Andrew Chen Archives

One of the more interesting lessons I’ve learned since joining the startup world has been the widely differing perspectives on failure and innovation. It’s a tremendous source of tension primarily built on a simple stylistic tension, which seems to occur in many different companies.

On one side, you have the free-wheeling Innovators, who instinctually move from one iteration of a project to another. Their goal is to throw up lots of ideas, see which ones stick, and build on those. Of course, this process continues indefinitely, piling failure upon failure to generate the successes. The attitude here is pretty much, "We don’t know what will succeed, so let’s do something simple, learn, and go from there." This is most commonly part of the Startup-Guy archetype, someone who is constantly on the move and shooting from the hip.

On the other side, you have the Analysis guys. Their goal is to make sure that everything’s buttoned up, that decisions are made for the right reasons, and that you have a repeatable process for success. Because most large companies have "cash cows" to protect and grow, these folks are very useful to scaling out a process in order to generate large-volume, low-risk, and high-margin products. They often have a systematic way of approaching issues, and rely more on logic and analysis than on instinct.

For me personally, I lean more towards the top (the no-holds barred Innovator) than the Analysis guy. There’s obviously lots of gray area, but I definitely tends towards one more than the other. Sometimes, when the going gets tough, I can bristle when interacting with Analysis folks. They can seem unimaginative and overly focused on process – but most importantly, their focus on making things repeatable (and big) may make something shitty into something shitty AND big. This may be what happens when you stick a bunch of these people on an unproven idea – rather than focus on verifying assumptions, testing hypotheses and coming up with new solutions, instead they want to focus on scaling up the unproven idea. Often, this leads to disasterous results. Grrr!

That said, I now realize that there’s a lot of gray area, and one is not mutually exclusive of the other.

For example, take the Big Pharma drug discovery process. The way it works is, you stick 100,000 (or whatever) possible drug candidates on one end of the "pipeline," and screen them out for all sorts of different things. Does it seem like it would work? Does it seem to attach to the right "targets?" Is it toxic to animals? Is it toxic to humans? Etc. Then at the very end, after 99,999 failures, 1 drug comes out that is approved by the FDA which makes a billion dollars a day. When rolled up, all the failure costs upwards of $800MM to produce one drug.

Now, from the outside, if the Analysis people didn’t understand this process, they’d be pretty worried. After all, they’d see drug candidate after drug candidate (99,999 in all) fail over and over. It might be depressing if you didn’t know that the entire process is only meant to produce 1 good drug, and that the payouts at the end will more than make up for all the costs.

But similarly, the pure Innovator types need to understand that they are, ultimately, making a drug. And that they probably need to make 99,999 failures to make 1 success. And that there are X number of steps to go through to get to that success rate. That way, they are hitting "innovation" metrics on the way, to ensure a deterministic march towards the billion-dollar drug. If they just focused on one thing, and morphed it quickly, they might not have enough people, drug permutations, or resources to get to the success.

So ultimately, I’ve come to believe that the startup types need to think ahead and plan, at least roughly, against the major metrics of success. For startup types, that might be a rough schedule or some ideas on how many prototypes need to be generated to be successful. Or how many users need to join in order to pronounce the user mechanics as "fully baked." So for them, if they are comfortable with micro-levels of failure, they need to keep their eye on macro-level success, and what that entails.

And for the big-time Analysis types, they need to understand how failure plays a role into their overall plan. Just because you are scaling something out, and you are hiring a big team, doesn’t mean you’ve overcome the fundamental issues in your business model. Some of that requires change, failure, and innovation to drive it forward.

The most important part is for these two groups to communicate and share a common language. Just as the drug pipeline process has been able to tie innovation with expected failure rate, dates, deliverables, and financial metrics, entrepreneurs need to be able to talk about their approach towards innovation and how that ties into ultimate success. Without it, you’ll have big companies with shaky business models, or small companies with no strategic direction.

I write a high-quality, weekly newsletter covering what's happening in Silicon Valley, focused on startups, marketing, and mobile.