How do you forecast demand for last-mile delivery in new cities?

Honestly, for new markets I'd start with secondary data first — census demographics, e-commerce penetration rates, and competitor density from tools like Placer.ai or even Google Maps scraping. Layer a small survey on top (200-300 respondents per city) to validate willingness to pay and frequency assumptions. Saves budget and you go into the survey already knowing what to pressure-test.