Competitive intelligence for site selection is the disciplined read of who else is competing for the same demand in a specific trade area, and what that leaves for you. It is not a logo map and it is not a market share deck. It is a bounded, pre-commitment estimate: how much category spend exists inside the polygon you actually serve, how much of it existing supply already absorbs, and whether the remainder clears the number your format needs.
The distinction matters because the word “competition” gets used at two very different altitudes. Reading a rival’s national expansion strategy — where they are opening, what formats they are testing, which markets they have quietly abandoned — is a portfolio-planning exercise, and we cover it separately in retail competitor analysis for expansion. This piece is the other altitude: one candidate site, one trade area, one decision, made before the LOI.
Competitive intelligence at the trade-area level comes down to five moves: count direct competitors and near-substitutes (the substitute list is almost always too short), estimate the category spend pool inside your real drive-time polygon, subtract what existing supply plausibly absorbs, decide whether proximity to a strong rival helps or kills you based on whether your category is destination or impulse, and write the competitive assumption down before the forecast so it can be graded when the store opens.
What Is Competitive Intelligence?
Competitive intelligence is the systematic collection and analysis of publicly observable information about rivals in order to make a specific decision better. It is an ethical and legal discipline built on open sources, not espionage. In retail real estate, a competitive intelligence report is the written output: a bounded document that names the competitors in a trade area, rates their strength, states what is known versus assumed, and converts that into an explicit view of available demand.
The failure mode is familiar. A brand runs a radius ring, counts six competitors, declares the market “competitive but viable,” and moves on. Nothing in that sentence can be wrong, because nothing in it is falsifiable. A real competitive read makes a claim precise enough to be graded later: we believe the three nearest rivals capture roughly 55% of category spend inside this polygon, leaving a pool that supports a unit at our average volume. That sentence can be checked against reality twelve months after opening. The vague one cannot.
If your competitive read cannot be proven wrong, it is not intelligence. It is reassurance.
How Do You Do a Competitive Analysis for a Retail Location?
Start with the real trade area, inventory everything inside it that competes for the same wallet occasion, score each competitor on strength and proximity, then estimate what demand is left. The order matters: analysts who start with the competitor list before defining the boundary invariably count stores that never see the same customer, and miss ones just outside an arbitrary ring.
1. Draw the boundary before you count anything
A one-mile ring is a geometry exercise; a drive-time polygon is a behavior model. Rivers, highways, one-way grids and school catchments all distort who can realistically reach a door, which is why trade-area analysiscomes first. Every number downstream — population, spend pool, competitor count — inherits the error in this boundary.
2. Count direct competitors, then count substitutes properly
Direct competitors are easy and almost never the problem. The problem is the substitute list, which is systematically underestimated because teams define their category the way their brand deck defines it rather than the way a customer experiences a Tuesday at 6pm. A fast-casual bowl concept does not compete only with other bowl concepts. It competes with the grocery hot bar, the supermarket meal kit, the two delivery-only kitchens operating out of a ghost facility, the taqueria with a drive-thru, and the customer’s own fridge.
- Occasion test: list the three occasions your customer uses you for, then list everything in the polygon that solves each one.
- Channel substitutes: delivery-only brands, click-and-collect, and subscription boxes consume category spend without appearing on any competitor map.
- Adjacent formats: a large-format operator carrying your category as a department may absorb more of your pool than a same-size rival.
- Weight, don’t binary-flag: a substitute capturing a third of the same occasions should count as a third of a competitor, not zero and not one.
3. Estimate the pool, then subtract existing supply
Convert trade-area population into households, multiply by a defensible category spend per household per year, and you have a spend pool. Then ask the only question that matters: how much of that pool is already spoken for? This is the same arithmetic underlying retail market capacity, applied to one site rather than a whole market. A pool of $48M split across seven effective competitors leaves a very different story than the same pool split across two.
4. Score proximity in the direction your category actually works
Distance to the nearest strong competitor is the single most decision-relevant variable in the whole exercise, and its sign flips by category. More on that below.
When Being Next to a Strong Competitor Helps — and When It Kills You
Proximity to a dominant rival is an asset in destination categories and a liability in impulse categories. In a destination category, the customer plans the trip and comparison-shops, so clustering concentrates traffic and the incumbent has effectively pre-validated the demand for you. In an impulse category, the customer buys whatever is in front of them at the moment of need, so the rival on the near corner intercepts your demand before it ever reaches your door.
Furniture, auto dealerships, bridal, and home improvement cluster for a reason: nobody drives forty minutes to see one sofa. Coffee, convenience, quick-service and car washes behave the opposite way, because the purchase decision is made in the last ninety seconds of a trip someone was already taking. The same 0.4-mile separation that reads as “proven corridor” in the first case reads as “pre-empted” in the second.
Ask: would a customer standing at my door, who knew the competitor existed, still choose me? In a destination category the honest answer is often yes, on assortment, price or experience. In an impulse category, the answer is usually that they never got to your door — which is a traffic and access problem, not a brand problem, and no amount of marketing fixes it after the lease is signed.
Co-tenancy complicates both cases in useful ways. A rival inside a center that drives the trips you depend on may be worth tolerating; the same rival on an isolated pad is pure leakage. Read this alongside whitespace and void analysis, which looks at the inverse question: where demand exists with no supply at all.
Map the Competitive Field for One Candidate Trade Area
Describe a trade area you are actually considering: how many direct competitors sit inside it, how many near-substitutes you can honestly name, how far the nearest direct rival is, how strong that rival looks, whether your category is destination or impulse, and your own spend assumptions. The tool computes the spend pool, an even split, and a strength- and distance-weighted headroom figure, then shows where the demand is going and what verdict the arithmetic supports. Change the category type with everything else held constant — the proximity effect is the thing most competitive decks never model.
Describe one candidate trade area. Every figure below is computed from your own assumptions.
a competent, unremarkable operator
people buy what is in front of them, so a close rival is far more damaging
The share of your modelled spend pool left unclaimed once existing supply is weighted for count, substitution, strength and distance — roughly 17.3% of the pool.
There is room, but not much. Only pursue this trade area with a site that wins on access, visibility or format — a parity site will split a pool that is already thin.
- Households in the trade area
- 16,800 (population ÷ household size)
- Category spend pool
- $5.4M (households × your spend assumption)
- Effective competitor count
- 5.2 (3 direct + 4 substitutes × 0.55)
- Even share, you included
- $867K across 6.2 seats
- Nearest rival pull factor
- 0.60× (strength 1× , 1.2 mi proximity 0.27)
- Weighted headroom
- $927K (17.3% of pool)
Spend-per-household and trade-area population are your assumptions, not measured data. This is a structured way to make a competitive assumption explicit before you forecast — it is not a revenue forecast, and it does not replace one.
What Data Can You Legally Collect About Competitors?
Publicly observable and commercially licensed information is generally fair game: locations, hours, formats, visible pricing, job postings, building permits, liquor and health licenses, public filings, franchise disclosure documents, review volume, and aggregated anonymized mobility data from licensed providers. What is off-limits is misrepresentation, trade secrets, confidential material from current or former employees, and anything obtained by deception. These are general practices, not legal advice; rules vary by jurisdiction and by the terms of the data you license, so confirm unusual collection methods with counsel.
| Source | What it reveals | Reliability |
|---|---|---|
| Store locator & POI data | Footprint, format mix, white space | High — but stale within weeks of a closure |
| Licensed mobility data | Visit counts, dayparts, dwell, where visitors come from | Medium-high for trends; weaker as an absolute level |
| Building permits & licenses | Openings 6–18 months before they appear | High signal, low coverage — varies by municipality |
| Job postings | Staffing levels, new-market intent, format changes | Medium — directional, easily over-read |
| Observed pricing & menus | Positioning, likely ticket, promotional posture | High for price; says nothing about volume |
| Public filings & FDDs | Average unit volumes, build costs, unit economics | High for the chain average; poor for a specific store |
| Search demand data | Category and brand intent at city level | Medium — strong for relative interest, not sales |
| Site visit & lot counts | Peak throughput, queueing, condition of the asset | High for what you saw; a sample of one day |
Search demand deserves a specific note, because it is the one source most retail teams underuse. Semrush’s research puts US “near me” searches at roughly 7.1 million per month and growing 29% year over year, with “near me open now” up 38% and “near me tonight” up 41% — and their keyword database spans 26.7 billion keywords across 142 geographic databases with city-level volume. That makes branded and category search a usable proxy for which rivals own intent in a specific market, as we explore in near-me searches and site selection.
How Do You Find Out a Competitor’s Sales?
For a private operator, you usually cannot, and the professional answer is to model a range rather than invent a number. Combine visit counts from licensed mobility data with an observed price point and a category conversion assumption, sanity-check it against published average unit volumes for comparable chains, and present the result as a band with its assumptions visible. A single confident figure for a competitor’s store-level revenue is almost always either licensed from a provider who modelled it the same way, or made up.
The honest boundary is worth stating explicitly in your report, because unmarked guesses migrate into pro formas and then into board decks. Three things are typically unknowable from the outside: a competitor’s actual store-level P&L, their lease terms and occupancy cost, and their internal plans. Anything built on those should be labelled as an assumption, not a finding — the same discipline that makes new store sales forecasting defensible.
Write the Competitive Assumption Down Before the Forecast
The single highest-return habit in this entire discipline costs ten minutes: before the model runs, write one paragraph stating what you believe about the competitive field and why. Name the competitors you counted, the substitutes you weighted, the share you believe existing supply absorbs, and the one fact that would change your mind. Date it and file it with the site package.
The reason is simple. Once a forecast produces a number, the competitive read gets retrofitted to justify it — every analyst has watched a substitute quietly drop off a list because including it broke the pro forma. Writing the assumption first makes that retrofit visible. It also makes the assumption gradeable: twelve months after opening you can check whether the competitive field behaved the way you said it would, which is the only way a team’s judgment actually improves over time. That feedback loop is the subject of grading site-selection forecast accuracy.
- →The polygon: how the trade area was drawn, and what it deliberately excludes.
- →Direct competitors counted, with distance and a strength rating for each.
- →Near-substitutes counted, with the occasion overlap weight applied to each.
- →The spend pool, the share we believe existing supply absorbs, and why.
- →Whether proximity helps or hurts here, with the category logic stated.
- →The one observation that would change this read.
Competition Is a Number, Not a Mood
The brands that get this right are not the ones with the longest competitor lists. They are the ones that convert the list into an explicit, falsifiable claim about available demand, before capital is committed, and then grade themselves on it. At Locate we build that competitive read into the revenue forecast itself rather than bolting it on as a slide, because the forecast is where the assumption either survives contact with reality or does not — and because the same team that runs the analysis negotiates the lease. If you are weighing a specific trade area and want a second read on what the competitive field actually leaves you, talk to us.
Common Questions
- What is competitive intelligence?
- Competitive intelligence is the systematic collection and analysis of publicly observable information about rival businesses in order to make a better decision. In retail site selection specifically, it means the disciplined read of who else is competing for the same demand in a specific trade area, how strong they are, and what that leaves for you. It is built from legally observable signals — locations, formats, hours, pricing, hiring, permits, mobility data — not from secrets.
- How do you do a competitive analysis for a retail location?
- Draw the real trade area first (a drive-time polygon, not a ring), then inventory every direct competitor and every near-substitute inside it. Score each one on strength, format, distance from your candidate site, and whether it sits on the same traffic pattern. Estimate the category spend pool for the trade area, subtract the demand existing supply plausibly absorbs, and write down what is left as an explicit assumption before you build a revenue forecast.
- How do you find out a competitor's sales?
- You generally cannot know a private competitor's store-level sales, and you should not pretend otherwise. What you can do is estimate a range from observable proxies: visit counts from mobility data, dwell time, parking-lot fullness at peak, staffing levels inferred from job postings, menu or shelf pricing, and published averages for public companies in the same category. Treat the result as a modelled range with stated error bars, not a fact — and never rely on information obtained improperly.
- What data can you legally collect about competitors?
- Anything publicly observable or commercially licensed is generally fair: store locations, store hours, formats and square footage, visible pricing, public filings and franchise disclosure documents, job postings, building permits and liquor licenses, press coverage, review volume, and aggregated anonymized mobility data from licensed providers. What is off-limits is misrepresentation, trade secrets, confidential documents from employees, and anything obtained by deception. Rules vary by jurisdiction and by the terms of the data you license, so confirm with counsel before an unusual collection method.
- How many competitors is too many in a trade area?
- There is no universal number, because the answer depends on the size of the demand pool, not the competitor count. A trade area with 60,000 people and strong category spend can support several units; one with 12,000 can be oversupplied by two. The useful test is the leftover demand pool: estimate total category spend in the trade area, weight existing supply for strength and proximity, and see whether what remains clears the revenue your format needs to pencil.