The idea in one line. Signal strength is how much a signal can be trusted as a predictor — judged by three things: independence of sources, repetition over time, and spread across audiences. Strength decides the action: watch, test, or move.
Why it matters. A researcher who cannot rank signals treats every observation as equally important — and drowns. Strength turns a pile of observations into a priority list: what deserves budget now, what deserves a small test, and what only deserves a calendar reminder.
The three dials. Independence: how many UNRELATED sources show it? One blogger is one source; ten reposts of that blogger are still one source. Repetition: does it keep appearing over weeks and months, or did it spike once? Spread: is it crossing borders — cities, age groups, price tiers, platforms? A signal that jumps from street style to mass-market drops has crossed the most important border of all.
The action scale. Weak (one source, one appearance, one bubble): watch — note it, set a check-back date. Medium (several independent sources, repeats, still one segment): test small — a capsule, a content probe, a limited buy. Strong (independent sources, sustained repetition, crossing segments): act — this is no longer early information, and the window is closing.
Example. "Slow fashion" as a signal ladder: weak — one Copenhagen designer labels a capsule "made slowly" and gets one article; medium — five brands in three countries use the language and the theme appears at a trade fair; strong — mass-market players adopt the vocabulary, EU regulation pressure appears, consumer surveys register shifted priorities.
Common mistake. Counting echoes as confirmations. Ten TikToks quoting the same runway clip are ONE signal with one passport — loud, not strong. Independence is measured at the origin, never at the repost.
Check yourself. Take one signal from your board and score the three dials in one sentence each. If you cannot name a second independent origin — it is weak, whatever its view count.