In this post:

Data Driven Cultures in startups should be better at prospection than other cultures.

Data Driven Cultures

Carl Anderson, 2015 (Data Scientist at Warby Parker) defines a data driven culture as:

  • Is continuously testing;
  • Has a continuous improvement mindset;
  • Is involved in predictive modeling and model improvement;
  • Chooses among actions using a suite of weighted variables;
  • Has a culture where decision makers take notice of key findings, trust them, and act upon them;
  • Uses data to help inform and influence strategy.


There is tremendous variation in how people think about the future. There’s a lot of variation in how people think about how people think about the future.

If I were to use a very strong magnet to interfere with your DLPFC, your Dorsolateral Prefrontal Cortex, I’d mess with your ability to think about the future.


Somewhere in that tangle of blood, cholesterol, and neurons there’s a neural network. It enables you to think about the future.

It’s a shame that the DLPFC is supposed to be responsible for the regulation of emotion. It’s too messy in there.

Better at Prospection

The brain is a hungry organ.

The planning function in an organization is also a hungry organ.

This post would be done if you drew a straight line from data to knowledge, and knowledge straight to prospection. It’s an axiom in the world of Data Driven Decision making (DDD) that this is always the case. A data driven culture is engaged in predictive modelling, so it ought to go without saying that it’s always the case that it is better at prospection.

Data Driven Cultures are engaged in forward looking modelling, and since reasoned thinking and testable hypotheses are superior to hockeystick hopes, they’re superior at prospection. That’s it. The argument is over. Stop reading.

But that may not always be the case.

Those engaged in the standard act of model improvement may be intensely resistant of testing an adequate volume of what-if’s.

Worse, their own neural networks, deep in their DLPFC’s, may not be in a state to imagine wide enough.

Investment in a paradigm blinds

Not all findings are treated equal, even among scientists. Kuhn (1962) observed that the findings that didn’t fit into an established scientific paradigm were frequently disregarded. It was taught to me as a puzzle analogy.

Imagine a huge puzzle on a table. And you’re standing there with a group of scientists. You start categorizing the pieces. Others will start putting it together following some theory. “It’s a horse!”, or “It’s a picture of a castle!”. Under the castle paradigm, it might be possible to get pretty far before a bunch of pieces that don’t quite fit start to turn up. These tend to be, at best, put off to the side in an odds and ends pile until it reaches a critical mass. Most of the time, if you’re in a competition with others framed by the theory that it’s a castle, there’s less incentive to wonder what six pieces that look like a hoof are about.

Founders are especially vulnerable to convenient reasoning and anchoring. (And in part, they have to be!)

Hope, an emotion, is regulated in roughly the same tangle of nerves that prospection is. It might also be really close to where rage is felt as well.

Let’s say that you have a fairly large investment riding on the puzzle turning out a certain way. How receptive do you think you’re going to be when members of your own team – people you pay for their expertise – to start to tell you about pieces of the puzzle that don’t add up.

Founder: “It’s a castle!”

Underling: “Dunno Boss, it might be a horse!”

Founder: (Thunderous roar) “IT’S A CASTLE!”

It has to be a castle because an investment is riding on it. Not just of money, but of feelings, ego, and pent up preferences about the future. Moreover, the founder might not know what to do with the finding that it’s a horse.

Worse, pointing out that it’s a horse may be interpreted as a form of obstructionism. It may be viewed as an attempt to destroy morale.

This has little to do with trust and everything to do with the tendency to anchor.

Investment in a paradigm blinds.

And it’s an equal opportunity blinder.

Data Driven Cultures in startups should be better at prospection than other cultures.

Taken on its own, an involvement in predictive modeling and model improvement does not guarantee a sustainable competitive advantage at prospection. The same forces that cause paradigm lock-in in scientific communities can also happen to communities of data scientists and especially their founders.

All too often, many reach into the past record to argue against particular what-if’s. There’s no evidence in the historical record to suggest that a particular scenario is likely, so why even entertain the scenario in the first place. This is a major source of friction.

You don’t landing-page optimize your way to a billion dollars in revenue alone. You don’t cost cut to a billion either. But, you don’t get there by being blindingly hopeful without reason, or generally, pure dumb luck.

The friction doesn’t have to exist.

When combined with strategy formation, a Data Driven Culture is likely to be much, much better than alternative cultures at prospection.

Strategy formation, as defined in the data driven culture paradigm, is the deliberate selection of choices amongst weighted variables. If I tweak Carl Anderson’s definition by just a little bit, and argue that strategy formation involves choice amongst alternatives and likelihoods about the future, a bunch of breakthroughs are realized.

For one, the search for alternatives is more likely to be treated as a routine scientific course of action. A recurrent theme in James March’s work is that the search for alternatives is driven by dissatisfaction. If-it-aint-broke-dont-fixit-ism is a common attention conservation pattern in North America. Dissatisfaction drives search. That seems pretty dangerous to me.

Instead of search-for-alternative actions being driven by the emotion of dissatisfaction, it is driven by routine or habit. The behavior of alternative-seeking, when habitual, is disconnected from the roller coaster of feeling, and it becomes consistent. Consistency is a key competitive advantage in most industries, so, it may very well be an advantage in this context. Consistent search, powered by behavior instead of emotion, is likely to generate more reliable performance.

And for two, when choices are weighted against future expectations, a lot more attention is going to be paid to why those expectations exist. Expressed slightly different, hope-based argumentation is unlikely to persist in a truly data driven culture. It’s less likely you can get away with arguments in the format “something something hockeystick hooray.”

If a strategy is derived from reason, and constructed with measures from reality, the culture is likely to learn to make complex decisions more reliably over time.

When combined, better search and better choice are likely to generate better strategies. Just getting better at generating better strategy should cause an incentive for better prospection that is strong enough to overcome the emotional blinders and lock-in.

For those reasons:

Data Driven Cultures in startups should be better at prospection than other cultures

Posts in this series include:

The Data Driven Culture

The Data Driven Culture: Strategy

The Data Driven Culture: Startups