Amazon’s HQ2 selection process has been described as a beauty contest, which misses the point. Amazon is definitely not going to pick a city based on popular opinion or consensus.
An excellent theoretical framework to think about Amazon’s choice process for HQ2 is the idea of Innovation Tournaments. A good resource to learn more is the wonderful book Innovation Tournaments by my Wharton colleagues Christian Terwiesch and Karl Ulrich, which I highly recommend.
Innovation tournament typically involves several contestants going through a series of rounds, as pictured above (under a selection-criterion and pre-announced rules), until a “winner” is chosen. American Idol is a TV-show that typified this idea.
Innovation tournaments are a genuinely great method to brainstorm and generate new ideas. For example, a startup can pick a logo by inviting several designs, running an open tournament where the designs compete until a final design is picked. In this case, Amazon is running open, multiple rounds, pure cascade tournament with absolute filters, with few twists.
The tournament is open because any (North American) city/region could enter; there are multiple rounds (at least two public rounds, so far); this is a pure cascade tournament because losing entries don’t seem to have a re-entry process (at least, so far); This is a tournament with absolute filters because Amazon is definitely using some credible yardsticks, — the supply of educated labor force, international airport access, etc.
Such open tournaments have been done before by organizations, (and even for-profit firms). For example, the $25000 Orteig Prize offered to the first aviator to fly non-stop New York City to Paris directly led to Charles Lindbergh crossing the Atlantic Ocean in the Spirit of St. Louis. X-Prize Foundation Prizes belong in a similar category. (I heard the 30M prize wound down earlier today).
So what’s Amazon’s twist?
Research in Operations recognizes that such tournaments are inefficient ways to generate new ideas, because a ton of effort is spent in developing opportunities and ideas, and most ideas are anyway abandoned.
What Amazon has done is different. The development costs are externalized from Amazon and borne entirely by the competing bidders. Unlike in a typical firm-level tournament, there is no idea exploration-exploitation trade-off for Amazon, as each city passes through the filters. (This explains why Amazon has multiple “cities” that cover the same location — e.g. around DC, NoVa, and Maryland).
Every choice has a shadow price for Amazon (based on Amazon’s estimated value of the location to Amazon vs. the friction/cost in implementation of the HQ2 plan). The competitive bids reduce the friction of implementation (with measures like tax cuts, land lease agreements, etc.). The top 20 contenders through their efforts corroborate the estimated value of the locations and reveal the true costs of execution that are not otherwise available to Amazon. Amazon then likely chooses the best bid that meets its shadow price.
This is not to say that it is a bad idea for cities to bid. (Bidding for a professional team is bad). In fact, for cities like Philadelphia, currently so dependent on health care and Pharma, the entry of Amazon will be a transformational positive employment shock.
Some imprudent forecasting:
I hope that Philadelphia wins, but… my own (uninformed) forecasts: [Remember, all point forecasts are wrong!]
NoVA/DC/Maryland area or Raleigh, NC — both based on lexicographic preference for the local university talent pool.
Dark horses (in my book) are Dallas and Toronto because of better access to universities and high availability of educated labor – Toronto is an AI hotspot — in addition to being cheaper locations.
Many, many smarter people have bet on NYC — I am not yet convinced.
Reference:
- Innovation Tournaments: Creating and Selecting Exceptional Opportunities by Christian Terwiesch and Karl Ulrich. 2011. Harvard Business Press.