Does Data Really Matter To Your Startup? Definitely, Maybe.
There’s been a huge amount of conversation over the past few years about data-driven startups, the power of great analytics and how to use data as a tool to make better decisions. So much so that a whole new crop of incredibly expensive and, apparently, incredibly smart people have appeared in Silicon Valley. I’ve even heard whispers that, as the owner of a startup, the next most important hire your company will make is the Head of Analytics, noting any founding team without such minds is lacking. It’s probably just a matter of time before we see our first “Analytics Fund” invest in companies that have this DNA at their core.
These companies dominate with analytics. Good for them.
Amazon is the role model, but just as impressive are Zynga and Groupon, who moved from non-existent to multi-billion dollar companies largely on the backs of their analytics systems. This is pretty impressive. Their ability to test different messaging and see in real-time how engagement is trending, responding accordingly, is mind-blowing. Plus, venture capitalists love analytics: after all, if you can measure it, you can manage it.
Uber is another, albeit earlier stage, famously data-driven company that talks almost incessantly about their “math team.” They make beautiful maps of traffic and users. It’s pretty impressive and really does seem to drive their business in hugely important ways.
OKCupid has an amazing blog on the topic as well, and they clearly practice what they preach.
For these companies, it seems data analytics may really be a significant differentiator.
But is this the right thing for all startups? After talking to some of the smartest analytical minds out there, I have a hard time buying that. I think the power of the herd effect, especially in Silicon Valley, is more often than not dangerous for first time entrepreneurs who think that “because [fill in the name of any hot company] does this, we better do it, too.” However, there is too large of a positive selection bias for that to be true when it comes to analytics. The number of “analytics-driven” companies that have failed could be in the thousands, although you’d never know… they just faded off into the distance.
Is it important to understand the key things that drive your business? Obviously. But that’s different than running your company on analytics alone; I don’t hear nearly enough about when and how to apply analytics with an appreciation for stage and type of company.
At Zaarly, we use Kissmetrics and MixPanel (both are awesome), and are in the process of evaluating some other very powerful tools. But does it really matter to our business? At this point, I’d say the answer to that is a definite maybe.
Data is great for optimization, but not invention.
Although it’s nice to know how things are going, very rarely does data tell us something our users haven’t already. This is partly due to the stage of our company; we are still pre-product/market fit, which means we have yet to invent some of the things that will end up making our business truly succeed. It’s impossibly difficult to imagine finding any dataset – even depicted in the most beautiful way – that would inform the new things we are currently building. However, data does act as an important guide when deciding where to spend time and energy when optimizing things we’ve already built. After we’ve user tested a new idea, prototyped it and built it, if it doesn’t quite have the impact we expect, our data usually highlights a step we missed. Sometimes it’s something that seemed to work in a small group but ended up behaving differently with hundreds of thousands of users. More often than that, it’s the one thing we assumed we knew how to handle and, therefore, didn’t bother putting in front of “real” people. Usually, it’s easy enough to address. But this is in the zone of optimization, not invention.
So, with the hype about analytics that is currently gripping the startup world, don’t forget this is just one tool in a very deep toolbox. The cool kids may be talking about it, but it’s our job to choose the right tool for our company’s stage in the startup lifecycle. For Zaarly, at this particular moment, user feedback and testing beats data every single day of the week. My guess is that, in six months, this will be much more core to us, and I’ll be happy we have the tools and systems in place. But today, analytics are a nice-to-have.
If you relied on analytics exclusively, you’d probably never start a company or work for a startup… the data doesn’t show it’s a smart decision. But that’s not why we’re doing this.
Bo Fishback is the Founder and CEO of Zaarly. Bo was previously the vice president of entrepreneurship for the Ewing Marion Kauffman Foundation and the president of Kauffman Labs for Enterprise Creation.