In his article “Strategy Under Uncertainty,” Jerry Neumann contrasts the traditional Porter model of business strategy with one more suited to startups, the former being modeled around mature organizations operating in known competitive spaces, the latter around startups moving in opaque environments with higher uncertainty and more moving parts.

In the piece he defines “strategy” as a framework for “how to make decisions in situations that are not yet known.” To have a purposeful, intentional approach to an objective, whether in war, sports, or business, you have to formulate a model for predicting the future. Without clarifying some future state you think will develop, it’s impossible to take the right series of actions to end in success.

Strategy in some businesses is all anyone wants to talk about. In others it’s taken for granted or not discussed enough explicitly. Each of these paths is incorrect.

In the second case, most anyone would acknowledge that little meaningful progress can happen in a business with no strategy. You’re merely subject to the whims of the environment and other players around you. You rely on getting lucky.

The first case, though, is also surprisingly common. In the modern business universe, being a “strategic thinker” is an accolade. The business literature surrounding us pushes this idea so much that in some environments, everyone wants to strategize. Without the right guidance and guardrails, this can go on far too long and can be actively detrimental to execution.

One of the dangers I see of over-strategizing in startup environments is that it tempts you into searching for information that may not be there in the first place. We think we can “map the competitive landscape”, but what if we can’t lay out an immediately clear picture? We begin to prioritize having a drawable diagram of our competitive landscape over the actual reality on the ground. In pursuit of having a legible map, we stretch definitions or cut corners. The map becomes the goal.

As Neumann points out, the environment in startups is riven with uncertainty and moving targets. You have more unstable dynamics than you’d find in many mature, large companies:

Startups operate as part of a complex system that encompasses not just their internal operations, but their customers, their suppliers, other companies that might compete or cooperate with them, financiers, the media, the government, and society at large. Each of these other entities also makes decisions, and the results of their decisions must factor into the startup’s decision model. The changes most likely to affect a startup are the ones that happen as a result of the decisions the startup itself makes, a complex feedback loop.

One of my core beliefs is that in spaces of high uncertainty, too much analysis and planning builds in more risk rather than less. Partially this is because of the time we spend in planning; more time spent equates to higher expectations and an inflated sense of what we know. I’ve analogized this situation to a high-wire act: the more assumption you make that your Big Prediction is correct, and the more you build toward that up front, the higher you raise that wire over the ground. As your predictions go farther into the future and your bets get bigger, the risk keeps rising.

Building strategy is so tempting. It always sounds like a good thing to spend time on, and it often is. Somewhat perniciously, it can be even more tempting in a startup environment, where the risk is high and the runway short. You perceive little margin for error, so you have to get the plan right.

The trouble is that there’s a limit to how much it can do for you in a startup. And the more novel your idea the higher the uncertainty as to what the future of the market holds. Bias creeps in about what you think you know about the predictability of the space; just because you’ve done n hours of analysis doesn’t guarantee you have any clearer an answer than at hour two. But human biases will tempt you to believe you’ve imbued it with more legibility than there really is. The only worse decision making environment than one with no information is one with actively misleading information.

Like many things, I believe the right way to approach strategy in startups is a nuanced middle.

You should always start with a clear vision of the future. What world is your business or product trying to create? How are you changing life for your customers or users? The answer to this question sets the course, but not the strategic roadmap, go-to-market, or many other things you need to figure out.

I like to think of strategy in medium term chunks. The goal is to build a hypothesis that we can push forward, rollout a strategy for, and test with real feedback on the order of weeks or months. More frequent, lower intensity strategy sessions that allow you to come up for air and work with real information about the world that you’ve learned, rather than speculating on your 5 year strategy at a whiteboard for a month.

Thinking you know more than you do leads to dangerous and risky plans. Usually it’s possible to know enough to take smaller, incremental steps based around much more reliable signal, with less interference from bias. There’s typically room for a couple of riskier moves here and there where you could earn outsized returns if you’re right. But the majority of the time, a tempered approach to just-right “Goldilocks” strategizing is the right way to go for a small team and a new product.