The Signal Audit: a 90-minute exercise to find what actually matters

Every business we meet has the same relationship with data: too much of it, too little certainty.

The dashboards are healthy. The decisions are not. Somewhere between the two, the information that would actually change a decision is buried under information that merely describes the past.

The Signal Audit is the exercise we run before any intelligence or strategy work. It takes about 90 minutes with the right people in the room. It doesn’t require new tooling. It does require honesty.

The premise

Not all data is equal. Most of what a business collects falls into one of three categories:

1. Comfort data. Metrics that exist because they have always been reported. They are read, nodded at, and never acted on. Page views. Total followers. Impressions. Comfort data makes meetings feel evidence-based.

2. Lagging data. Metrics that tell you what already happened — revenue, churn, closed-won. Necessary, but by the time they move, the causes are weeks or months behind you.

3. Signal. Information that changes a decision before it is made. A prospect asking a new question repeatedly. A competitor changing pricing structure. An enquiry pattern shifting from one segment to another. Signal is rare, uncomfortable, and disproportionately valuable.

The audit’s goal is brutal sorting: nearly everything falls into category one or two, and the room has to name what is actually in category three.

The three questions

Question one: What decision are we avoiding?

Every business has one. The pricing review that keeps slipping. The channel that everyone suspects is wasted but nobody will kill. The segment that underperforms but carries emotional weight. Name it explicitly. Signal always relates to a real decision — if you can’t name the decision, you’re not looking for signal, you’re looking for reassurance.

Question two: What would change our mind this quarter?

For the decision on the table, describe the evidence that would genuinely alter the choice. Not confirm it — alter it. Most teams discover they have never defined what disconfirming evidence would look like, which means no amount of data can ever change their course.

Question three: Who sees this first?

Signal decays fast. If the person closest to the customer — sales, support, delivery — sees something in week one and leadership hears about it in week six, the signal has expired. The audit ends by naming where signal gets captured and how fast it travels upward.

What the audit produces

Three things, and nothing more:

• One named decision that has been avoided, now scheduled
• A short list of disconfirming evidence worth watching — the actual signals
• One change to how signal travels from the edge of the business to the decision-makers

That’s it. The output is not a report. It is a small number of commitments about what the business will pay attention to.

Why it works

Because it inverts the usual question. Instead of “what data do we have?” it asks “what would change our mind?” — and then checks whether anyone is watching for it.

Businesses don’t lose to competitors with more data. They lose to competitors who notice things earlier and act on them. The gap is never collection. The gap is attention.

If your team runs this exercise and the third question exposes a six-week delay between noticing and deciding, you have found the first real problem worth solving. That is usually where an intelligence practice begins.