Data Science · 10 May 2026 · 5 min
How do you rank a market when no sales label exists?
GENAQ turns an unlabeled market-entry question into an explicit, reviewable scoring system.
01
No label means no conventional prediction target
For a product without historical sales in the target market, training a sales predictor would pretend that the missing outcome already exists. The project instead ranks 35,891 Spanish census sections using public demographic and climate evidence.
02
The score is an argument
When labels are unavailable, feature selection, normalization and weights encode the business hypothesis. Making them explicit matters more than wrapping the ranking in a more complex model, because each assumption can be challenged and revised.
03
A ranking is a starting point for validation
The output narrows a national search into prioritized areas. It should guide field research and future measurement, not be presented as proven demand. The repository publishes the ranking together with its methodology and limitations.
Takeaway
What I take from it
When the target is missing, make the decision logic inspectable and design the next experiment that can create evidence.