The R² on the Dashboard
Decide what you would do from the brief alone, including whether you would change anything at all. Everything below it is available, but the exercise stops working if you open it first.
The finance team's dashboard shows the house-price model at R² 0.89 and calls it "89% accurate". A regional manager says the model is "useless in our area" and wants it switched off there. Both are looking at the same model.
Retraining a separate model for the manager's region so its R² improves. A per-region model has an R² computed against the regional mean, which is what the manager needs, and it will look worse, not better — the "improvement" comes from tuning until the within-region R² is respectable, on far less data, with far more variance. The dashboard now has two incomparable numbers, and the underlying fact — errors in currency did not change — is hidden behind a metric that was never about the manager's question.
Read this even if you are confident. It is here rather than behind a button because it is the answer most teams actually ship, it passes review, and its cost arrives weeks later when the labels do.