§ Field Notes · a behavioural research library
Field Notes.
Not a blog. A growing body of disciplined research on how to read European markets through behaviour: what the continent is doing, how brands sit against each other inside it, and what the evidence can and cannot bear. Every figure names its source.
11 of 16 notes carry a dated correction. 2 had their central claim withdrawn and are kept in the archive. We print those because a research library that never corrects itself is not being checked.
Market intelligence · Note 0017 · Oct 2026 · 4 min
Italy is the only market where more nights meant fewer visitors.
Ten European markets published accommodation nights and arrivals moving the same way this quarter. Italy published them moving apart — and Ireland is the only market where nights fell at all.
- 11/11
- markets publishing a realised travel finding
- 1
- where nights and arrivals point apart
- 1
- where nights fell at all
- +0.18
- nights, Italy’s change in average stay
Market intelligence
How European markets are behaving this quarter, and how they differ.
0016 Which European markets actually moved this quarter. Eight of eleven are where they were. Three are not, and one of them moved the wrong way. The argument for reading a market against its own history rather than a single quarter, in one grid. 0012 Four European markets feel worse than they intend to behave. In four markets consumers report below-normal confidence and above-normal willingness to make a large purchase. Both are reported intentions, measured by the same survey, and the gap between them is where headline-keyed briefs go wrong. 0011 Cost of living is the top story in three of eleven markets. It was the only story. Europe’s shared economic conversation has broken into six different ones. The brands still briefing against “the squeeze” are briefing against last year. 0010 Four European markets have almost no artists of their own in their own top 50. We only found this by splitting a metric we had been publishing for months. The number we were calling home-grown share was measuring something else. 0007 France and Germany share a position. One of them only just arrived. Both markets read below par and deferring in Q3 2026. Germany has been there a while; France moved into it this quarter, one of only three European markets to change position at all. Same label, opposite directions of travel.
Competitive intelligence
Brand-level reads: who is where, in which market, against whom.
0014 A challenger sits above two of France’s largest banks in the app chart. One snapshot of the French finance chart, taken after the quarter closed. Seven brands Cadence observes appear in it, and the order is not the one a balance-sheet ranking would give you. It is an order, and nothing more than an order. 0013 The discounter doesn’t always win. Seven European markets published a named riser and a named faller in grocery search attention this quarter. In one of them the gainer is a discounter. The pattern a pan-European plan assumes is not the pattern the measurements show.
Research & methods
How the measure is built, and the discipline about where the evidence stops.
0018 We tested whether intentions predict spending. They did not. Big-ticket purchase intent is the closest thing to a stated version of what people actually buy. We pre-registered whether it forecasts realised retail volume, ran it over 120 months in two markets, and it added nothing. Here is the result, and here is what it does not prove. 0015 Nobody can tell you what European inflation did this quarter. Eurostat closed the series in December and has not published its replacement. Cadence withholds inflation in all eleven markets rather than print a nine-month-old figure as current — and so, quietly, does every read that still quotes one. 0006 One quarter, three desks, three different answers. The same France edition answers a retail, a finance and a travel question differently, and what each desk receives is the clearest demonstration of what Cadence is actually selling. 0005 We rebuilt nine quarters of Europe on one engine. Europe moved more than we said. Two years ago this reconstruction reported stability. Rebuilt on the measures Cadence publishes now, a third of the quarter-to-quarter transitions are changes. The method held; the finding did not, and the two counts are not comparable. 0004 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. 0003 We backtested our own product five times. Every test failed. We pre-registered what the music signal can carry, tested it five ways, and it failed five times. We published every result, including the failures. Negative results are still results. 0002 We read what a market says and what it does. The gap is the finding. Cadence reads both instruments, because they answer different questions and are not always in the same place. A stated measure gives you a level; a behavioural one gives you a rate of change. The useful quarter is the one where they part company. 0001 The paper that started Cadence. And the five tests that closed the question. We did not invent this idea. We found robust science, tested it ourselves, learned where it breaks, and built a product that stays inside the evidence.
Archive
Notes whose central claim has since been withdrawn. Kept whole, with the correction above the piece, because a record of what we published is worth more than a tidy library.
0009 The neobank wave breaks differently in every market. Here’s where it wins, stalls and recedes. Per-market comparisons withdrawn: the figures did not reproduce at publication precision. 0008 Europe is trading down. Not by the same amount. Central claim withdrawn: brand-level search movement is not published in Q3 2026.