InsightsSite SelectionStrategy

Choosing a Great Location for Your New Coffee Shop

By August 11, 2016 September 18th, 2018 No Comments

Let’s say you’ve developed a new coffee shop brand that is a direct competitor to Starbucks. You strongly believe that your product and execution strategy are better than Starbucks, and your locations won’t suffer from the “brand fatigue” that has supposedly affected Starbucks. You’ve come up with the catchy name “Buckstars” for your new venture.

You’re ready to select locations for your next stores, and, by definition, you know that locations that are good for Starbucks, are good for Buckstars.

How would you go about finding the best location?

You might follow a strategy like this:

Site Selection, circa late 1900’s
(1) Identify customer demographics
(2) Find neighborhoods high in those demographics
(3) Profit

Let’s see how that strategy works out in central Austin, TX.

The first step of this strategy is to identify demographics that match well with Starbucks Buckstars. You do your brand research and find that there is a strong consensus about the “typical Buckstars” customer. That typical customer base is thought to be younger, highly-educated-high-income females. Great, you’ve completed Step 1.

Now that you’ve identified your target customer types, how would you go about finding people like that? Perhaps you’d look up some data from the US Census Bureau; you’d end up with total population statistics for various demographic groups. You might be able to find zip-code or county-level information with that tells you the number of females in your target age range, income ranges, as well as the number of people holding advanced degrees.

But now you’re stuck. Those numbers are tallied at the country level (or, if you’re lucky, at the zip-code or census-tract level). How do you get those numbers for smaller geographic regions (for instance, neighborhoods or block groups)? Even if you could get those for smaller regions, there is an even bigger problem – those numbers for females, education and income are often tallied separately. How do you combine all of the numbers into a useful metric for evaluating a particular neighborhood?

What you really want to do is to identify a particular combination of demographic attributes, and locations that have the highest densities of people with those attributes. You want the most desireable people, per acre. Luckily, you find a great website that helps you with this – IdealSpot provides an online site-selection tool that lets you find high concentrations of your target demographics.

Eager to get to Step 3, you create your IdealSpot account and set up your business. IdealSpot’s software automatically identifies your existing Buckstars’ locations, as well as identifying lots of potential competitors (including Starbucks). You set your demographics filters to “Female Population: ages 25-39” and “Household Income: $100K+” (you’ve left off Education for now). The map instantly shows you the “hot spots” in the city where the confluence of both factors is highest:

There are definite “hotspots” on this map – I’ve dropped several pins – one in downtown Austin, one in Hyde Park, one in the “up and coming” east Riverside area, and one in Cherrywood. Three have higher densities of highly-educated females in our age range, and the Cherrywood location is chosen because that area of town has been gentrifying quickly and the demographics are good.

Now let’s add Education to our demographic filters and see how we can improve our site selection. So, in addition to the Female Population and Household Income filters, we turn on the Education demographic filter and set it to “Bachelor’s Degree to Doctorate”:

Heatmap of central Austin displaying areas with the highest densities of women, aged 25-39, in households earning more than $100K and holding a Bachelor's Degree or higher.

Heatmap of central Austin displaying areas with the highest densities of women, aged 25-39, in households earning more than $100K and holding a Bachelor’s Degree or higher.

Luckily, we see that the scene hasn’t changed much; many more areas became darker, but our original choices are still pretty good; no areas are really standing out as better than them. This is probably due to the correlation between income and education, but that’s a topic for another post. However, the “hot” spots don’t look as hot as before. Is this the best we can do with demographics? Can we go to Step 3 yet?

At this point, we need to ask ourselves:


How good is our assumption that our demographic correlates with Buckstars demand? How do we know if, in a particular area, our target demographic are even interested in Buckstars or even coffee shops? Also, just because our target demographic lives in a particular area, does this mean they actually want their coffee shop experience in their home neighborhoods?


Luckily, IdealSpot knows the answers to these questions. For over 500 different category interests – including coffee shops, IdealSpot collects and aggregates data from social networks. Using our “Social Interest” tool, you can map real-time interest based on the local population’s social media signals. So, across the entire town, we spot where coffee shop lovers are living, working and enjoying their leisure time. Unlike demographics, this interest is not tied to the place someone calls their “home”.

Here we’ve turned off the Demographic map layers, and turned on the “Cafes, Coffee and Tea Houses” Social Interest layer:

Heatmap of central Austin displaying areas with the highest social interest in coffee shops.

Heatmap of central Austin displaying areas with the highest social interest in coffee shops.

Uh-oh. Two of our target locations – downtown and east Riverside – have mediocre social interest in coffee shops. We can take them off our list. But both the Hyde Park and Cherrywood locations show above-average-to-strong demand. Already, IdealSpot’s Social Interest data feeds are showing their value. We can “see” where the demand is strong throughout the day; we don’t have to rely on tying someone’s home location to their interest. We don’t have to guess where the demand is! Perhaps more importantly, we can rule out locations that are risky – they don’t have strong demand for our product.

So, given our filtering on demographics and Social Interest, we’ve identified a couple neighborhoods that exhibit strong demand for our new coffeeshop. It’s obvious that we’re now following a new site-selection strategy:

Site Selection, circa 2016, using IdealSpot Social Interest(tm)
(1) Identify customer demographics
(2) Find neighborhoods high in those demographics.
(3) Find neighborhoods that also have high interest in our product.
(4) Profit

Now are we ready to move onto (the new) Step 4?

Not yet.

We don’t know what each neighborhood’s retail landscape looks like. For instance, we don’t know if there is an oversupply of coffee shops, or an undersupply of “feeder” stores in the area that often drive traffic to coffee shops (clothing and fashion accessory stores come to mind…).

How do we identify areas that have too few or too many coffee shops? Once again, IdealSpot has the answer.


The Data Science team at IdealSpot has tracked the interactions of over 4,000,000 restaurants and retail stores in the nation. We have found both the brand synergies and incompatibilities that promote – or limit – the common good of surrounding business. The result of this effort is a single number that is unique to every neighborhood and business category. This number is called “Market Gap”, and it tells you how much opportunity remains for newcomers to a neighborhood in a particular business category.

For instance, there might be high demand for sushi in an area, but there might be too many sushi places there already. Or demand is high but there are too many stores that run counter to the “sushi experience” – pawn shops and cheap liquor stores, for example. Or there are a lot of sushi places in an area, but there are not enough to meet the demand generated by the shoppers at surrounding businesses – more are needed. Market Gap takes all of this into account, identifies which categories are out-of-balance, and by how much.

Armed with Market Gap, we can now decide between our last two locations:

Heatmap of central Austin displaying areas with the highest opportunity (blue) or oversaturation (red) for coffee shops.

Heatmap of central Austin displaying areas with the highest opportunity (blue) or oversaturation (red) for coffee shops.

Market Gap is showing us that both of our two target neighborhoods could use some more coffee (and that Downtown definitely has enough coffee!). So, even though we’re not seeing a strong Market Gap difference between our top two choices, both choices are good ones. Our confidence that we’re on the right path has increased. Demand is high in both locations, and there is a retail gap for coffee that needs to be filled. Our site-selection strategy has further evolved:

Site Selection, circa 2016, using IdealSpot Social Interest(tm) and Market Gap(tm)
(1) Identify customer demographics
(2) Find neighborhoods high in those demographics.
(3) Find neighborhoods that also have high (daytime or nighttime) interest in our product.
(4) Find neighborhoods that have a Market Gap that our business can fill.
(5) Profit

I hope this post was helpful to you – if you’d like try out our products, sign up at

Marc Smookler

About Marc Smookler

Marc Smookler has founded 6 companies—2 of which have been acquired and 3 of which are market leaders in their respective spaces—the leading brick-and-mortar retail analytics company (, a leading online retailer (, and a cutting-edge marketing services platform ( Marc’s companies have generated over $300M in lifetime revenues and sold over 150,000 products worldwide.

Leave a Reply

Your on-demand data and mapping platform

With our easy-to-use, all-in-one platform, anyone can get actionable location intelligence within 5 minutes of signing up.

Request a demo