KWL
← Field Notes

Methods · Field Note 0004 · Jun 2026 · 6 min · KWL Research

To compare eleven markets, you need one ruler.

Raw signals cannot be compared across countries. Different chart sizes, languages and brand footprints distort every measure. This is how we put a number in Spain on the same footing as a number in Sweden.

From raw signals to one comparable read
Raw signals eleven markets, not comparable as they are
Position-weight a No.1 counts for more than a No.50
Year-on-year compare like quarter to like quarter
De-mean across markets strip out what is moving everywhere
Unified cultural origin home-grown means the same thing everywhere
Comparable behavioural read a number in Spain reads like a number in Sweden

The problem

You cannot compare markets on raw numbers. A bigger market has a bigger chart. A campaign lifts a brand everywhere at once. Home-grown means one thing in France and another in Ireland.

Lay one market’s raw figures next to another’s and you are comparing rulers, not markets.

It is a commercial problem, not only a statistical one. Without comparable measures, France always looks bigger than Belgium and Germany always looks louder than Portugal. Comparability is what lets you see a behavioural difference instead of a difference in scale.

Four steps to one footing

How a signal is made comparable
StepWhat it does
Position-weightingA No.1 song tells us more about a market than its No.50. We weight every chart entry by its rank.
Year-on-yearSome quarters always run hotter than others. We compare each quarter with the same quarter a year before, so seasonality is not read as a trend.
De-meaning across marketsIf a brand rises everywhere at once, that is a campaign, not a country behaving differently. We remove what moved everywhere and keep only what was unique to the market.
Unified cultural originHome-grown must mean the same thing everywhere. We resolve origin from ISRC, citizenship and genre, one method for all eleven markets.

De-meaning, worked

The subtlest step is the most important. Suppose a brand is up around forty percent in search in every market at once. Raw, that looks like momentum everywhere. But if it is rising the same amount everywhere, it is one campaign, not eleven local stories.

Illustrative: a brand up ~40% across Europe
MarketRaw moveAfter de-meaning
Germany+40+1
France+390
Spain+41+2
Italy+38−1

The Europe-wide average is about +39. Strip it out and the residual is almost nothing: no market is doing anything distinctive. The raw forty was the campaign. The de-meaned figure is the truth.

Subtract what is moving everywhere, and what remains is the market.

Why it is worth the trouble

One comparable measure is what makes everything else possible. The cross-market control room, the Four Europes, the two-year reconstruction: none of it works if a number in one market does not mean the same as a number in another.

Without one ruler you have eleven reports. With it you have one continent, read side by side.

Where we are honest

Every normalisation is a choice. Different choices produce different numbers. That is why we publish ours, and why every figure names its public source and date. The ruler is open to inspection, which is the only way a comparison earns trust.

Method and sources

Sources
BehaviourSource
MoodSpotify chart audio
Search and intentGoogle Trends
AttentionWikipedia
News toneGDELT
MacroEurostat

A read, not a forecast. The method is comparable across markets and quarters by design, which is the whole point of building one ruler.