Why Data Alone Won't Predict Amazon's New Location
Originally published in LinkedIn.
Amazon is contemplating the location of its next massive headquarters, and everybody is collectively losing their minds. The prize- which will be contested in a manner reminiscent of the search for a World Cup venue and, perhaps, the urban planning equivalent of the Triwizard Tournament- is significant, with an estimated injection of $5 billion capital investment and 50,000 high-paying jobs. Proposals are due in October, and pundits have already projected winners, narrowed down from a shortlist of viable candidates, from Denver to Toronto to Philadelphia. These assessments are typically based on an analysis of very sensible parameters such as the cost of housing and office space, the availability of skilled workers, ease of travel both to and within the city, and the willingness of the local governments to play ball with respect to tax breaks and incentives.
Here’s why very little of that will matter.
1. We have data, but no insight into the data that matters
What is more important to Amazon in a location: an existing skilled worker base for them to hire or poach, or a cost of living so deliciously low that they could immediately lure skilled workers from nearby cities and build cheaply to their heart's content? Both would solve the same need for Amazon, yet the former would give a massive advantage to Atlanta or Washington DC, while the latter would do the same for Pittsburgh or Baltimore. The deciding committee will likely have a bias for one or the other path, and without that information, we are left to guess, knowing that one of the two groups will likely be preparing a response with no real chance of success. Components of a decision-making process are rarely all on a level playing field. Weighting the relative importance of each component is difficult but crucial to get a true understanding of which criteria will truly drive the decision.
Resulting prediction: Amazon will weigh existing skilled work force, eastern geography, and access to really good chili dogs and fried peach pies above other components, and choose Atlanta as its new headquarters.
2. Key data is often hidden from everyone except the decision-makers
It’s all armchair quarterbacking, but it’s fair to suggest that Austin would have made far fewer shortlists were in not for Amazon’s recent decision to buy Austin-based Whole Foods for $13.4 billion. Over and above the Whole Foods acquisition, in the past year, Amazon announced an array of new things, including Prime Wardrobe, a Try-Before-You-Buy clothing service, a visual echo home assistant, and a partnership with the NFL to stream Thursday Night Football. Amazon innovation and experimentation is astoundingly diverse in scope, location and vertical, and the big bets, both common knowledge (e.g., Alexa) and unknown, will necessarily help forge growth decisions. Now we must consider that there are at least three other pending launches or acquisitions in the next twelve months that have not been announced which are influencing the decision making team at Amazon. And we are utterly unaware of those.
Resulting prediction: Amazon will announce a surprise deal to acquire a controlling stake in FedEx and choose Memphis, TN as its new headquarters.
3. Human factors can influence or outweigh data inputs
Whether you are designing a bathroom remodel or choosing a facility for 50,000 of your future employees, there will always be a solid nod to practicality (e.g., we need a toilet). Past that however, there are a number of factors that are subject to radical personal preference or even whim (e.g., we need this toilet) which may, given the right circumstances and leadership, even grow to equal parity with the practical concerns. While it is unlikely that Amazon would choose to build in a city primarily because it has great Mexican food or because the fall foliage is lovely, it is entirely possible that the deciding committee might make a decision at the expense of some practicalities in order to make a positive change in the world, to elicit a public affairs victory, or to otherwise align with leadership interest or values.
Resulting prediction: Amazon will announce “Project Motown,” a bold urban revitalization initiative as it chooses Detroit, MI for its new headquarters.
For most of us, predicting the site of Amazon’s new headquarters is no different than wondering how a TV show season might end- an amusing thought exercise sparked by personal interest. But the cities who elect to participate in this process will spend millions of public dollars crafting their proposals, hedging their chances on a data set that is at best incomplete, and at worst is misleading. Indeed, it doesn’t require complete cynicism to entertain the hypothesis that Amazon has already decided on a location partner and is merely playing the role of Willy Wonka to apply leverage for better tax incentives from them. Either way, we can be certain that Amazon, the one player that has all of the data required, will make a decision that best aligns with their business needs, while we continue to sit back and watch in awe of their success.