From Signal to Strategy: How Marketers Can Anticipate Market Shifts

Picture a soccer scoreboard three minutes into the match. Nil-nil. If that scoreboard were your marketing dashboard, it would say nothing had happened. But on the field, players are already moving into space that has not opened yet. The best ones are not watching the ball. They are reading where it is going to be.

The scoreboard tells you what happened. The field tells you what is about to. Most marketing teams are staring at the scoreboard, and I understand why. It is the certain one. The hardest part of prediction is not finding more information. It is separating a meaningful shift from short-term noise, and deciding when the evidence is strong enough to act.

No one sees the future

So who is actually good at that? The answer is humbling. In a multi-year forecasting tournament run for US intelligence, ordinary volunteers working from public information beat professional analysts with classified data by a reported 30%. They did not have better information. They had a better habit. They updated constantly, in small steps, and never made one big call and then defended it.

The same lesson turns up wherever someone keeps score. Microsoft found that only about 1 in 3 ideas it tested improved the metric it was built to move, and nobody could reliably pick the winners in advance. Nor does more data rescue you. Google Flu Trends had more of it than anyone and still overestimated flu in 100 of 108 weeks. The pattern was real in the data and wrong in the world: a correlation with no cause behind it stops predicting the moment conditions change.

Put those together and the conclusion is freeing. No one sees the future. The edge is small bets, honest scoring and constant revision.

AI is a prediction machine. Most teams use it as a content machine

Then AI arrived, and it looked like the crystal ball we had been waiting for. Strip away the hype and AI is a prediction machine. A language model predicts the next word, a recommendation engine the next click, a scoring model the next buyer. That should have made marketers better at reading the field. Instead, most marketers bought a prediction machine and pointed it at production: more assets, faster, with the same decisions underneath. Today 88% of organizations now use AI somewhere, and 39% can show any impact on profit. I call it AI theater. When prediction becomes abundant, judgment becomes the scarce part, and judgment is the step we skipped.

Four tests separate a real shift from noise

Judgment, in practice, is a method. Nobody in B2B is short of signals. The problem is choosing which ones deserve a move, so at Park & Battery we run every signal through four tests. Does it persist across several periods, or is it one spike? Do customer, culture, competitor and category signals converge? Is anyone paying a cost, in money or time, or are they only talking? And can you name the mechanism behind it? That is the test Google Flu failed.

Then you count, because the count tells you how big a move the signal has earned. Zero or one test passed: watch it. Two: probe with synthetic audiences and message tests, which costs days, not dollars. Three: pilot in one segment with real spend. Four: scale.

The tests also protect you from both ways of being wrong. Chasing noise is the mistake everyone fears. Ignoring a real shift is the one that costs you the category. Consider the 94% of B2B buying groups that rank their shortlist before they ever contact a seller. Nothing in the dashboard flags that, because the dashboard starts counting at the form fill. The fix for both is the same: write the prediction down, attach a probability, score it later. Misses are training data.

The new operating system

Work this way for long enough and your operating system changes. The annual plan as a fixed document is over. What replaces it is a thesis about where your market is going, a set of bets sized to the evidence, and a loop that never stops: sense, simulate, launch, learn. Launch stops being the finish line and becomes the first data point.

Tori Young, who leads global brand strategy and marketing at Thermo Fisher Scientific, has the best name for this I have heard. Her team lives by the Wayne Gretzky line about skating to where the puck is going, and alongside every QBR they want a QBA, a quarterly business anticipation. The review tells you the score. The anticipation tells you where to skate.

Which brings us back to the field. Pick one decision due in the next 90 days. Write down what you think will happen and how confident you are. Run the four tests on the signal behind it. Then score it when the result comes in.

The best teams do not wait for the game to come to them. They read the field.