The Data Rush

Today “data” is all the rage. The Big Data Buildout is on, and the “data network effect” is the new competitive moat. First one to the data wins.

Maybe.

Or maybe we’re just in transition from times of data poverty to data abundance. It could be that we’re so unused to data that we think it’s all worth gold. Once we’re done instrumenting, collecting, analyzing and dashboarding, we might realize only a few metrics mattered at all.

Data used to be difficult to collect, to store, to analyze and to operationalize. And now those things are getting easier: a combination of technological innovation and fad-driven (but useful) management strategies are bringing data to the front. We are asymptotically approaching knowing everything that can be measured by a sensor or input into a spreadsheet.

But data is a mid-term advantage. In the long run, in a world of increasingly perfect information, data will be worth everything, and simultaneously, worthless.

That is to say, we should expect the world to be profoundly shaped by data and yet few companies will be known as data companies, just we are now losing the “e-” in front of commerce.

If anyone tells you that their advantage is “data”, start asking questions. Because over and over we’re learning that data is often less proprietary than we first thought, and less useful as a long-term advantage.

A few non-representative examples that may each be worth their own post:

  • For most products (on Yelp, Amazon, Uber, etc.), the value of many reviews are rarely defensible and less useful than one or two well-chosen reviews.
  • Quantified Self products (Fitbit, etc) are great novelties left in drawers, with one simple question providing most value: have I moved enough yet today?
  • The consumer Internet of Things space has ended up being kind of boring and introduced new modes of failure. The true value of the “connected toaster” (or thermostat, or door, or lightbulb, etc.) has not been discovered.
  • The navigational data advantage of Google is being challenged on all sides, by Tesla, Apple, HERE, OpenStreetMaps and others. How much more useful are the highest fidelities, and at what cost?
  • Even in the many machine learning races going on, I think (but don’t yet know) that vast reams of data will end up being a less useful competitive advantage than expected. On the consumer side, clever UI trade-offs that mask failings while being focused on very targeted use cases will do just fine.

As a society, we should expect great returns from data. But the data won’t be valuable because it is Big, but because it is Just Right. Data will be everywhere, but like gold, not every vein will make you rich.

To receive future posts via email, sign up here.