Field Notes · Note 002 · Market intelligence
How to read a market
that moves fast.
A monthly ranking tells you who is ahead today. It does not tell you who is accelerating, who is falling behind, or what moves the lines. A market that changes every week needs another way of looking. A concrete example: beauty social commerce on TikTok Shop, six countries, five months.
The question
How do you follow a fast-moving market without redoing a full study for every new question?
What we observed
Of observed revenue made by the first 10 shops in the French beauty top 200.
The growth of a skincare brand in France, from March to June. Invisible in a frozen ranking.
Creators tied to the seller: a small share of relationships, a quarter of creator revenue.
The US market against observed continental Europe, January to June 2026.
Estimated GMV amounts (Kalodata source), on the observed scope. Descriptive findings, not universal laws.
The data
The ground covers France, the UK, Germany, Spain, Italy and the US, from January to June 2026:
- 95,955 observations at month × country × category × shop level, for 29,728 distinct shops and 28 categories;
- a detailed panel of the top 200 beauty shops in France: 1,397 products, 1,950 creator × shop relationships, 1,952 videos and 1,903 lives.
Amounts are revenue estimates (GMV), not accounting figures. Every figure stays tied to its scope and its period.
The method
Four rules turn a one-off study into a continuous reading:
- Keep historyKeep every monthly snapshot instead of replacing the previous one. The sequence of snapshots is what shows a trajectory.
- SeparateDo not confuse observations with entities: the same shop appears across several months and categories.
- Drill downRead the market, then the category, the shop, the product, the creator and finally the commercial lever. Each level answers a different question.
- DefineWrite down denominators and scopes. A figure is not information until you know what it is relative to.
What the reading shows
Share of observed revenue, top 200 beauty shops in France
1. The market is highly concentrated
A few shops make most of the revenue. Following the top 50 closely, and the rest by category, is enough to see most movements.
2. Trajectories matter more than ranks
A fragrance brand grew about 27-fold from February to June, a skincare brand about 114-fold from March to June. In a monthly ranking, they “climb a few places”. In a historised series, they are the two most important signals of the period.
3. Same category, different structures by country
In the UK, skincare and make-up make about half of beauty revenue. In Germany, Spain and Italy, it is personal care appliances and fragrance, around 44 to 46%. France is more balanced. A launch plan copied from one country to another aims at the wrong target.
4. Levers show up in the detail
Creators tied to the seller make 5.3% of relationships but 25.3% of creator revenue, with a median per relationship of €4,380 against €630 for independent ones. And videos tied to the platform’s commercial operations generate clearly more per video than generic hashtags. These are leads to test, not proven causes.
The limits
Estimates. Revenue comes from a third-party estimation tool, not from shops’ accounts. Orders of magnitude are reliable, decimals less so.
Five months. Enough to see accelerations, not to separate a trend from a season.
A panel. Findings on creators concern the observed panel; they do not describe every creator on the platform.
What to do about it
Do not order a study: build a reading that repeats. In practice:
- Week 1Write down the three questions management really asks, and the level of detail that answers them (country, category, player, month).
- Every monthAdd a market snapshot without overwriting the previous ones.
- ContinuouslyTrack accelerations and breaks rather than ranks, with an alert when a player leaves its trajectory.
- On demandReuse the same base for the next question: pricing, portfolio, entering a new market.
Frequently asked questions
What is GMV?
Gross Merchandise Value is the total amount of sales placed on a platform, before returns, discounts and fees. It is a good size indicator, not net revenue.
Which indicators should competitive monitoring track?
Market share by category, three-month growth, concentration of the top players, and levers (price, creators, content). Always over time, never on a single snapshot.
How often should it be refreshed?
One snapshot a month is enough for most decisions. Daily data only helps to explain a specific peak: a campaign, a stock-out, a launch.
Baptiste Pigaux founded HOCUS after more than 20 years turning data and machine learning into business decisions, for more than 30 organisations. LinkedIn profile