Tuesday, October 10, 2017

Questionable statistics alert – – Whole Foods Markets edition

From Yahoo finance:

High-end grocer Whole Foods is already poaching new customers from Walmart (WMT) and other competitors as the result of its new owner Amazon (AMZN) cutting prices, according to a report from Thasos Group, a data intelligence firm founded at MIT.

Beginning Aug. 28, Whole Foods, which had earned the nickname “Whole Paycheck,” slashed prices on some of its best-selling staple items, including avocados, bananas, organic brown eggs, Fuji and Gala apples, almond butter, and farm-raised salmon and tilapia.

Whole Foods immediately experienced a 17% increase in foot traffic year-over-year for the week beginning Aug. 28, the Thasos report found.

For the week ending Sunday, September 3, the weekly YoY change in foot traffic to Whole Foods peaked at 17%. On August 28, the daily YoY change in foot traffic peaked at 31%.

Of that foot traffic, new customers came primarily from Walmart, accounting for 24% of the new shoppers that week, while Kroger (KR) shoppers accounted for 16% and Costco customers for 15%.

Some of the Whole Foods’ other competitors also saw significant defections during that time. Nearly 10% of Trader Joe’s daily customers shopped at Whole Foods during the first week of price cuts, while 8% of Sprouts (SFM) and 3% of Target (TGT) customers did so.

Lots to unpack here, but in consideration of time constraints (both yours and mine), let's just focus on one aspect.

Possibly the most important and certainly the most underemphasized piece of context in the coverage of the acquisition is the size and distribution of Whole Foods Markets. This is a small player highly concentrated in a few areas (such as New York City, DC, San Francisco, and Los Angeles). These areas also tend to have a highly disproportionate representation of the people who write about and invest in business trends (which explains a lot about the coverage). These areas also (and this is important for this story) tend to have a disproportionately low number of Walmarts.

Because of its size and economic diversity, Los Angeles makes for an interesting case study (it also helps that I've been here for quite a while and have lived on both sides of the county). Keep in mind, when Angelenos talk about the town they are generally referring to the county and Los Angeles County has a land area of over 4000 square miles. There are a lot of good reasons for going from one side to the other, but picking up a few groceries is not among them.

If you know the town and you look at a map of LA Whole Foods Markets, it is immediately obvious that the stores are heavily concentrated in very upscale neighborhoods, particularly in the general vicinity of Santa Monica. Not only are you unlikely to find Walmarts in these neighborhoods; you are unlikely to find them anywhere in reasonable driving distance. Likewise, the parts of town are you are likely to find Walmarts have very few Whole Foods. It is also important to note that this relationship does not hold with Target stores, which have long done a remarkable job adapting their model to both sprawling suburbs and high density urban areas. Anywhere in LA will put you fairly close to a Target.

Obviously, there are exceptions (Austin comes to mind) and we would certainly expect some of the traffic from a WFM spike to come from regular Walmart shoppers, but any analysis that shows them making up the plurality must be viewed with suspicion, even more so when we consider the surprisingly low numbers for Target. These numbers are almost exactly the opposite of what we would expect.

I know what I'm about to say goes against everything you've been told, but when data and common sense diverge sharply, the first thing you should do is question the data. Yes, counterintuitive results can be the most interesting if true, but it is essential to remember that counterintuitive results are also far more likely to be untrue.

No comments:

Post a Comment