Cities actually spent quite a bit of money on hiring consultants, site planners, doing impact studies, etc. I don't think it's unreasonable to say mid to high 7 figures.
-https://www.consulting.us/news/290/consulting-firms-recruite...
-https://technical.ly/philly/2017/09/13/site-selection-consul...
This is not to mention that cities likely had to hand over demographic data that would've otherwise been very expensive to collect: https://www.businessinsider.com/amazon-hq2-search-data-2018-....
Look at some of the cities that made the finalist list:
-Columbus (population ~880k)
-Denver (population ~700k)
-Indianapolis (population ~870k)
-Nashville (population ~667k)
-Newark (population ~285k)
For these cities, an additional 25k jobs would have been huge. Scott Galloway has advanced the theory that a large goal of the HQ2 search was to secure competing bids, which Amazon then took back to NYC and DC in an attempt to secure better terms. If these "finalist" cities were simply used as stalking horses with ~0% chances of actually winning, I can't help but feel it was a little wrong to induce them to spend the necessary fees on consultants and planners to put together a decent bid.