ass several times, particularly in the early, fragile days of the company.
And now for the mistakes.
While we of course made our share of operational mistakes (personnel stuff, accounts that went awry, etc.), I tried to focus on only the foundational, tectonic stuff; the stuff that I missed on that likely had a profound impact on the company. The mistakes that I won’t make again.
So here goes:
We Didn’t Listen to our Customers Enough
One of the ways the level of our ambition manifested itself was in our belief that we could see the future, and that we knew what our customers needed better than they did. We were consumer people in an enterprise world, and we acted like it.
In a nutshell, we felt that the category of event technology would very soon become a data business, and that the vendor that could best detect buying signal (or employee engagement signal at internal events) would win, and win big.
Working backwards from this data capture endgame, we optimized our mobile event app roadmap almost exclusively for adoption / engagement of the attendee, as opposed to the requests of our customers, the event owners / organizers (aka the ones who were paying us). Our belief was that our relentless focus on adoption / engagement would generate a bunch of actionable data signals for our customers, which would massively expand the ROI of our customers’ events.
On the Vision Led vs. Customer Led continuum, we were off the charts Vision.
For a while, this was precisely the right strategy. We came into the industry unlike anything anyone had seen before in the world of events. We attracted all the press, all the capital, all the talent, all the early adopter customers. We were different and radical, and for a while we were unstoppable.
But when a customer would go into Year Three with us and we still hadn’t fixed / built that seemingly simple little thing that was crucial for the them, Because Vision, customers start to get frustrated.
And if you let that frustration go for too long, you start to have a churn problem.
I can vividly remember some of the moments that we could have listened more to our customers, yet still preserve our Vision led culture. But we mostly didn’t bend, and this would end up hurting us. A consumer ethic in a B2B world can be a powerful force; but you can’t be so rigid in your vision that you lose sight of your customers.
Too Slow to Properly Instrument the Business
And speaking of churn. Events are a curious business, and not a completely natural fit for a Saas business model. Some events are predictable and repeatable, and some are one time, non recurring (read: automatic churn).
To account for this, we created two segments for reporting purposes Standard, and Non Standard. At the time, we didn’t have much in the way of a Finance department so we mostly allowed the revenue team to make their its own calls as to what non standard was.
This approach did not have a happy ending.
One day we woke up with a ~$20M run rate business in which a shocking 40% of our business was in the Non Standard bucket, running at Really Bad churn rates, with the other 60% performing at OK but not great retention levels. You net these two buckets out and you have a big problem.
There was sort of a perfect storm of stuff that went wrong here.
- Lack of governance, underinvestment in Finance / G&A / reporting
- Misaligned incentives to give revenue oversight of nonstandard bucket assignment to revenue team
- A business that was growing topline historically quickly with all of the related people / growth distractions
- An admittedly tough category with low switching costs, and extremely high service expectations.
It’s easy enough to grow your way out of churn problem in the early days. But that hole in the business gets awfully big, awfully fast if you let it fester.
The Rewrite
One of the “distractions” of hyper-growth was scaling our system to keep up with the exploding usage of our apps. Live events are not easy to scale for. Usage tends to be incredibly bursty (think 20K people responding to a survey within 5 seconds). And if your app goes down during a CEO keynote, it’s an automatic churn - you simply don’t get the chance to win back your customer’s trust in our industry. Unlike most categories of enterprise software, the success of your customer relationship often comes down to your stability and performance during a few hours of the year. It’s incredibly stressful for everyone.
So in 2015, after looking at the insane growth of the business, our engineering and product leadership came to the decision that we needed to get out ahead of things and completely rewrite our legacy, monolithic system.
(not everyone was on board with this call - two of our most senior engineers came to me directly, multiple times to protest the decision).
Ultimately, this rewrite never shipped, despite the talent and effort of the engineering team that worked on it. It ended up costing us about $5M, not including the morale and opportunity costs associated with the project.
There were multiple levels of mistakes here. The decision to kick off a project of this magnitude without properly vetting just how big the project was, which resulted in wildly miscalculated timelines and budgets. The process decision to try and build out of our Amsterdam office, which kept the engineers who were building it away from the all of the institutional knowledge in HQ. And finally, we certainly could have killed the project earlier, and we didn’t - someone please remind me about what “sunk costs” are. Ouch, just a lot of pain here.
Too Timid on Cuts
When the extent of the churn problem became apparent, we moved to reduce burn through two painful RIFs. Our board wanted us to cut deeper, and we resisted. Of course it’s an emotional time and very hard to think clearly as a founder during these moments. These are good people that are losing their jobs, people that bet on you, people that are high performers, giving it everything that they had. Your inclination is to fight to keep every person.
But I wish we had pulled back even harder than we did - you can always run the business with fewer people than you think, and the key is to preserve your cash so you have enough in the tank to get out of the hole when you have fixed the business fundamentals.
We did end up mostly fixing our fundamentals through a disciplined, focused 2018. But ultimately we didn’t have enough cash in the bank to get back to growth, and no new cash was coming in on top of our heavy cap table until growth returned. Put another way, cutting deeper in 2016 likely would have made 2019 a different year for us.
Other mistakes I made that could easily could have made the list:
- too broad of a customer definition; almost certainly related to our churn problems. It’s hard to serve wildly different customer profiles, well. As Mike Maples likes to say, “find the customers that would be batshit crazy not to use you, and focus (only) on them.”
- too slow to get out of SF; until this place starts to build some housing, this is not the best place to scale a team because, cost, high attrition, and uh, the inflated expectations of folks who haven’t been through a cycle yet. We opened our PHX office in 2016, and should have done it earlier.
- too slow to platformize; as quarters turn in to years, selling a point solution against a platform that is determined to kill you is a tough slog.
- Not enough focus on sustainable, cost effective demandgen; when your cost of capital is low, it’s easy to throw $$ at SDR, PPC, and other high cost channels. But eventually you need a sustainable engine.
- We chose a tough category; we were a software company in a f2f, services world, and it caused ongoing stress. (I’m lukewarm on this one, as Cvent has shown it’s possible to build a very good business in this space, and we certainly weren’t complaining about the limited and fragmented competition).
As I parse through this history, it’s interesting that I don’t see a single mistake of consequence post 2016. It was the decisions that we made during and around hyper growth that likely wired our destiny (though we fought like hell to win until the sale).
Ultimately, all of these mistakes (and more!) were my fault. As founders, perhaps our greatest superpower is our stubbornness; a superpower that can certainly cut both ways. There are certainly some calls I wish I had back, and some mistakes that I made that I won’t make on the next one.
But man, it was a hell of a ride.
So What’s Next?
DoubleDutch has been a huge part of my identity for the last nine years, so the last 8 weeks have been a bit surreal.
For the first few weeks, I didn’t want to do anything even remotely work related. I tried to read a motivational business book and made it like three pages in before switching over to science fiction. Podcasts were out of the question, and VC / startup twitter was nails on a chalkboard. I dodged all coffee and call requests, even from people I like and admire. I completely overdid physical exercise. Looking back, I was pretty beat up emotionally, maybe more than I realized.
But about a month of downtime, I started to get the itch again. I started kicking around ideas with my wife and friends, and writing some things down.
And now I feel ready to roll. While I can’t share fully what the next thing is, here is what I can put out there:
- I’m going to keep Operating, and I’m going back to Zero to One.
- I’m going back to Consumer. I enjoyed the B2B run, but having done both B2C and B2B, you can stay truer to the Mission / Vision in Consumer. At this stage of my career, this is important for me.
- In terms of domain, I’m doubling down on my work at the intersection of the physical and digital spaces. I believe that we are at a pre Web 2.0 world for hybrid digital / analog companies, and that the opportunity is enormous for software to make f2f better. The next great consumer software company will have a strong physical world presence, and I don’t think we need to wait for a new Platform like AR to start building this company.
If you would like to hear about updates for this next project, please tap through.
To DoubleDutchers, thanks for the ride - hit me up if you want to grab coffee / beer.
Add a Comment
Recent Comments