A New American Geography: Cultural Ecoregions for Consumer Segmentation
December 11, 2012
I’d like to introduce my first pass at a new American Geography with regional definitions based on common culture, climate and environment rather than political borders.

I started this project because I needed a better way to define different American regions so I can help my clients, who supply or distribute goods in the US, better understand how to “cluster” their store locations and their customers. Demographic segmentation systems have become a popular tool for analysis of customer data to make retail location decisions; they’re also quite useful in a variety of other marketing strategy campaigns. I’ve used these data to build sales forecasting models and while they certainly are effective for identifying locations with similar socioeconomic status they don’t usually take into consideration environment/climate or unique cultural attributes that exist in different parts of the country. This can be critical for certain types of consumer purchase behavior. For example, if you’re in automotive, you know that consumers on the West Coast and the East Coast are far more likely to buy foreign name plates rather than domestic vehicles. You also know that 4-wheel drive vehicles are going to sell better in locations with more snow or with uneven topography. So clustering consumers by income variables doesn’t really capture the many differences in purchase patterns for companies selling cars, car parts or services related to specific vehicle types. Same goes for home improvement and many other consumer products industries. Should Home Depot and Lowes stock the same garden supplies at the same time in all parts of the country? Of course not.
Another problem is the use of political boundaries. Often the most granular location information for model development is State of residence. For example, in building replacement rate models for particular car part categories we’ve made model bias adjustments at the State level due to limitations in the available survey data. This is certainly better than using Census regions or divisions but still falls short when it comes to large states like California, Texas, Florida…even Oregon, Washington, Colorado, Illinois. The problem is that there often exist clear cultural and/or environmental differences within a particular state that need to be considered. Let’s say you run a chain of stores that sell tires, like Discount Tire, Les Schwab or even WalMart. Do you really want to stock all stores in California with the same inventory assortment, fitting the same profile of vehicles? Something tells me that you’re likely to find a pretty different proportion of Toyota Prius models in Truckee vs Tustin. Do you think there may be a few more Ford F-150 pickup trucks in Modesto vs Mill Valley?
So, I’ve tried my best to come up with a set of regions that can be used for these types of analytical endeavors and, perhaps, they can be used to help better understand America’s unique cultural landscape. My colleague and fellow geographer, Steve Scobie, and I started by defining our own set of regions and then comparing – our own little “double blind” study. From that we came up with some shared boundaries and now I’ve taken it a bit further with some modifications and a set of names to go with each region. I’ve also created a set of 8 divisions to group the regions when a smaller number of regions is needed.
I would love to get some feedback on this. I know the Western US fairly well, I’ve lived in the Midwest and I’ve traveled to most parts of the country but there are plenty of areas that I don’t know that well including most of the South and the Northeast. If you see a boundary that doesn’t make sense to you I would really like to hear what you don’t like and how you might change it. I would also love to hear thoughts on the region names. Steve contributed several but I’ll take blame for any that seem incorrect or inappropriate. My apologies in advance for any place names that someone may find offensive. I’m sure there are better names out there. I’m especially interested in names that are already in use by people living in or near the region. But please don’t submit names with any sort of negative connotation.
Please share your thoughts in the comments section below. Thanks!
By the way, I had so much fun making this map. It was produced “old school” by tracing an outline of the “Lower 48” and then using a variety of maps (political, environmental, topographic, demographic) to select boundaries and draw/trace them by hand. Takes me back to Cartography 101 with the legend of Oregon cartography, Bill Loy; now gone but not forgotten.
35 Comments
Justin this is an excellent approach. I have been thinking about something similar but based principally on population density, resource availability / distribution, and land use as the core factors. I have been working on a book: ‘The Science of Man’ for a couple of years now [old title but new approach] and am trying to look at the global picture but with specifics on Western Democracy. Mapping the data [physically and statistically] is an important part. Your approach – particularly if a temporal dimension could be added e.g. changes for 10 years segments linked to census data could be really useful. I hope I can keep informed about your progress.
What sort of database do you have?
George F. Hart,
[email protected]
George – thanks for the comment! I’ll post here as I make progress. I like the idea of the a temporal dimension. No coherent database on this just yet but coming soon. Cheers, J.
Brilliant piece of work! This completely changes the way folks like me analyze customer data for strategic planning.
Bravo!
Ganesh
Thank you, Ganesh! Let me know if you have specific application ideas, especially if I might be able to contribute in some meaningful way. Best, Justin
This is great stuff. I did some similar thinking when I did analysis for an apparel firm, specifically jeans. There were clear differences in the type of jeans people purchased, that couldn’t always be described by demographics. True, there is a correlation between ethnicity and jean size (waist & inseam), but other predictive factors come into play, such as climate, topography, proximity to water, and proximity to major universities.
And in terms of the auto industry, I noticed while living down in NC that I saw many more RWD cars (BMW, Mercedes, etc) than I did when living in Ann Arbor. There’s just less perceived need for FWD/AWD down there. And while they still need window scrapers for ice storms, they don’t need floormats that hold a gallon of water.
I like how you’ve separated the Mississippi Delta from nearby areas such as NOL and Dixie. It’s definitely it’s own unique area. And yes, the eastern halves of KY, TN, NC, and SC are very different from their western halves.
If I may make a suggestion, I think New England could be split in two. Not sure of the exact border, but the southern half of New England is much more affluent, urban, and multicultural.
Hans – thanks for these comments! I agree with the inclination to divide New England but, like you, I wasn’t sure how to do it properly. At one point I considered putting the upper Maine coast in a separate region but I couldn’t convince myself where the border should be and ultimately left it alone. Any New Englanders out there care to lend a hand? Thanks again, Hans! Cheers, J.
I personally would split it at MA’s northern border. Keep it simple. Except maybe include Nashua since a lot of Boston commuters probably live there.
[…] It’s way too difficult to predict that far out into the future. That’s why I prefer to look no further than 1-2 years out for business forecasting. Will miles driven increase in the next year or two? Perhaps but probably not by much. Will miles driven decrease in the next year or two? Perhaps but probably not by much. Will cars still be the primary mode of transport in 2 years? Yes. There you go. Three forecasts you can hang your hat on. Obviously those forecasts aren’t worth much. But, I would consider paying good money for a forecast of miles driven in Q1 and Q2 2013, especially if it were available by region. […]
Justin,
This is very interesting! Thank you for sharing. I like what you’ve done and just wanted to offer some feedback from someone who lives on the east coast.
From my experiences living in the southeast, the Coastal Carolina/Southern Piedmont/Appalachia regions are spot on. I think the Northeast region is a little more problematic in terms of defining distinct segments. As someone who grew up in northeastern Pennsylvania, the eastern part of the state always felt culturally distinct from western areas out near Pittsburgh, which I always likened to the other Rust Belt cities (Buffalo, Rochester, Cleveland, Detroit). You can see this most apparently in dialect (“soda” versus “pop” for instance) as well as in sports followings. In other words, rather than a north/south distinction between the Niagara/Steeler Country regions in your conceptualization, there may be an east/west distinction (you could approximate this as the Susquehanna River Valley to the east). I am not sure that these kinds of idiosyncrasies are of importance for the purposes of this kind of segmentation, but it would be interesting to see if others familiar with the northeast can offer any additional insight that could be of help.
Hi John – thanks very much for these comments. Glad to hear that the Coastal Carolina, Southern Piedmont and Appalachia regions felt right to you. I agree with the Northeast and I will look into an east-west border. At one point Steeler Country was significantly smaller containing greater Pittsburgh north to Erie roughly. I may want to fall back to that definition and give the rest of Western/Central PA its own region and name. Can you think of a good name for West/Central PA squeezed between Steeler Country and the Delaware Valley? Thanks again! Best, Justin
Hi Justin,
Long time no talk, you remember we did a UK grocery project together back in the day?
Great minds clearly think alike 🙂 I did something similar across the UK a while back during a project for a UK home improvement chain. Your logic is on the money, some obvious examples from here in the UK are beer (North is ale South is lager), wall coverings (North is wallpaper south is painted plaster) and horticulture (soil dependant). Maybe I should dig them out and we can get ex-colleagues/friends to do Canada, Australia and New Zealand. We end up with the Anglosphere’s ‘true’ regions!
Best from this side of the pond.
Blair
Blair – of course I remember! Great to hear from you! Hope all is well. I seem to recall using some of your region definitions in that very project. I would indeed be interested in having something similar for Canada. Let me know if there’s something I can do to get the ball rolling. Thanks again for the note and very best wishes from this side of the pond. Cheers, Justin
Justin, this is interesting work. You may be interested by a book entitled “The 9 nations of North America” written by Joel Garreau who was a journalist working for the Washington Post if my memory is good… I read it 25 years ago). You will find a description of his excellent book and his map of the 9 nations in Wikipedia since it has become a classical book: http://en.wikipedia.org/wiki/The_Nine_Nations_of_North_America
As you noticed, those boundaries don’t follow state or province boundaries but rather cultural ones. The only exception is Quebec where both political and cultural boundaries fit very closely and the only canadian region which is not merged with an american subculture.
These boundaries also fit roughly with elections results in Canada and in the USA.
Hope this adds to your interesting work !
Hi Yvan – thanks for the comment and for the book recommendation. I am aware of Garreau’s book but haven’t read it yet (ordered it from Amazon earlier this week). I’ve started reading Woodard’s American Nations (http://www.amazon.com/American-Nations-History-Regional-Cultures/dp/0670022969/) and also have Chinni and Gimpel’s Our Patchwork Nation (http://www.amazon.com/Our-Patchwork-Nation-Surprising-America/dp/159240670X/) waiting in the wings. I have to be careful because this project is too much fun and seems to be distracting me from other projects that are probably more likely to pay the bills. Thanks again! Cheers, Justin
“The Foundry”, now that’s a creative name, lol! If you only wanted 9 segments, I think those are right on the money.
Very interesting. Having lived most of my adult life in New England (and may still live there according to your map, although I’m in Saratoga Springs, NY; more details needed), I agree with the comment above that northern and southern New England are different. I would include much of western CT in “New York.” I probably would divide northern and southern New England farther south than the northern border of MA, though. Many communities near that border are quite New Englandee – Greenfield, Deerfield, etc. The dividing line might not be straight either, extending farther into MA on the west and perhaps even including part of southeastern NH in southern New England.
On the matter of names, what you call Niagara essentially encompasses the major and historic east-west transportation corridor made famous with the building of the Erie Canal and today followed by I-90/NYS Thruway and railroads. People don’t identify with the term Niagara vey far east of Buffalo, but all know what “Erie” refers to.
Wally – thank you for these comments! This is a big help. Much obliged! Best, Justin
[…] the map you see in this post to help determine the similar regions for auto parts. He has a post that explains the map and the fact the state boundaries are not helpful for forecasting auto […]
[…] has examined interior US geography from a cultural and economic point of view and has conceived of a new map of the nation focusing on micro-regions that he calls “cultural ecoregions.” I think the map needs more work, but it comes from an accurate theoretical presumption and it is […]
[…] It’s way too difficult to predict that far out into the future. That’s why I prefer to look no further than 1-2 years out for business forecasting. Will miles driven increase in the next year or two? Perhaps but probably not by much. Will miles driven decrease in the next year or two? Perhaps but probably not by much. Will cars still be the primary mode of transport in 2 years? Yes. There you go. Three forecasts you can hang your hat on. Obviously those forecasts aren’t worth much. But, I would consider paying good money for a forecast of miles driven in Q1 and Q2 2013, especially if it were available by region. […]
[…] the map you see in this post to help determine the similar regions for auto parts. He has a post that explains the map and the fact the state boundaries are not helpful for forecasting auto […]
Thanks for sharing – any way to access a shapefile version of the map (or an ArcGIS server map)? It would be fun to delve in a little deeper. My initial thoughts are that the desert southwest is a tad simplified and the the Sierra Nevada is too large. Also, I am curious where Boise is located in your representation – I’d include it in whatever region encompassed Salt Lake City. I could spend some serious time thinking about this…
Definitely get ahold of Nine Nations of N America. I’ve put American Nations on my library request list based on your post. Thanks!
Bob, thanks for the comments; I appreciate your feedback on the southwest and Sierra Nevada regions and will keep your thoughts in mind when I revisit and revise. Boise is currently in the Northern Rockies region. I doubt I’ll include it in the same region as Salt Lake City because the percent Mormon population is significantly higher in the eastern part of Idaho and that’s, essentially, what I’m using to define the northern and western border of the Salt Lake region. I can be convinced otherwise but I would need a good data element to sway the decision. Let me know if you have any further thoughts. I don’t have a shp file yet because I haven’t finalized the boundaries. Please stay tuned! Thanks again! Best, Justin
Nice job! Unlike you, I don’t know the Western USA well, but I do know the Eastern half of the USA. This certainly corresponds to what I know.
– Jim
Thanks, Jim! I appreciate the feedback!
From the Central/Finger Lakes/ Western New York Region. It’s a little tricky to discern the dividing line of the New York City from the rest of the State. However I think you are spot-on about the divide there. I work at a College here in the Finger Lakes and it is interesting to me that many of our students and some of the newer faculty are surprised by the expansive and rural nature of Upstate and Western New York; and how different the lifestyle here is relative to the downstate area. I like the fact that you named it Niagara, as the Niagara escarpment and other Geological features tend to play an important role in this difference between ‘upstate’ and ‘downstate’.
I see that Northern New York (Watertown, NY and St. Lawrence County) would be in New England. I had not thought of that before but having lived in that region of New York for several years as well I can rationalize that. Although I would very much characterize it as rural New England. To take that a step further, when I went to College in the ‘North Country’ as it is often referred to, you can be struck by how the language and accent often reflects more West Virginia than New England.
Very interesting map. This something I often think about having to confront the expectations of others when they move here and expect it to be more like New York City or New England; and they find they live more in Ohio or Indiana. I have often thought this is area is more of a ‘Great Lakes’ region to include Northern Ohio along the old ‘Rust Belt’ but different from Pittsuburgh due to the way Great Lakes trade and the Erie Canal influenced the patterns of life here early on.
Thanks for this map. Interesting.
Thanks for these comments, Robert! Great to hear positive feedback from the Niagara Region! Best, Justin
Hi Justin,
I know I’m coming late to this post, but I just came across it. You may be interested in this book on exactly the same topic of regional cultural differences in North America. I found it thought provoking as it goes much deeper than “north vs. south” and provides an imense amount of historical context.
http://freyfogle.tumblr.com/post/44777749377/book-review-american-nations-by-colin-woodard
Good to meet you last week, stay in touch.
Thanks, Ed! Yes, I have Woodard’s book. I’ve skimmed through but still need to give it a solid cover-to-cover read. Thanks for sharing the link to your review – definitely will motivate me to dig in. Look forward to keeping tabs on your progress with http://www.opencagedata.com/ and other projects. Cheers!
Have just stumbled across your blog and am having fun clicking through thoughts and links – thanks! Fun to pour over this map, check out the different places I’ve lived and traveled, and see how many of your regions make strong intuitive sense! As a Rochestarian who’s also lived in other parts of NYS, as well, I agree with on of your earlier commenters that most people in the region you label as “Niagara” wont identify with that language. I would think that “Erie Canal” or “Canal Corridor” would have a must stronger resonance for folks, or even “Erie-Niagara” if you want to keep the reference to the Niagara Escarpment. Thanks again, so much interesting stuff here to explore…
Glad you found me! And, thank you very much for the suggestion re Erie area. Very helpful indeed! Best, Justin
I lived for my first twenty-three years or so in New England and will accept the inclusion of the Adirondacks culturally into New England. Strangely enough they almost seem more like northern Maine to me than elsewhere in New England, possibly due to the distance from coastal and urban influences.
I have a bit of an issue with the southwest border of the Chesapeake along the bay – I’d say the break point is the Middle Penninsula (I currently live there). Further south there is more urban development and more emphasis on lifestyle associated wtih the military than here. In addition we are separated by the only toll bridge in the area. If that doesn’t say we’re not wanted what does? (The bay bridge tunnels are also toll to the eastern shore). We can link but it’s as though we really aren’t fish or fowl as to what should be our MSA, Richmond or the Newport News/Hampton/Virginia Beach MSA.
Personally I think the southern Rocky Mtn/Adobe Mesa border could (maybe should) be a dotted line. To me from the point of view of prehistory and cultural influences there always seemed to be a very permiable border there. (Yeah, did archaeology for several years out of Denver and the west. ) I know you need to call borders somewhere, but some are more permiable than others and that was one that seemed permiable to me if you think about the areas where the native peoples adapted to all those environments and thus influenced the modern cultures too. Let’s face it, there’s a good number of adobe style homes built in the Denver metro,too.
My two cents from years of observations of those areas.
Winnie – thank you so much for sharing your perspective! I especially appreciate the input in regards to Chesapeake as I’m not familiar with the area. Re the Rocky Mtn/Adobe Mesa border remember these are cultural “ecoregions” and the borders are meant to reflect not just cultural homogeneity but also environmental/climatic consistency. As I’m sure you know if you’ve traveled south through Colorado into New Mexico the landscape and the climate change in a fairly dramatic way. There are also major differences in terms of racial/ethnic population composition. So, I think I’ll keep the border a solid line, at least for now. Thanks again for taking the time to comment and share your thoughts! Best regards, Justin
Justin – One of the problems I have is that your map has no links to the major ecozones except the outline borders of the US. I’ll be darned if I can pick out if your Rockies are the high Rockies or??? Are you trying to tell me Adobe Mesa is the Llano Esticado and the High Plains? Sorry but I took this by culture being the important aspect since your clients were looking at modern humans.
Again, this is meant to define regions based on a combination of human and physical factors. The Adobe Mesa borders are driven as much by % Hispanic population as they are by topography and climate. I readily admit my map is lacking. To me it was important to produce the conceptual framework and gather feedback before producing a publication-quality atlas. I appreciate your notion that I should produce overlay maps to define/defend/justify chosen borders.
One question. You assumed culture would trump physical geography because we’re “modern humans”? I don’t quite understand. Are you implying climate and other environmental factors are less important thanks to our many technological advances? If that is indeed your position I disagree wholeheartedly. If I’m mistaken please clarify.
In any case, thanks again for your input. Best, Justin