Choose the geography before comparing the evidence.
See where market signals concentrate.
Read the evidence before using the result.
Choose a question, then open the right evidence.
Keep metro, county and city scopes separate.