Market research data is a set of estimates of how large each product category is in each country, how fast it is growing and which companies hold the largest shares. Strategy teams use it to plan. It is also called market sizing data. A typical dataset covers hundreds of categories, from packaged food to pet care, in about a hundred countries, with figures for past years and forecasts for the next five.
How are the numbers made? No single source counts a whole market. So analysts build each estimate from many sources: government figures on production, imports and household spending; company annual reports; retail measurement figures; trade surveys; visits to shops; and interviews with manufacturers, distributors and retailers. They check the sources against each other. For example, the sales of the five largest brands added together cannot be larger than the whole market. They then publish one figure for each category in each country. The same definitions are used everywhere, and the figures are given in local currency and in one common currency, so that countries can be compared.
Imagine the strategy team of an Indian ice cream maker that wants to sell abroad for the first time. The team compares four nearby countries: Bangladesh, Nepal, Sri Lanka and the United Arab Emirates. For each one, the dataset shows the size of the ice cream market over the past five years and the forecast to 2031. It shows how much is sold through supermarkets and how much through small shops. It also names the three largest brands. The team finds that one market is growing fastest, but two strong local brands lead it. Another market is smaller, but no brand clearly leads it. The team plans for the second market, and checks the figures in its own interviews before the board meeting.
These datasets are estimates, not counts. A forecast five years ahead depends on guesses about incomes and prices. Categories sold mostly in informal markets are hard to measure. So planners use market research data for the direction and the rough size of a market, not for decisions from month to month.
Buyers usually subscribe to an online database and download tables from it. More and more, analysts find these figures through financial research platforms, which search reports with a language model. Selling a dataset of this kind is an example of data as a service (DaaS).
How is it different from retail measurement? Retail measurement counts what actually sold, product by product, each month, from shop data. Market research data estimates the size of a whole category, and is used to decide where to compete, often years ahead. Both belong to retail and audience measurement.