Market Saturation Analysis: How to Measure Supportable Units

A market can have customers, traffic, income, and strong category demand, and still be saturated for your brand.

Saturation usually appears while the growth story still looks healthy. New units open. Revenue rises. The map fills in. The board deck still shows runway.

Then the unit economics start moving the wrong way. New-store cohorts ramp below plan. Existing locations give back volume. Delivery zones overlap. Franchisees complain about encroachment. AUVs flatten. The company keeps adding locations, but the market stops adding enough incremental value.

Market saturation analysis is the discipline of finding that point before the lease is signed.

In site selection, a market is saturated when the next location's incremental network contribution falls below the operator's threshold after accounting for cannibalization, competition, access, capacity, unit economics, and operating cost.

Saturation is a marginal-return question. The decision turns on how much incremental contribution the next unit creates after transfer, cost, and risk.

Retail site selection software should make that marginal-return test visible. The useful output is not just a site score; it is a site approval brief that explains whether the next unit creates enough net-new value for the network.

1. What is market saturation analysis?

Market saturation analysis evaluates whether a market, corridor, trade area, service area, delivery zone, or franchise territory can still absorb additional locations without weakening the network.

For a multi-unit operator, saturation analysis helps answer:

• How many additional units can this market support?
• Is the next site likely to create net-new demand or mostly transfer demand from existing units?
• Are new stores ramping below historical expectations?
• Are same-store sales weakening in mature markets?
• Are competitors expanding, stable, closing, or consolidating?
• Is the market saturated for the current prototype but still viable for a smaller, faster, cheaper, or channel-specific format?
• Should the operator open, pause, reformat, relocate, close, or wait?

A market can be saturated for one prototype and under-served by another. A full-size restaurant may fail the saturation gate while a smaller pickup format passes. A dense urgent-care corridor may be overbuilt for general visits but under-supplied for a specific service line.

Saturation is a portfolio condition. It cannot be evaluated from a single site score alone.

2. Saturation vs white space vs cannibalization

Saturation analysis belongs in the same strategic system as white space analysis and cannibalization analysis.

White space identifies possible growth. Cannibalization measures internal demand transfer. Saturation determines whether the next unit clears the operator's hurdle.

A market can still contain demand and be saturated. A candidate can project strong gross sales and still be a weak network decision. A dense market can appear overbuilt and still be rational if density creates measurable operating advantages.

3. The core pattern: growth can mask declining incrementality

Most saturation mistakes do not begin with a weak market. They begin with a strong market that keeps accepting locations after the best opportunities are already taken.

The early openings work. The brand gains share. The operator adds more units. Revenue rises because the footprint grows. At some point, the next unit still generates sales, but those sales come from weaker sources:

• customers who already visited nearby stores
• delivery orders reassigned from existing kitchens
• patients who would have used another clinic in the same system
• members who transfer from one club to another
• franchise customers split between too many territories
• competitor demand that is harder and more expensive to win
• low-frequency customers with weaker lifetime value
• marginal trade areas with worse access, rent, labor, or visibility

A supportable-unit claim is really a marginal-return claim. If a company says a market can support ten more stores, it is saying units one through ten can still clear the hurdle. If a company begins closing stores, relocating units, or changing formats, it is usually admitting that some part of the footprint no longer clears that hurdle.

4. Case studies: how saturation appears in real networks

Saturation rarely appears in filings under one clean label. It shows up as store closures, lower new-store productivity, same-store sales pressure, franchisee economics, channel overlap, branch consolidation, delivery-radius conflict, or a shift from unit growth to existing-unit performance.

Starbucks: saturation as a cohort and format problem

Starbucks' 2008 correction remains one of the clearest restaurant examples of footprint rationalization. In its fiscal 2008 annual report, Starbucks described a plan to close approximately 600 underperforming company-operated stores in the U.S. market.

The site-selection lesson is not just that Starbucks closed stores. Saturation often appears first in the newest cohort. Later openings may enter trade areas already covered by earlier openings, shrinking the gap between gross sales and incremental network value.

Starbucks' 2025 restructuring shows a different version of the same discipline. In its Q4 FY25 results, the company reported 627 stores closed in Q4 FY25 as part of its Back to Starbucks restructuring plan, including 584 in North America.

Krispy Kreme: revenue growth masking unit-level decay

Krispy Kreme is a clean example of topline growth hiding unit-level weakness. The early-2000s expansion combined rapid store growth, franchise complexity, and broader distribution through additional channels.

The site-selection lesson is the gap between system growth and unit productivity. A company can keep opening stores, increasing distribution points, or pushing more product through a network while the underlying unit economics weaken.

Krispy Kreme's more recent history reinforces the same principle in a different channel. In its Q1 2025 results, the company said it was reassessing the McDonald's deployment schedule, would no longer pay quarterly cash dividends, and was withdrawing prior full-year guidance amid macro softness and uncertainty around that deployment.

Subway: franchise overbuild and weak unit economics

Subway shows how saturation can persist in a franchise system for years. QSR Magazine reported that Subway lost a net of 631 U.S. restaurants in 2024 and finished the year with 19,502 domestic units, below 20,000 for the first time in about 20 years.

The franchise lesson matters. A franchisor can benefit from footprint scale and royalty streams while franchisees experience lower AUVs, weaker margins, and more same-brand competition. Franchise saturation analysis should include franchisee economics, protected-area logic, transfer risk, and store-level cash flow.

Domino's: fortressing as deliberate density

Domino's gives dense-market expansion an operator vocabulary word: fortressing. In its 2024 10-K, Domino's describes fortressing as increasing its presence in existing markets by condensing delivery areas for better service and adding locations closer to carryout customers.

The same filing warns that building additional stores in markets where Domino's already operates may negatively impact existing-store sales and could lead to closures if executed too rapidly. That is the right way to think about density: network benefits have to exceed transfer, rent, labor, and operating complexity.

Sweetgreen: channel-specific saturation

Sweetgreen's 2024 10-K is a precise example of channel-level saturation language. The company says new restaurants in or near existing markets could hurt sales at existing restaurants, especially in highly concentrated markets. It also says Sweetgreen may selectively open restaurants near existing restaurants operating at or near capacity, while warning that cannibalization may become significant if delivery radii overlap.

For delivery-heavy concepts, saturation can appear by channel before it appears by store count.

Chipotle: AUV step-down during continued expansion

Chipotle is useful as a current growth-counterexample because it shows why saturation analysis should read new-unit growth, same-store sales, and average restaurant sales together. In its Q1 2026 results, Chipotle opened 49 company-owned restaurants and reported a return to positive comparable restaurant sales of 0.5%. At the same time, average restaurant sales stepped down from $3.186M in Q1 2025 to $3.094M in Q1 2026, and management guided to about flat comparable restaurant sales for the full year.

The AUV step-down is the more useful saturation signal. A high-growth operator can return to comp growth while existing-unit productivity quietly steps lower, which is exactly why a saturation brief should separate macro pressure, cohort ramp, same-brand transfer, market-level incrementality, and whether openings expand reach or divide existing demand. This does not prove Chipotle is saturated. It shows what the model should watch.

Pharmacy chains: saturation and access gaps can coexist

Pharmacy closures show why saturation analysis must be local. AP reported that nearly 30% of U.S. drugstores operating during the prior decade had closed by 2021, with closures disproportionately affecting Black, Latino, and low-income neighborhoods.

This is not a simple too-many-stores story. A chain may rationalize an overbuilt footprint while vulnerable neighborhoods lose access. One corridor can be saturated while another becomes under-served.

Bank branches: de-saturation and re-saturation inside one network

Banking shows the same pattern in a different category. Bank of America announced in 2025 that it planned to open more than 150 new financial centers across 60 markets by the end of 2027, while continuing to invest in and reshape its branch network.

That is saturation and white space operating simultaneously. A bank can close or consolidate branches in mature corridors while opening in growth markets where physical presence still improves deposit gathering, advisory relationships, or customer acquisition.

Wal-Mart: density can be rational

Thomas Holmes' NBER paper on Wal-Mart is the academic counterweight to simplistic saturation thinking. Holmes found that Wal-Mart maintained a dense, contiguous store network during its rollout and inferred the value of density economies from the sales cannibalization Wal-Mart was willing to sustain from closely packed stores.

This does not mean every chain should accept overlap. Wal-Mart's density economics were tied to its logistics, distribution, and big-box operating model. The broader lesson is that saturation analysis should measure network economics, not just proximity.

5. The history of saturation analysis

Retail saturation and supply-demand balance

The classic retail planning concept is the Index of Retail Saturation, often expressed as:

IRS = (Households x annual category expenditure per household) / retail facilities

Retail facilities can mean square footage, stores, selling area, or category capacity depending on the source and use case. A high IRS suggests more spending per unit of retail supply. A low IRS suggests too much supply chasing the same demand.

The idea is useful for screening, especially when comparing markets in the same category. It is weak as a final approval tool because it does not model trade areas, access, competitors, cannibalization, customer choice, channel mix, or unit economics.

Analog methods and customer spotting

The operator-side ancestor of modern site selection is the analog method: compare a candidate location to existing locations with similar trade-area characteristics. Saturation enters the analog method when comparable stores share high same-brand density, weaker new-store cohorts, higher overlap, lower AUVs, or heavier demand transfer.

Gravity, Huff models, and spatial interaction

Huff-style models estimate the probability that demand from an origin chooses a location based on attractiveness, travel cost, and competing alternatives.

Pij = (Aj^alpha x Dij^-beta) / sum(Ak^alpha x Dik^-beta)

Esri describes the Huff model as a spatial interaction model in which shopping probability depends on distance, site attractiveness, and the distance and attractiveness of competing sites.

For saturation analysis, the model compares demand allocation before and after the next unit enters the market. If the new unit mostly reallocates probability from existing same-brand stores, the market may be approaching saturation.

6. The formulas operators actually need

No single formula can define saturation. A useful model combines several diagnostics.

1. Store density

Store density = stores / target population

Example:
12 stores / 500,000 target residents = 2.4 stores per 100,000 target residents

Store density is useful for comparing markets. It should not approve or reject sites by itself. A better version uses target households or target consumers instead of generic population.

2. Index of Retail Saturation

IRS = (Households x annual category spend per household) / retail facilities

Example:
Households = 80,000
Annual category spend per household = $650
Retail facilities = 120,000 sq ft

IRS = (80,000 x $650) / 120,000
IRS = $433 per sq ft of category supply

3. Supportable units

Supportable units = capturable market demand / target annual unit volume

Additional supportable units =
net-new capturable demand available to the brand / target AUV

The stronger version subtracts demand already captured by existing units and adjusts for competitor capture, transfer risk, and unit economics.

4. Cannibalization-adjusted net-new demand

Net-new demand = candidate gross demand - demand transferred from existing locations

Candidate gross demand = $2.4M
Expected transfer from existing units = $900K

Net-new demand = $1.5M
Cannibalization rate = $900K / $2.4M = 37.5%

A site can have attractive gross sales and still fail the saturation test if net-new demand is too low. For the mechanics, see the cannibalization analysis guide.

5. Existing-store impact

Existing-store impact =
transferred demand from existing store / existing store baseline sales

Store A baseline sales = $3.5M
Expected transfer to candidate = $420K

Existing-store impact = 12.0%

6. Marginal cash-on-cash return

Marginal cash-on-cash return =
incremental annual cash contribution / net investment

Incremental annual contribution = $450K
Net investment = $2.5M

Cash-on-cash return = 18%

7. Cohort AUV decay

Cohort AUV ratio = new-store cohort AUV / mature-store market AUV

Recent cohort AUV = $1.9M
Mature market AUV = $2.6M

Cohort AUV ratio = 73%

New-store cohorts often reveal saturation earlier than mature stores. If each new wave ramps lower, the market may be running out of high-quality trade areas.

8. Demand-weighted overlap

Demand-weighted overlap = overlapping target demand / candidate target demand

This is stronger than area overlap because a small overlapping area can contain dense, high-value demand, while a large overlapping area may contain little relevant demand. The same distinction is central to trade area analysis.

9. Incremental network contribution

Incremental network contribution =
new unit contribution
- lost contribution from existing units
- added fixed and operating costs
+ measurable network benefits

Open if incremental network contribution >= hurdle.

This is the cleanest definition of saturation.

7. How to measure market saturation and competition

Market saturation is measured by comparing remaining capturable demand with current supply, competitor coverage, trade-area overlap, recent cohort performance, same-store sales pressure, and the marginal contribution of the next unit. The approval question is whether unit N+1 creates enough net-new network value after cannibalization, cost, and risk.

A strong saturation analysis uses multiple measures because each one catches a different failure mode.

A practical saturation review follows this sequence:

1. Define the market unit: corridor, trade area, DMA, delivery zone, franchise territory, or service area.
2. Measure demand: population, spending, visits, patients, members, workers, vehicles, or orders.
3. Measure supply: own units, competitors, square footage, seats, provider hours, delivery capacity, bays, appointment slots, or branches.
4. Measure overlap: trade-area overlap, delivery overlap, demand-weighted overlap, commute overlap, patient-flow overlap, or membership overlap.
5. Estimate transfer: how much of the candidate's volume comes from existing units?
6. Calculate marginal return: does the network get stronger after opening?
7. Define validation: what will be measured after opening?

8. Location-allocation, market share, and capacity models

Saturation analysis can use formal location-allocation methods. Esri's location-allocation documentationdescribes coverage, capacity, and market-share problem types that are useful for portfolio planning.

Coverage, capacity, and market share are separate signals. A strong saturation model uses them together.

9. The saturation gate

A saturation gate is a required approval checkpoint before a candidate advances.

Example saturation gate, illustrative

A useful saturation gate does not just say high risk or low risk. It shows the unit economics, transfer assumptions, affected stores, and approval conditions.

10. When dense markets still work

Dense markets can still support growth when the new unit solves a measurable network problem.

Fortressing benefit =
delivery time savings
+ added carryout demand
+ competitor defense
+ capacity relief
- same-brand transfer
- added fixed cost

Domino's and Wal-Mart show why dense networks cannot be dismissed automatically. Both examples reinforce the same discipline: density needs a measurable benefit.

11. Category differences

Saturation behaves differently by category. The model should match the way customers choose, travel, order, receive care, and repeat.

12. Data sources and limitations

A defensible saturation model uses several data layers.

Existing network data

Useful fields include location, opening date, prototype, sales, visits, transactions, AUV, average ticket, contribution margin, rent, labor cost, channel mix, daypart mix, customer origins, delivery zones, capacity, same-store sales trend, cohort performance, and closure or relocation history.

Customer and transaction data

POS data, loyalty records, delivery addresses, patient ZIP codes, appointment records, membership addresses, CRM data, and first-party customer origins are the strongest inputs for measuring transfer after openings.

Demographic and household data

ACS data is a common foundation for population, household, income, employment, housing, and commute variables. It is authoritative, but it is survey-based and should be paired with behavioral or transaction evidence where possible.

Workforce and commuter data

LEHD data is useful for categories shaped by daytime population, commuting, and workplace density. The brief should note that LEHD products are tabulated and modeled administrative data with nonsampling-error caveats.

Competitor, mobility, real estate, and operating data

Competitor locations, format, hours, delivery coverage, reviews, openings, closures, mobility or visit estimates, rent, buildout cost, parking, ingress, zoning, delivery feasibility, supply-chain reach, and field management coverage all affect whether a market is truly saturated for the current prototype.

The brief should also name data providers and vintages where possible. Mobility, vehicle traffic, zoning, POI, and real estate layers can vary sharply by source and coverage area; unattributed layers are useful for screening but weaker as approval evidence.

13. Saturation inside a Geod-style brief

Geod's core score should stay simple. The methodology page defines the score as a weighted linear model using Reach, Demand, Competition, and Accessibility. Saturation belongs beside that score as a Strategize or portfolio overlay.

The saturation overlay answers a specific committee question:

Is this candidate strong because it reaches valuable demand,
or is it strong because it sits inside demand we already capture?

That distinction should be explicit before approval. It belongs in the same explainable decision record as the site brief, maps, methodology, component scores, assumptions, and validation plan.

14. Post-opening validation

A saturation model earns credibility after the opening. Pre-opening analysis should define what will be measured once the site is live.

For major openings, stronger causal designs may be appropriate: matched-market comparisons, difference-in-differences, or synthetic controls. The point is choosing a credible comparison group so the team can distinguish cannibalization from seasonality, macro softness, competitor changes, or operational issues.

15. What to do when a market is saturated

Saturation does not always mean stop opening forever. The response depends on the diagnosis.

A practical sequence for a 30-500 location operator:

1. Add a saturation gate to every site brief.
2. Review markets with three or more existing units.
3. Compare recent cohort AUV against mature-market AUV.
4. Identify high-overlap candidates before approval.
5. Define a cannibalization budget by category and format.
6. Build a closure, relocation, and reformat playbook.
7. Validate every major opening against forecast.

16. Common mistakes in saturation analysis

Mistake 1: Treating stores per capita as the whole answer

Stores per capita is a useful screen. It misses trade areas, commuters, category spend, channel behavior, competitors, and unit economics.

Mistake 2: Using the metro as the unit of analysis

Saturation is usually local. One corridor can be overbuilt while another part of the same metro remains under-served.

Mistake 3: Confusing saturation with cannibalization

Cannibalization is demand transfer around a specific opening. Saturation is the market condition where additional openings produce weak incremental returns.

Mistake 4: Confusing saturation with operational underperformance

A store may underperform because of staffing, pricing, hours, service, merchandising, or local execution. A saturation model should separate structural location problems from operating problems.

Mistake 5: Ignoring competitor exits

Competitor closures can create de-saturation. A market that looked overbuilt two years ago may become viable after rivals leave.

Mistake 6: Ignoring channel saturation

Delivery, pickup, drive-thru, dine-in, appointments, and in-store visits can saturate differently.

Mistake 7: Ignoring cohort performance

Recent openings often reveal saturation earlier than mature stores. Watch AUV and contribution by vintage.

Mistake 8: Hiding the cannibalization budget

Every dense-market strategy needs an explicit transfer tolerance. If the team cannot state the acceptable level of transfer, the approval logic is incomplete.

Mistake 9: Treating saturation as permanent

Markets can de-saturate through demographic change, competitor exits, format shifts, operational improvements, or new channels.

Mistake 10: Skipping validation

A saturation model that is never compared with post-opening outcomes becomes a permanent assumption.

17. FAQ

What is market saturation in site selection?

Market saturation occurs when the next location in a market is unlikely to create enough incremental network value after accounting for cannibalization, competition, capacity, access, and operating cost. The cleanest test is whether unit N+1 clears the operator's marginal-return hurdle.

How is saturation different from cannibalization?

Cannibalization is the demand transferred from existing same-brand locations to a new location. Saturation is the broader market condition where additional locations produce declining incremental returns, often because cannibalization becomes unavoidable or competitor supply is already high.

How is saturation different from white space?

White space asks where valuable demand is under-served. Saturation asks whether another unit can still profitably serve the market. A market can contain white space for one format while being saturated for another.

What is the Index of Retail Saturation?

The Index of Retail Saturation compares category demand with retail supply. It is useful for early market screening, but it should not replace trade-area, competition, cannibalization, and unit-economic analysis.

How do you calculate supportable units?

A simple supportable-unit estimate divides capturable market demand by target AUV. A stronger version subtracts current network sales, expected competitor capture, and likely same-brand transfer before dividing by target AUV.

How do you know when to stop opening locations?

Pause or stop opening when the next unit's projected incremental contribution, net of expected transfer, falls below the hurdle rate. Other warning signs include declining cohort AUVs, high demand-weighted overlap, weak same-store sales in otherwise healthy markets, and competitor closures.

Can a saturated market become unsaturated?

Yes. Markets can de-saturate when competitors close, population grows, work patterns change, a new format becomes viable, existing locations close or relocate, or demand shifts by channel.

Is overlap always a saturation signal?

Overlap is a risk signal, not a verdict. Domino's fortressing strategy and Wal-Mart's density economics show that overlap can be rational when the network benefits exceed the transfer cost.

What is a saturation gate?

A saturation gate is a site-approval checkpoint that evaluates trade-area overlap, expected transfer, cohort performance, same-store sales trend, competitor density, marginal contribution, and strategic network benefits before a new unit is approved.

What data is needed for saturation analysis?

Useful data includes existing-store performance, POS records, customer origins, trade areas, delivery zones, ACS demographics, LEHD workforce data, competitor locations, foot traffic or mobility data, real estate economics, same-store sales trends, cohort AUVs, and post-opening validation results.

How often should saturation analysis be updated?

High-growth operators should refresh saturation analysis quarterly for active markets and after every major opening, closure, competitor move, format change, or material shift in same-store sales.

18. Glossary

Market saturation

A market condition where additional locations produce weak or negative incremental returns after accounting for cannibalization, competition, capacity, access, and operating cost.

Supportable units

The estimated number of additional locations a market can absorb while meeting performance thresholds.

Index of Retail Saturation

A ratio comparing retail demand with retail supply, commonly expressed as households times category expenditure divided by retail facilities.

Cannibalization

Demand captured by a new location that would otherwise have gone to an existing same-brand location.

Net-new demand

Demand that expands the network rather than shifting from existing locations.

Demand-weighted overlap

Trade-area overlap measured by relevant demand, such as population, spending, visits, orders, patients, members, or households.

AUV

Average unit volume, often used to compare store sales productivity.

Cohort AUV

Average unit volume for a group of stores opened in the same period.

Marginal contribution

The incremental contribution created by the next unit after subtracting transfer and added costs.

Fortressing

A Domino's term for adding stores in existing markets to condense delivery areas and move closer to carryout customers, while accepting some risk of existing-store sales impact.

Density economies

Operational or strategic benefits from a dense network, such as distribution efficiency, delivery speed, brand presence, or field management leverage.

Location-allocation

A class of models that choose facility locations and allocate demand to them based on objectives such as coverage, capacity, market share, or travel cost.

Huff model

A probabilistic spatial interaction model that estimates customer choice based on location attractiveness, travel cost, and competing alternatives.

Saturation gate

A site-brief checkpoint that determines whether another unit can clear the operator's marginal-return threshold.

De-saturation

A market shift that restores room for profitable growth, often through competitor exits, population growth, format changes, or network rationalization.

19. Conclusion

Market saturation analysis makes expansion discipline measurable.

Before a candidate advances, the site brief should answer five questions:

1. How much demand remains available?
2. How much of the candidate's volume would transfer from existing units?
3. Are new-store cohorts still ramping to threshold?
4. Are competitors expanding, stable, or consolidating?
5. Does the next unit's incremental contribution clear the hurdle?

Strong markets can be saturated. Dense markets can still support growth. The difference is marginal network value.

Open when the next unit clears the hurdle. Reformat, relocate, close, or wait when it does not. Measure the result after opening, then update the model.

Related reading

• Trade area overlap and cannibalization
• White space analysis
• Drive-time trade areas
• Defensible site selection score
• Site selection scoring models

References

• Domino's 2024 Form 10-K
• Starbucks fiscal 2008 annual report
• Starbucks Q4 and full fiscal year 2025 results
• Krispy Kreme Q1 2025 financial results
• QSR Magazine: Subway's U.S. count keeps declining
• Sweetgreen 2024 Form 10-K
• Chipotle Q1 2026 results
• AP: Nearly 30% of U.S. drugstores closed in one decade
• Bank of America branch expansion announcement
• NBER: The Diffusion of Wal-Mart and Economies of Density
• ArcGIS Pro: How Huff Model works
• ArcGIS Pro: Location-allocation analysis layer
• Census Bureau: About the ACS
• Census LEHD data