Most Decisions Don't Need Better Data

This is the post I've been most hesitant to write, because it cuts against the thing I've spent years doing professionally and am now building a company around.

But I think it's true, and I think the industry would be healthier for saying it plainly.

Most decisions don't need better data. They need the courage, the clarity, or the organisational alignment to act on what's already known.

---

The pattern

There's a pattern I've seen enough times that I've stopped being surprised by it.

An analytics team spends weeks on a piece of work. The model is sophisticated. The segmentation is nuanced. The insight is real. It gets presented to leadership. Leadership nods, says it's interesting, and then makes a decision based on something the CEO said in a hallway conversation three months ago.

Or the reverse: a decision that should have been made a year ago finally gets made, and the justification is a number so basic — monthly active users fell for four consecutive months, average order value is down — that any analyst in the building could have pulled it in ten minutes. Nobody needed a model. They needed permission.

The bottleneck was never the data.

---

What actually blocks decisions

When decisions fail to get made — or get made badly — the cause is almost always one of a short list of things, and "insufficient data" is rarely at the top.

Organisational inertia. The right answer is known but changing course requires disrupting existing plans, upsetting stakeholders, or admitting that a previous decision was wrong. No amount of better data overcomes this. The data was never the objection.

Misaligned incentives. The person who needs to act on the insight is not the person who benefits from acting on it. Or the person who benefits is not the person with the authority. Analytics can surface this misalignment but it can't fix it.

Risk aversion without accountability. It's professionally safe to ask for more data. It's professionally risky to make a call that might be wrong. Requesting another analysis buys time and diffuses responsibility. Analysts are sometimes unwitting participants in this delay — producing another piece of work when what's actually being asked for is someone to make a decision.

The insight isn't trusted. Not because it's wrong, but because the person receiving it doesn't understand how it was produced, doesn't have context on its limitations, or simply doesn't trust the team that built it. This is a relationship and communication problem, not a data problem.

---

The uncomfortable truth about most business decisions

Operational decisions — the vast majority of what organisations actually decide — turn on a relatively small number of simple, well-understood metrics. Revenue up or down. Customer retention improving or declining. Costs in or out of control. Product engagement growing or shrinking.

These numbers are not hard to produce. Most organisations of any maturity have access to them. The question of what to do about them is rarely answered by more granular analysis. It's answered by leadership, judgement, and the willingness to move.

The sophisticated analytical work — the models, the segmentation, the causal inference — earns its place at the margins. It helps with genuinely hard questions where intuition fails. It catches things the simple metrics miss. It occasionally produces a genuinely counterintuitive insight that changes strategy.

But the ratio of sophisticated analysis produced to sophisticated analysis that actually influenced a decision is, in most organisations, not flattering.

---

What analytics is actually for

I don't think this means analytics teams are wasted. I think it means we've been slightly wrong about what the value is.

The value of good analytics work is less often the novel insight and more often the confidence it gives people to act on what they already suspected. A leader who thinks customer retention is the core problem doesn't need a model to tell them that. They need a clear, trustworthy number — ideally with a trend — that lets them stand in a room and say: this is the problem, here's the evidence, here's what we're doing about it.

That's a different job than finding the insight. It's closer to producing the argument. The analysis isn't the discovery, it's the justification for the decision that common sense already suggested.

This sounds deflationary. I actually think it's liberating.

If your job as an analyst is to produce a novel finding that nobody else could have found, you're setting yourself up to fail most of the time. If your job is to give decision-makers the clarity and confidence to act, that's achievable on almost every project. And it's genuinely valuable — organisations that act decisively on clear evidence consistently outperform ones that wait for certainty.

---

What this means for how we work

If you accept this framing, a few things follow.

Simplicity is a feature, not a compromise. A clear, well-framed single number that a decision-maker trusts is worth more than a sophisticated model they don't understand. Before reaching for complexity, ask whether the simple version already answers the question.

Communication is not secondary to analysis. It is at least coequal with it. The analysis that never leaves your laptop is worth nothing. The simple analysis that arrives clearly framed, at the right moment, in a form the decision-maker can act on is worth a great deal.

The question "will this change a decision?" should come first. Not "is this technically interesting?" or "is this methodologically rigorous?" Those things matter — but they matter downstream of whether the work will actually be used.

Relationships and trust are load-bearing. If the person receiving your analysis doesn't trust you or understand how you work, they won't act on it regardless of quality. The best analysts I've known invested as much in those relationships as in the technical work.

---

The honest version of this

I've spent years building analytical capability and I'm now building a tool for analysts. I have obvious reasons to believe that analytics work matters.

I still believe that. But I think the industry — including me, at various points — has overstated the case that better data or more sophisticated analysis is the answer to the question of why organisations don't act on evidence.

The answer to that question is more human than technical. And the parts of analytics work that are genuinely most valuable are the ones that meet humans where they are: clear, trustworthy, well-framed, delivered at the right moment.

That's what good publishing does. Not a coincidence that it's what I'm working on.

---

Max is the founder of Infigured. He spent years doing analytics at a football club and has strong opinions about what actually makes analytical work valuable. Sign up for updates, or if this triggered a disagreement, he'd genuinely like to hear it.