How many stores can one metro support before the next opening starts taking sales from the last one? A franchisor needs that number before it sells territories or commits to a development schedule. Only a few tools claim to produce it directly, and they get there by different routes.
Modeled Answer or Mapped Answer: How These Tools Reach a Number
Two questions sort the seven. The first is whether the tool returns a modeled answer, meaning a count or ranking from its own model, or a mapped answer, where the team reads demand, competitors and coverage off a map and does the arithmetic itself. The second is who runs it: the vendor, with its own data and analysts, or the team, with an analyst or a spreadsheet.
In list order, the seven fall like this:
- Mapped answer, vendor data: Placer.ai.
- Modeled answer, vendor-run: SiteZeus, Audiense In-Person (Buxton) and GrowthFactor.
- Modeled answer, team-run: Esri ArcGIS Business Analyst and CARTO.
- Mapped answer, team-run: Maptive.
A modeled count arrives faster. A mapped one takes longer, but every input behind it is in plain view. A model trained on other brands’ stores can miss what sets one concept apart, while a mapped count depends on the ratio the team picks, such as people per store.
1. Placer.ai (Mapped Answer, Vendor Data)
Foot-traffic analytics firm Placer.ai supplies the observed-demand check, and it makes no claim to count how many stores a market can support. It measures visits at a brand’s own stores and at competitors’ stores from mobile-device panels, so a team can see how visits per location change across a metro as units open.
If each new store in a metro draws fewer visits than the last, the market is closer to its ceiling than a demographic model might suggest. Void Analysis adds a second view, showing which kinds of retailer a shopping center is missing.
Placer.ai also measures shared audiences, which shows where two units already draw the same shoppers. The count itself is left to the team, and subscriptions are custom-priced.
2. SiteZeus (Modeled Answer, Vendor-Run)
SiteZeus ties territories to store counts. On the SiteZeus homepage, a quote from Dave’s Hot Chicken credits the company’s territory management with showing how many stores each territory could accommodate.
The suite also includes a nationwide white-space analysis that projects revenue potential, plus impact analytics that estimate how much a nearby opening or closing would move sales between stores. Sales transfer, or cannibalization, measures how much of a new unit’s revenue is revenue moved from a sister store, and it is the input that caps any capacity estimate.
The model belongs to SiteZeus, and so does the pricing conversation. The company publishes no pricing page, and every sale goes through a demo. The suite is built for franchise and multi-unit brands.
3. Esri ArcGIS Business Analyst (Modeled Answer, Team-Run)
Esri’s Business Analyst gives an analyst the parts to build a capacity estimate rather than handing one over. Its Huff model estimates how much of a market a site can capture based on distance and attractiveness. Market Penetration shows how deeply existing stores already reach into their areas, and the platform’s overview lists market potential among its uses.
Stack those together and an analyst can work out how many more units a metro can absorb. The catch is tiering. Huff and Market Penetration are not included in Web App Standard, the entry web tier, and Esri publishes no US price for any tier. This is a tool for a team with someone who already works in ArcGIS.
4. Audiense In-Person (Buxton) (Modeled Answer, Vendor-Run)
Audiense In-Person (Buxton) is Buxton’s location intelligence business, now sold under the Audiense brand, with SCOUT as its web platform. It is the only tool here that says outright it will determine how many locations a market can support. Its market optimization analysis covers the full question: the count, where white space exists and how territories should be configured.
The estimate comes from Buxton’s side of the table. The company cites more than 50 data sources and describes its coverage as reaching across the USA. Audiense In-Person also scores sites and forecasts performance, so the count and the individual picks come from the same vendor.
Pricing is shared only after a demo.
5. GrowthFactor (Modeled Answer, Vendor-Run)
GrowthFactor‘s Market Planning ranks white space across metros, meaning markets with room for the brand that no store serves yet. That tells a team which markets to work first, though a ranking is not a unit count. Onboarding imports a brand’s existing locations and configures its scoring model, so the ranking starts from stores the brand already runs.
The platform also plans trade zones, and its cannibalization analysis is listed on the Enterprise plan. Custom revenue forecasts come through GrowthFactor Labs, which generally needs 40 or more mature stores with revenue data. Discovery, a 30-day engagement, is a one-time $5,000 fee. Since September 23, 2026, new customers can’t buy a single seat.
6. CARTO (Modeled Answer, Team-Run)
CARTO supports the build-it-yourself route. An analyst working in the business’s own cloud data warehouse can combine sales, demographic and competitor layers in CARTO’s low-code Workflows and write the capacity model the team trusts. CARTO names Site Selection and Territory Planning among its use cases. Travel-time areas around existing stores can run in the same workflow, and the result stays in the warehouse beside the sales data.
The trade-off is that someone on staff has to build that model, and the estimate is only as sound as their work. Every CARTO plan is sold by quote.
7. Maptive (Mapped Answer, Team-Run)
Maptive won’t compute a market-capacity number, but it supplies the inputs for one. US or Canadian census data can join any radius or drive-time pop-out, so the area around each store shows the population inside it. The Boundary Tool can fill ZIP codes or counties by marker count, turning a list of competitor addresses into counts by region. A POI feature checks those locations against Google Places data.
The team does the arithmetic, for example by reading population per store in markets that already work from pop-outs in Maptive location intelligence software, then applying it to the market under study.
Questions for Any Vendor Selling a Unit Count
A modeled count depends on what went into it, and the demo is the place to find out.
What are the inputs? Ask whether the model uses the brand’s own sales, visit data, demographics or all three, and how recent each one is.
What cannibalization threshold does it assume? The count rises as the model allows more sales to transfer between stores, so ask what share it treats as acceptable and whether that share can be changed.
Then ask whether the model has seen stores like these. A model built on large suburban formats may misread a small urban concept. GrowthFactor’s 40-store floor for Labs forecasts is one public example of how much history a model can need.
A team working from a mapped answer should put the same three questions to its own spreadsheet.