Starbucks/Walmart update

Alex F. commented here about problems with our Starbucks and Walmart data. Elizabeth Kaplan, who collected the data for me, replied:

Yeah Walmart was a bit of a pain to find the locations for as you can not search just by state on their website, like for Starbucks. In order to find the locations, I relied on the yellow page results. Even though I looked through to eliminate double postings for walmarts with the same address, after I looked into it again tonight, it appears the yellow pages dramatically over-represented the number of Walmarts per state. I have attached the correct data. All of these numbers come from this website ( which I was unable to locate before. Hopefully, that will help people who want to visit a Walmart store. Try to go before the store is closing as cleaners can sometimes leave equipment on the floor which has caused some accidents previously. This has led to some cases where a Walmart injury attorney has had to be called. Hopefully, Walmart staff will take more precautions, but shoppers should be careful too.

As far as the data for starbucks that should be correct as I got it straight from their website. The one thing is that they don’t list all affiliate stores (that is stores not own and operated by the company). There is no reliable source of data on affiliate stores by state, and obviously the yellow pages are not a good source. So the data I sent to you just includes Starbucks owned and operated stores.

Also for population I used the 2006 Census Bureau estimates.

This sort of thing happens all the time to me, so I certainly don’t think Elizabeth should feel too bad about this. I’m just glad that Alex noticed and pointed out the problem. Anyway, here are the corrected maps:



and scatterplot:


And also, following Seth’s suggestion, the scatterplot on the log scale:


And, following Kaiser’s suggestion, a reparameterization showing people per store (rather than stores per million people):


28 thoughts on “Starbucks/Walmart update

  1. I'm struck by the degree to which home base continues to matter — even for two very mature companies. Starbucks started in Seattle and maintains a heavier presence in the west despite its seeming ubiquity. Walmart started in Arkansas and maintains a southern footprint despite its ubiquity. Does the negative correlation hold up once distance from corporate headquarters is accounted for?

    I also wonder if you would see the same pattern with the global expansion of the two companies. The two companies serve different clienteles, so it is possible they see different opportunities abroad.

  2. You won't see quite the same pattern in the UK. Starbucks has expanded from big cities (probably starting in London). Walmart's geographic distribution is almost entirely down to the distribution of Asda stores it bought (still run and primarily branded as Asda). Barriers to market entry for supermarkets being very different to that for coffee shops – especially in the UK.

  3. Surely this video is relevant to the discussion:

    It looks like a diffusion process, and if you think about it, it makes sense. It's not like Wal-Mart can just establish new stores in random locations on the map. I'm just guessing, but I would bet it's driven by the distribution centers. New stores have to be close enough to an existing distribution center to be supplied economically with merchandise. So they build new stores along the highways that are within a certain distance of an existing distribution center. Then after some time the distribution center reaches its capacity, so they build a new one "further out". The process repeats …

  4. I appreciate these maps and scatterplots. Thank you.

    Walmart started with smaller stores in small towns. As they expanded nationwide, they discovered that large supercenters in suburban locations are much more profitable. The number of stores per capita may vary more than the square footage per capita when comparing early Walmart states with later ones. Assembling data on square footage is probably an impossible task, and I'm not suggesting anyone try.

    The Walmarts in very dense cities – Los Angeles, for example – probably generate more sales per square foot than do the stores in Oklahoma and Arkansas and Missouri. A more enlightening comparison might be Walmart's per capita sales by state – though I doubt the company releases that data.

  5. Found this via and his comment made me wonder, does the density of Starbucks or Wal-Marks have a correlation with the state's tendencies in elections. I have not done any analysis other than compare these maps with election maps, but it looks like states that tend vote democrat have more Starbucks per capita.

    Reference map:

  6. I was going to prod you about not including Alaska, but, never mind: I think we only have five in the entire state, with two being located at the Anchorage airport.

    Move along; nothing to see here!

  7. In spite of a number of comments here about the way that Walmart spread, the actual approach is called "reverse hierarchical diffusion," and in the case of Walmart shows up as the idea that instead of the traditional approach of putting big store in big cities, you put a Walmart in the smallest towns that could possibly support such a store. The reason and result? In towns small enough that they can just barely support a Walmart, there is nothing that compete with them: when I Walmart is put into such a town, nobody is able to compete. That is the source of many of the complaints against Walmart, obviously. Starbucks, on the other had, had followed an approach of traditional diffusion: they go where there is a market, like following a seam of gold in a mine shaft.

  8. Walmart filled a gap in rural America where residents had to drive a hundred miles or more to find a department store. I'm not sure what gap Starbucks filled other than giving those who think wasting $5-$7 on a cup of coffee somehow proves they are successful. Wasteful spending is still wasteful spending any way you slice it. So compare those who want the government to support your every need against those who value a dollar because they have to work hard for each and every one.

  9. Rich State, Poor State…?
    PhD State, High School Diploma State..?
    Red State, Blue State..?

    What's the point here…? Why the big fascination with breaking the country down into two camps, especially when most of the divisions are almost equal. What does that do for us except disenfranchise half the country..?

  10. John S., as noted the distribution of walmarts has some geographical dependence – except look at Illinois — quite close to Arkansas geographically, yet the same inverse relationaship as most coastal states (with VA a notable exception). Since no one else has mentioned it, one might cite union membership (a requirement stated as a reason there is no walmart in San Francisco), urban centers (there seems to be a correlation between high population cities and the states that have more starbucks (NYC, LA, San Diego, San Francisco, Boston, …).

    Disposable income is tied with this; my ability to pay 1.65 for a regular cup of coffee translates into others being able to afford 100K as a down payment (the cost of a house in many places).

  11. I will say after living in Colorado for several years — the reason Starbucks' numbers are so different is that they bought out a couple of other coffee house chains in recent years and switched their locations over to the Starbucks brand. One example of this from just about a year and half or two years ago was Peaberry Coffee. So you sometimes had coffee stores located just a couple hundred feet apart that became two Starbucks right side by side (almost). I'm sure that as time goes by, these duplicative stores will be consolidated to one location and then the other shuttered, dramatically cutting back on the number of stores in the state and thus changing this data.

  12. The Rand McNally Road Atlas has a complete listing of Wal-Marts and Sam's Clubs in North America. Only Road Atlases sold at Wal-Mart have the special "wrap" containing all the locations. And it gives actual addresses and services of each store, unlike, so you could plot every location with GIS and do much more detailed analysis, if desired. You'd have to enter them all by hand, unfortunately, as that list is not available electronically as far as I know.

  13. G —

    There's no Wal-Mart in San Francisco because it's a difficult place to get around in by car. It doesn't have many arterials, land is too expensive to have large parking lots => parking is a pain. While many people *have* cars, hopping in the car to go five miles is not as easy/pleasant as it is in the burbs.

    This means that people spend more time in their own neighborhood, buying smaller amounts more frequently.

    This is similar to Vancouver, BC, where I live. It's a real pain to drive here, and there are no big-box shops. Out in the suburbs (e.g. Richmond), it's much easier to get around by car and they have all the big-box shops.

  14. Wow, so many comments. Thanks! In brief response:

    Many people noted the diffusion pattern. That's what was interesting to me too, but I hadn't thought about Walmart's small-town strategy.

    Yes, I'd be happy to see the Dunkin Donuts map too. At some point it gets a little silly, but I thing a few more maps would still be fun.

    I did make some maps showing Democratic and Republican vote. As I recall, Walmarts was strongly correlated with Republican vote in the state, but Starbucks was not particularly correlated with either party.

    Sara: I'm not a coffee drinker myself, but I can only assume that people pay $5 for a Starbucks coffee because of some combination of: (a) liking the taste, (b) having a spare $5, and (c) not having a lot of high-quality $4-or-lower coffee options out there.

    GadgetGav: My fascination with breaking the country into two camps is that I don't like that pundits, etc., often break the country into two camps. I'm hoping that exploring and understanding political differences will ultimately reduce the stereotyping. I agree with you that it's horrible when people seem to think that everybody, or even the majority, of people on the other side politically are evil/stupid/etc.

    Jason: Good point. Another reason why these maps shouldn't be taken too seriously…

    Kaitlin: Interesting site. I'll have to try it out.

  15. So, does anyone know an official count of Starbucks within San Francisco? I'm currently in town visiting from Raleigh, NC where there is a healthy mix of Starbucks, Caribou and independents. This place (San Francisco, that is) is ridiculous, and I imagine that the case is similar in Seattle and other West Coast cities. I can literally stand on a corner and see two or more, oftentimes across the street from another. Google searches only yield a few of the locations, and I am curious just how many DO exist here…

  16. The first map just couldn't be BLACK enough to represent the Starbucks in Washington State vs. all the rest.

  17. If you type in Starbucks to GoogleMaps Find Businesses it comes up with 109,150. That may include some stores that are not cafes, but I doubt that's a very large number difference.

  18. One thing to note is that I am originally from CT, and Starbucks is not as popular here as Dunkin Donuts. We don't have that many Wal-Marts per capita, so I think that your ratios for New England are thrown off by different coffee preferences.

Comments are closed.