Investors — Why love Picasso and ignore Cezanne?
Malcolm Gladwell’s excellent podcast series “Revisionist History” recently dug into approaches to innovation in a way that helped me understand why I’m uncomfortable w/ the tendency of the venture community to triage quickly and to bet only on the relatively obvious winners.
I’m referring specifically to the episode “Hallelujah” which refers to the Cohen song of the same name that, it turns out, Cohen agonized over for years.
To steal one of the punch lines of the excellent podcast — and you should listen to it if only to hear some of the stories about what it was like to be Elvis Costello’s producer (really) or countless other anecdotes woven together into an entertaining whole — Cohen was, like Cezanne, someone who never quite finished their work; he spent 5 years working on Hallelujah whereas for example Paul Simon in the case of “A Bridge Over Troubled Waters” sort of heard it all in his head and just wrote it down in less than an hour. Apparently Picasso, especially early in his career, was similarly able to crank out nearly finished works at a prodigious rate whereas Cezanne was never satisfied.
What does Cezanne (and Cohen) vs. Picasso (and Simon) have to do w/ start-ups?
Well, there is a widely shared rule of thumb in the venture community that there are really only 12–15 companies out of several thousand founded a year that really matters. The job of good investors is largely to find those companies and make sure you get accepted into investing in them. This way of thinking largely explains the unicorn phenomena and, let’s be clear, is pretty much correct. Venture investing is a power law business. To pick perhaps the most extreme data point, Facebook alone supposedly returned more than every other investment the long time top tier fund Accel has ever made.
However, what if performance is a lagging indicator and, what is more, “past performance is no guarantee of future performance?”
What if companies that struggle and never give up, all the while investing in their technology and thereby building greater and greater barriers to competition are actually better able to sustain long term success?
We all know of rocket-ships — those companies, like Facebook, that generated so much momentum and showed so much promise early on that they were apparently destined for success. These are the glamor stories.
But, then again, if you look at one of the top Y-Combinator exits of all time (I think second behind Cruise), Twitch, that company emerged from a really tough struggle that included pivoting the company away from JustinTV.
And a company I’m close to, Cloudian, resulted from a pivot away from an already successful company, Gemeni Mobile Technologies and towards built from the ground up S3 / AWS friendly, Cassandra leveraging smart object storage. www.cloudian.com
And other examples abound, including Pandora which is now doing about $1bn in revenues and yet was founded back in 2000 and after spending their initial capital had to beg the team to work for roughly 2 years without pay.
Why might it be the case that companies that iterate their way to genius, like Cezanne, are more able to sustain their advantages versus those that achieve faster growth earlier?
- Innovating in technology takes time. And no one really knows how long it takes before you start.
That second point is key and simply true and almost always ignored. Yes, experienced engineering leaders can do a pretty good job of scoping a solution to a problem if it is similar to something they have done before. However, current approaches to building and operating software — DevOps — really depend upon leveraging agile to discover and address needs massively faster than the discovery heavy approaches of waterfall. In a sense, waterfall approaches to upfront discovery and specification were so worthless and even destructive, that we generally don’t even bother any more.
What’s more is that it may well be that the better you can scope a problem, the less value, from a long term technical differentiation perspective, it gives you. Let that one sink in a little; so if you ask your entrepreneur how long to release X and they say 6 weeks, the quick and exact reply may fit with the mental model of competence you’ve learned as an engineer and professional in highly competitive environments and yet it may be a negative signal of the value of the work being done.
2. Innovating in technology takes time. And enterprises know this, and won’t touch anything < 3.0.
Well, this one is obviously an exaggeration and yet there is some truth to it. I work w/ a number of start-ups that are selling into some of the most important users of information technology in the world. I don’t mean important in the value sense — but in the sense that if their technology fails to perform they could directly impact the operations of “too big to fail” firms. It is no wonder that these buyers remain somewhat conservative.
In the case of Cloudian, their heritage of having first innovated in distributed object storage while serving their carrier customers in their initial business builds confidence on the part of top tier financials and others that the system is going to work, and scale, for years and years as petabytes turn to exabytes.
3. Transparency and community builds trust — and takes time.
Related to the above point, if you want an enterprise like a Netflix or Pearson or Cisco to adopt your solution — as was the case in our second year at StackStorm — and use it to help operate their core business, it often helps to be transparent. While earlier in my career I admit to having closed large deals on Powerpoint alone, these days at least in key pieces of the business, there is exhaustive testing. One way to shave costs and enable self service of POCs and tech reviews is to open source your software and solutions. And this, in many ways, limits your ability to sell only to a vision.
And yet that open source community is itself a flywheel that is hard to replicate. However, it too takes time and, ironically, can limit at least for the first couple of years a start-ups ability to monetize.
So what’s an investor to do?
In my case, where to date I’ve largely invested as the first check invested, I look for the market potential and depth of the team and especially the promise of sustainable technical differentiation as well as, yes, early traction. Those are all positive markers. However what I’m most looking for is someone who is scrappy and stubborn. When I take on a project I know it will succeed — I just don’t know exactly how long or how painful it will be to get there. Starting things for me isn’t a way to make money, or to research for the sake of research, or even to change the world — those are all real and important motivations. Starting things for me is who I am.
These days I’m trying to find ways to filter for entrepreneurs who are like that — who are committed from their very core to building. I agree w/ some comments I’ve heard on the 20 minute VC and elsewhere that one marker may well be the arc of your life or, put somewhat differently, were you born w/ a silver spoon in your mouth, or maybe w/ an extra 10–20% of IQ in your head? And if so, did you sort of slide into your start-up idea having never been stressed and shaken to your core? If so — hmmm, you may well be in for a shock that you simply won’t be able to handle. This perspective must help explain why many immigrants do so well as entrepreneurs — not because it is so easy to immigrate to the United States — but because it is so freaking hard to leave your family and community to follow your dream.
In short, if you over-emphasize momentum and validation as markers of success, then you risk fighting your way into companies that has been inflated based on expectations and that is vulnerable, over time, to competition from those companies that took bigger bets on riskier technologies and that were knocked down, only to get back up as fundamentally different and better solutions to current and emerging markets. You risk bidding up the works of a young Picasso (which can work) and ignoring those of a never satisfied Cezanne.