As part of HQ2 search, Amazon officials have already visited 10 of 20 finalist cities, which are providing them with data including SAT scores of local students
Finalists for the company's second home base break out bicycles to win over the internet giant — Don't get too fancy.
Context & Ripple Effects
Amazon's HQ2 contest has moved from paper to fieldwork: after narrowing the field to 20 cities from 238 proposals in January, company officials are now touring half the finalists — and asking hosts for granular civic data, down to local students' SAT scores. The visits signal that the shortlist stage is less about brochures than about Amazon auditing each city's actual talent pipeline.
The later arc of this story shows why those requests matter: Amazon reportedly kept a wealth of city data from all 238 entrants that could inform siting decisions well beyond HQ2, and the eventual pick emerged from a process where cities had offered everything from employee relocation funds to an exclusive airport lounge, per the incentives roundup. The SAT-score request is one data point in a broader pattern of cities outbidding each other with information as much as tax breaks.
First-order effects
- The 20 finalist cities are now actively courting Amazon with site visits and disclosures of education and workforce data, effectively opening their civic records to a counterparty that is simultaneously evaluating all of them.
Second-order effects
- Data gathered across the full 238-city pool gives Amazon a proprietary comparative map of US and Canadian labor markets it can reuse for future facilities like warehouses, independent of where HQ2 lands.
Third-order effects
- If mega-bids keep working this way, corporate site selection hardens into an asymmetric auction: cities compete by surrendering granular public data, and the winner-take-most employer exits with intelligence no single bidder holds.
The trend: Cities competing for anchor employers are shifting from tax-incentive bidding wars to data-disclosure contests, handing prospective tenants a comparative dataset of their own workforces.