Our methodology

A strong decision shows
its working.

See what drives a shortlist, how much evidence supports it, and what still needs a site visit. Every answer should be open to scrutiny.

Read the SiteFinder sample report

The analysis, step by step

A clear question.
A traceable answer.

This is how SiteFinder moves from a chosen franchise and city to a shortlist. Each stage keeps its assumptions and evidence limits visible.

  1. Set the scope

    Confirm the franchise, the city — any city, anywhere — and available capital. Format requirements establish which local conditions matter to this decision.

  2. Build the local picture

    Read mapped businesses, demographics, access, competition and property evidence in small analysis areas. Label estimates and reported inputs by their source.

  3. Compare candidate areas

    Apply format-specific weights to city-relative signals. Check confidence and saturation, then diversify the shortlist across localities. Read any evidence limits or budget shortfalls alongside the ranking.

  4. Stress-test and verify

    Compare financial scenarios, show the constraints and turn open questions into field checks. Preserve the report's scope, assumptions and version at delivery.

Starting with a property? FitFinder keeps one site fixed and compares the categories and formats it can support. It does not perform a citywide location search.

Follow the method in the sample

From evidence to a better field visit

The method, in practice.

Two real workspace views show how the analysis informs your next checks. Open either screenshot to inspect the original.

SiteFinder / Catchment

Read the trading area, then check the route.

Catchment bands bring mapped businesses, household proxies and competing outlets into the same view. The bands describe approximate analytical areas; confirm actual access, barriers and journey times on the ground.

Explore SiteFinder
SiteFinder CatchmentActual product screen
SiteFinder catchment map around Deccan with inner and outer hexagonal bands and a household, POI and competition inspector. A closer look · original screenshot Enlarge & inspect
Naturals Ice Cream · Deccan. These H3 catchment bands are approximate analytical areas, not road-network travel-time isochrones.

The analytical grid

We divide each city into cells using the H3 spatial grid. The default grid uses H3 resolution 8 (~0.74 km² per cell), with cell area stated as an approximate average; resolution can differ by market. Scores describe the area around a site, not the condition of a particular shop. A cell boundary is not a walking-time or delivery-time boundary.

The signals behind the score

What moves the result.

Different business formats need different surroundings. The analysis weights the relevant signals and explains which drivers raise or lower fit.

Demand and customer context
Residential and workplace activity, nearby businesses, and demand and affluence proxies help describe the customer base. A modelled footfall index is not a measured pedestrian count.
Competition and local draw
Similar formats, same-brand outlets and complementary businesses help distinguish a trading cluster from an overcrowded catchment. Saturation can remove a candidate from the shortlist.
Access and occupancy costs
Road and transit context inform access. The format's space requirements and estimated rent are assessed against modelled sales; the actual unit and lease still need verification.

Two questions, kept separate

Good fit is only
half the answer.

Fit score

How suitable is this location for this format, relative to other locations in the same city?

Fit combines weighted, city-relative signals and affordability checks. A score of 80 is neither an 80th-percentile rank nor an 80% chance of success. Where shown, the separate city percentile tells you how many cells score lower for that format. Scores cannot be compared across cities.

Evidence confidence

How well is the area's mapped activity supported by source coverage, useful detail and survey evidence?

Confidence is not a check of every financial assumption or proof that every shop is current. The evidence share and source labels show the support behind a result. A promising score with thin evidence needs investigation.

The highest fit may not be the recommended site.

SiteFinder orders the shortlist by fit, then checks financial viability under the model's assumptions. A lower-ranked site may have more workable economics. If none meets the model's viability criteria, the analysis says so. Small score gaps still call for a closer look at rent, frontage and territory before choosing a winner.

If no site fits your stated budget, SiteFinder can still show strong fits with their capital shortfalls. Appearing in the results does not mean a site is affordable.

How we test the economics

A scenario you can question.

Financial outputs are Modelled planning scenarios. They are not actual outlet results, brand offers or guaranteed returns.

Start with the assumptions
Sales depend on assumed transactions and average spend. Rent, staffing, margins, royalties and other operating costs shape the result. The full cash-flow model also includes owner pay, tax and equipment replacement assumptions. Check borrowing costs separately.
Test what could change
Conservative, base and upside scenarios change sales and rent assumptions together. Sensitivity checks change one input at a time. These are tests of the assumptions, not probabilities or statistically validated forecast ranges.
Separate earnings from payback
Operating break-even is when monthly operating earnings turn positive. Capital payback is when cumulative modelled cash flow recovers the initial investment. A business can reach the first without reaching the second within the forecast period.

The free financial calculator is a first check. It estimates one month's operating surplus from your inputs. It does not forecast site demand, tax, financing costs, equipment replacement or investment payback.

Try the financial preview

From analysis to the field

What a field visit still needs to establish

  • Actual rent, deposit, lease terms, frontage, floor and access.
  • Observed footfall at relevant times and the competitors currently trading.
  • Franchisor approval, territory availability and nearby same-brand outlets.
  • Delivery access and travel patterns; a hexagonal catchment is not a road-network travel-time boundary.
  • Permitted use, utilities, fit-out requirements and the obligations in the proposed lease and franchise agreement.

The plan supplies analysis and a verification checklist. An on-ground visit, legal review and franchisor approval are not included. Use qualified local advisers for the checks that need them.

Evidence standards

Inspect the evidence standards.

Open a reference below to examine the validation checks, source labels and scope behind the analysis.

Input validation

Before a number earns its place

Checks matched to the source.

Property surveys and franchise submissions use checks suited to their fields and source. Mapped business records are normalised and deduplicated separately. Rejected or review-held submissions remain in the audit record and do not enter scoring; other concerns can reduce their evidence weight.

  1. Admissibility

    Check whether an input is physically possible and whether its units and required fields make sense.

  2. Plausibility

    Compare the value with locality benchmarks where available. Unusual figures can be flagged or given less weight.

  3. Duplicate detection

    Check recent repeated reports of the same property from the same contributor. This reduces double-counting; it does not prove that all reporters are independent.

  4. Cross-field consistency

    Read related facts together. Rent and floor area may each look possible while their ratio does not.

Validation is not independent verification. Unverified customer-reported rent, floor area and deposits can be retained in the record without entering the scoring evidence. These property observations require field verification before they can change that evidence base.

Sources, freshness and missing data

Read the coverage as well as the result

A snapshot, with visible limits.

Sources differ by market
Public mapping, business listings, geographic data, field observations and customer or brand submissions contribute where available. A source used in one city is not necessarily available in another.
Collection timing is not a live check
Competitor freshness is reported at city level, not as a verified date for each shop. A recently processed dataset can still contain an old listing. Confirm that nearby competitors are trading and that a proposed unit is available.
Coverage changes what can be concluded
A strict map filter can return no areas, and FitFinder requires sufficient evidence for the selected site. If SiteFinder's normal filters leave no candidates, it can still show the best-supported areas available. Read their confidence before treating them as suitable. Missing optional exhibits are labelled; a blank map does not establish that there is no competition.
Transparent calculations still need real-world checks
Demand, affluence and revenue can rely on proxies and format assumptions. Model consistency does not establish commercial success. Compare alternatives within the same city and confirm local sales and cost assumptions before committing.
Data-quality labels

Read the label, then the number

What we know. How we know it.

Data labels explain how a figure was produced. Source tiers explain where its supporting evidence came from. A modelled estimate stays an estimate, however polished the chart.

Scroll sideways to see all columns

Four data-quality labels, with a separate state for missing information
LabelMeaningExample
ObservedSupplied by a source or collected in the field.A mapped pharmacy or a staff-measured rent.
DerivedCalculated directly from source data.Competitor density within an analytical cell.
ModelledEstimated using stated assumptions and a model version.A revenue scenario, payback estimate or rent proxy.
User-providedEntered by a customer or brand without independent verification.A quoted rent or declared franchise royalty.
UnavailableRequired data or geometry is missing; the report explains why.An urban-fabric exhibit where building coverage is insufficient.
Five tiers of supporting evidence

Supporting evidence

Five tiers of supporting evidence

Scroll sideways to see all columns

Source provenance is shown alongside the strength and coverage of evidence.
TierSourceWhat it establishes
1Staff / field-verifiedAn accountable field observation.
2Customer-reportedA customer's account of a real property.
3Franchise-providedThe brand's own account of its commercial terms.
4External / publicInformation from a public or external source.
5Platform-estimatedA model estimate where stronger evidence is absent.

The tier describes the source, not a guarantee that every supplied figure is correct. Admissible, sufficiently supported observations can change the calculation before ranking; repeated independent observations carry more weight. Locations without usable survey evidence retain their existing estimates.

Franchise profile sources

Franchise profile sources

What a franchise profile's source tells you

Engine profile
A profile the engine can evaluate, with operating requirements and financial assumptions. Curated profiles use publicly reported benchmarks for the business format; their inputs are not necessarily supplied by the brand.
Marketplace listing
A published opportunity retained with its original source. When used in the analysis, published terms are combined with labelled business-format assumptions for missing inputs. A scoreable scenario is not an independently verified brand offer; confirm current terms and territory with the franchisor.
Franchise-provided profile
A published BrandReach submission contributes the brand's own terms. Supplied values replace relevant defaults, with their source label retained. Publication does not establish independent verification of every figure.
Scope and market readiness

Boundaries make the answer useful

A specific question. An accountable answer.

Scope is confirmed before purchase

SiteFinder: one franchise and one city. FitFinder: one city and one canonical site or analytical cell. A locality name can contain many different cells.

The delivered HTML and PDF are frozen. The report record keeps the decision scope, model and report versions, and run timing. This preserves the delivered analysis; it does not guarantee an identical future calculation after source data or evidence changes.

Readiness is checked at the decision level

Any city in the world can be bought, and every city is analysed with the same collectors and the same checks before its report runs — within 2 days of payment. Each country has its own rent basis, currency and franchise terms; a city is never scored on another market's. FitFinder also needs sufficient evidence for the exact site.

A ready city can still contain sites with insufficient evidence, and every report keeps the confidence label of each figure.

Product screenshot

Read details opens at the key analysis. Scroll to explore; Actual size shows every pixel.

Original product screenshot. Example values and assumptions require verification.