Capture-Demand Swarm · Territory-Analysis-Market-Scoring Agent · Cross-Location-Cannibalization-Detection Skill · Build pillar · Published September 27, 2026

How to build cross-location cannibalization detection for multi-unit franchise + multi-location retail operators

A 4-skill bundle (Measure + Compare + Attribute + Decide) layered above the existing Placer.ai + SafeGraph + Foursquare + UberMedia (Cuebiq) + Near Intelligence + AirSage + StreetLight Data + INRIX mobility ecosystem + the Toast + Square + Clover + Lightspeed + NCR Aloha + Oracle MICROS + Revel Systems POS substrate + the Salesforce Loyalty Management + Punchh + Paytronix + Thanx + Eagle Eye + LoyaltyLion + Smile.io + Yotpo Loyalty + Annex Cloud + Oracle CrowdTwist customer-cohort substrate + the Snowflake + BigQuery + Databricks + Redshift + Synapse + Firebolt + ClickHouse data-warehouse substrate + the Looker + Tableau + Power BI + Mode + Sigma + Hex + Omni + Metabase BI substrate + the Esri ArcGIS + Mapbox + Google Maps Geocoding + HERE + Pitney Bowes Spatial Analytics geocoding substrate. Anchored on FTC Franchise Rule 16 CFR Part 436 + FDD Item 12 territorial rights + Item 19 Financial Performance Representation + per-state Franchise Investment Law in 14 named states + per-state franchise relationship laws in approximately 25 additional jurisdictions + FASB ASC 606 revenue recognition + per-franchisee inter-company accounting + per-vendor mobility + competitor-discovery data licensing + FTC v X-Mode/Outlogic 2024 + FTC v Mobilewalla 2024 + Washington My Health My Data Act 2024 + Massachusetts AG v Copley Advertising 2017 + CCPA + CPRA + state-comprehensive-privacy + GDPR + NIST AI RMF + ISO 42001 + EU AI Act.

Start with the AI Readiness Assessment Engage the Fractional CMO with AI Swarm Take the 3-question scope quiz

The 4-skill bundle on the territory-analysis-market- scoring agent

Cross-location cannibalization detection is one skill on the territory-analysis-market-scoring agent. The skill decomposes into four operationally distinct sub-skills, each with its own success criteria and its own handoff to the next.

1. Measure

Per-location daily revenue from POS (Toast + Square + Clover + Lightspeed + NCR Aloha + Oracle MICROS + Revel Systems + Lavu + Touchbistro), normalized for known confounders (calendar-seasonality + weather + holidays + banner-wide marketing events + promotional periods). Longitudinal panel covering pre-open + post-open windows operator-defined (typical 90 days pre + 90 days post + 12-month monitoring tail). Per-location foot-traffic from Placer.ai + SafeGraph + Foursquare + UberMedia + Near Intelligence + AirSage + StreetLight Data + INRIX per per-vendor data-license posture. Loyalty- program customer-cohort data (Salesforce Loyalty + Punchh + Paytronix + Thanx + Eagle Eye + LoyaltyLion + Smile.io + Yotpo Loyalty + Annex Cloud + Oracle CrowdTwist) for same-customer-different-location pattern tracking.

2. Compare

Build a matched control set of locations far enough away from the new-location trade area to be unaffected but similar enough on relevant covariates (banner format + region + urban-vs-suburban + seasonality + demographic substrate + anchor co-tenancy + age-of-location + pre-open revenue trend) that their post-period trend is a credible counterfactual. Matching methodology operator- defined: propensity-score matching, synthetic control method, difference-in-differences, comparative interrupted time series. Matched set documented + version-pointed for reproducibility. Cannibalization effect estimate produced with explicit confidence interval — never point estimate alone.

3. Attribute

Decompose observed effect into contributing factors: cannibalization from new-location opening (estimated via matched control + cohort overlap when available), competitor-open effect (incorporates Anchor 3 competitor discovery from Google Maps Places + Yelp Fusion + OpenStreetMap), local-construction-or-road- closure effect (state DOT + municipal portals), banner-wide marketing-spend effect (operator marketing-budget registry), other observable factors. AI-driven Attribute (LLM-assisted causality vs correlation disambiguation) runs under NIST AI RMF + ISO 42001 + EU AI Act + per-vendor LLM zero- retention — LLM is NEVER in critical decision path; methodology selection + matched-set construction is the substantive decision.

4. Decide

Decision-grade record: cannibalization effect estimate with confidence interval, contributing- factor breakdown, FDD Item 12 + Item 19 cross- reference, recommended action (no action + monitoring tail + Item 12 mitigation negotiation + Item 19 disclosure refresh + counsel escalation), audit-trail entry. Decide output NEVER auto-publishes; routes to franchise development + franchisor counsel for review.

The real ecosystem this skill sits above

Mobility + foot-traffic + POS substrate

Placer.ai, SafeGraph, Foursquare, UberMedia (Cuebiq), Near Intelligence, AirSage, StreetLight Data, INRIX, Locomizer for mobility + foot-traffic. Toast, Square, Clover, Lightspeed, NCR Aloha, Oracle MICROS, Revel Systems, Lavu, Touchbistro, ShopKeep POS for per- location daily revenue capture.

Customer-cohort + data-warehouse substrate

Salesforce Loyalty Management, Punchh, Paytronix, Thanx, Eagle Eye, LoyaltyLion, Smile.io, Yotpo Loyalty, Annex Cloud, Oracle CrowdTwist customer- cohort. Snowflake, BigQuery, Databricks, Redshift, Synapse, Firebolt, ClickHouse data warehouse for longitudinal-panel analysis.

BI + geocoding substrate

Looker, Tableau, Power BI, Mode, Sigma, Hex, Omni, Metabase BI for franchise-development + counsel- review surfaces. Esri ArcGIS, Mapbox, Google Maps Geocoding, HERE Geocoding, Pitney Bowes Spatial Analytics for trade-area geometry referenced from the same registry as the FDD-territorial-protection- gating skill.

5-anchor compliance overlay

Anchor 1 — FTC Franchise Rule + FDD Item 12 + Item 19 FPR + per-state FIL + per-state franchise relationship laws (operationally distinctive)

Cannibalization detection is the post-open verification that the territorial promise + projected -revenue claim made at site selection still hold. FDD Item 12 (exclusive + protected + non-exclusive + reserved territory classes) defines what each franchisee was sold; when a new same-banner location opens close enough to a sibling franchisee territory and the sibling revenue declines, the affected franchisee can pursue an Item 12 encroachment claim. FDD Item 19 (Financial Performance Representation) covers projected-revenue numbers disclosed to the franchisee at sale; if realized post-open revenue diverges materially from those projections — particularly if cannibalization caused the divergence — Item 19 substantiation gets scrutinized. FTC Franchise Rule (16 CFR Part 436, amended 2007) requires both Item 12 disclosure + Item 19 substantiation; per-state Franchise Investment Laws in 14 named states layer state enforcement; per- state franchise relationship laws in approximately 25 additional jurisdictions govern post-sale conduct. Operationally distinctive — this skill exists to answer the regulator + franchisee question the moment it arrives.

Anchor 2 — FASB ASC 606 + per-franchisee inter- company accounting

When revenue moves between franchisee P&L statements (cannibalization shifts dollars from franchisee A to franchisee B), FASB Accounting Standards Codification Topic 606 (Revenue from Contracts with Customers) revenue-recognition treatment applies at the franchisor consolidated level; per-franchisee inter- company accounting handles the franchisee-level allocation per the operator-counsel-documented policy. The Decide output references the per- franchisee inter-company posture so accounting + franchise-development decisions remain aligned.

Anchor 3 — Per-vendor mobility + competitor- discovery data licensing + FTC v X-Mode/Outlogic + FTC v Mobilewalla

Mobility data from Placer.ai + SafeGraph + Foursquare + UberMedia + Near Intelligence carries per-vendor SDK + ToS license scope. Google Maps Places ToS limits Places-data retention to 30 days except cached place IDs + requires Google Map display when shown to end users. Yelp Fusion API requires attribution. OpenStreetMap data is licensed under ODbL with attribution + share-alike. FTC v X-Mode/Outlogic (2024) + FTC v Mobilewalla (2024) constrain how location data can be sold + used per consumer consent posture. Washington My Health My Data Act (2024) restricts health-related location inference. Massachusetts AG v Copley Advertising (2017) established geofencing-around-sensitive-locations limits.

Anchor 4 — CCPA + CPRA + state-comprehensive-privacy + GDPR

Customer-cohort tracking (same-customer-different -location pattern) is personal information under California Consumer Privacy Act + California Privacy Rights Act + 18 state-comprehensive-privacy statutes + GDPR in EU jurisdictions. DSAR overlay tagging preserves data-subject-access-request fulfillment evidence per cohort record.

Anchor 5 — NIST AI RMF + ISO 42001 + EU AI Act + per-vendor LLM zero-retention

AI-driven Attribute (LLM-assisted causality-vs- correlation disambiguation) operates under NIST AI Risk Management Framework + ISO 42001 + applicable EU AI Act articles + per-vendor LLM zero-retention posture. LLM is NEVER in the critical path of the decision; methodology selection + matched-set construction is the substantive decision. LLM proposal recorded with model + prompt-template + confidence in routing-audit-trail.

6-workstream pre-engagement-baseline reporting cycle

Cannibalization effect estimates are what the data shows after the workflow is built, not numbers Completions promises in advance.

Measure coverage. Per-location POS data completeness, per-location seasonal-normalization adherence, per-location mobility-data-vendor connection, per-vendor data-license posture freshness, per-loyalty-program customer-cohort data integration completeness.
Compare quality. Per-incident matched -control-set construction documentation, matching methodology version, per-covariate balance check, per- matched-set version pointer freshness.
Attribute quality. Per-incident contributing-factor decomposition completeness, per- factor confidence interval, per-causality-vs- correlation disambiguation methodology adherence, per- AI-Attribute LLM-classification accuracy where LLM- assisted.
Decide quality. Per-incident FDD Item 12 + Item 19 cross-reference completeness, per-incident counsel-review routing completeness, per-incident recommended-action rationale capture, per-incident audit-trail entry completeness.
5-anchor compliance posture freshness. FTC Franchise Rule + FDD Item 12 + Item 19 FPR substantiation + per-state FIL across 14 named states + per-state franchise relationship laws + FASB ASC 606 + per-franchisee inter-company accounting posture + per-vendor mobility + competitor-discovery data licensing posture + FTC v X-Mode/Outlogic 2024 + FTC v Mobilewalla 2024 + Washington My Health My Data Act + Massachusetts AG v Copley Advertising 2017 + CCPA + CPRA + state-comprehensive-privacy + GDPR + NIST AI RMF + ISO 42001 + EU AI Act posture.
Audit-trail completeness. Per-incident canonical decision record, per-matched-set construction record, per-Attribute decomposition record.

Frequently asked questions

What does cross-location cannibalization detection actually solve?

A multi-unit franchise or multi-location retail operator opens a new location 2 miles from an existing one. After 90 days the existing location shows revenue down 15 percent. Is that cannibalization (the new location absorbed customers from the old one) or coincidence (a competitor opened across the street + a road closed nearby)? Without a structured detection workflow the operator cannot distinguish the two cases — and the difference matters because cannibalization implicates FDD Item 12 territorial rights when the affected location is a sibling franchisee and Item 19 FPR claim substantiation when the original site-selection projected revenue was published to a prospect. The skill measures revenue trend pre vs post adjacent-location-open at the existing location + at the new location + at a matched control set of locations far enough away to be unaffected; compares against the operator-defined cannibalization-threshold policy; attributes the observed delta to the new-location opening vs other observable factors; produces a decision-grade record for franchisor counsel + franchise development + the affected franchisee.

Why is FTC Franchise Rule + FDD Item 12 + Item 19 FPR the operationally distinctive frame for cannibalization?

Cannibalization detection is the post-open verification that the territorial promise + projected-revenue claim made at site selection still hold. FDD Item 12 (exclusive + protected + non-exclusive + reserved territory classes) defines what each franchisee was sold; when a new same-banner location opens close enough to a sibling franchisee territory and the sibling franchisee revenue declines, the affected franchisee can pursue an Item 12 encroachment claim. FDD Item 19 (Financial Performance Representation) covers the projected revenue numbers the franchisor disclosed to the franchisee at sale; if realized post-open revenue diverges materially from those projections — particularly if cannibalization caused the divergence — Item 19 substantiation gets scrutinized. The FTC Franchise Rule (16 CFR Part 436, amended 2007) requires both Item 12 disclosure and Item 19 substantiation; per-state Franchise Investment Laws in 14 named states (California Corporations Code Sections 31000 et seq, Hawaii FIL, Illinois Franchise Disclosure Act 815 ILCS 705, Indiana FDA, Maryland Franchise Registration and Disclosure Law, Michigan FIL MCL 445.1501 et seq, Minnesota Franchise Act Minn Stat 80C, New York General Business Law Article 33, North Dakota FIL, Rhode Island FIA, South Dakota FIL, Virginia Retail Franchising Act, Washington FIPA RCW 19.100, Wisconsin FIL) layer state enforcement; per-state franchise relationship laws in approximately 25 additional jurisdictions govern post-sale conduct. The Measure + Compare + Attribute + Decide sub-skills emit the per-incident evidence the franchisor and counsel need to defend or settle.

How does the Measure skill capture revenue trend without confounding it with seasonality + external factors?

The Measure sub-skill captures per-location daily revenue from the operator-chosen POS substrate (Toast + Square + Clover + Lightspeed + NCR Aloha + Oracle MICROS + Revel Systems + Lavu + Touchbistro), normalizes for known confounders (calendar-seasonality, weather, holidays, banner-wide marketing events, promotional periods), and assembles a longitudinal panel covering pre-open + post-open windows operator-defined (typical: 90 days pre + 90 days post + a continuing 12-month monitoring tail). For mobility-pattern context the skill pulls per-location foot-traffic from Placer.ai + SafeGraph + Foursquare + UberMedia (Cuebiq) + Near Intelligence + AirSage + StreetLight Data + INRIX per the per-vendor data-license posture (each vendor SDK + ToS records the licensed-use scope; the skill records the per-source attribution + license-version pointer per record). When loyalty-program customer-cohort data is available (Salesforce Loyalty Management + Punchh + Paytronix + Thanx + Eagle Eye + LoyaltyLion + Smile.io + Yotpo Loyalty + Annex Cloud + Oracle CrowdTwist), the same-customer-different-location pattern is tracked directly so cannibalization can be verified at the cohort level rather than inferred at the aggregate level.

How does the Compare skill match locations to a defensible control set?

The Compare sub-skill builds a matched control set of locations far enough away from the new-location trade area to be unaffected by it but similar enough on relevant covariates (banner format, region, urban-vs-suburban, seasonality pattern, demographic substrate, anchor co-tenancy, age-of-location, pre-open revenue trend) that their post-period revenue trend is a credible counterfactual. Matching methodology is operator-defined (propensity-score matching, synthetic control method, difference-in-differences, comparative interrupted time series). The matched set is documented + version-pointed at the time of analysis so the analysis is reproducible. Per-location revenue trend at the affected location vs the matched control set produces the cannibalization effect estimate with explicit confidence interval — never a point estimate alone.

How do the Attribute and Decide skills produce a defensible decision-grade record?

Attribute decomposes the observed effect into contributing factors: cannibalization from the new-location opening (estimated via matched control set + cohort overlap when available), competitor-open effect (incorporates Anchor 3 competitor-discovery substrate from Google Maps Places + Yelp Fusion + OpenStreetMap), local-construction-or-road-closure effect (pulled from state DOT + municipal portals), banner-wide marketing-spend effect (pulled from the operator marketing-budget registry per Anchor 5), other observable factors. AI-driven Attribute (LLM-assisted causality vs correlation disambiguation) runs under NIST AI RMF + ISO 42001 + EU AI Act + per-vendor LLM zero-retention with the LLM never in the critical path of the decision — methodology selection + matched-set construction is the substantive decision. Decide produces the decision-grade record: cannibalization effect estimate with confidence interval, contributing-factor breakdown, FDD Item 12 + Item 19 cross-reference, recommended action (no action + monitoring tail + Item 12 mitigation negotiation + Item 19 disclosure refresh + counsel escalation), audit-trail entry. The Decide output never auto-publishes; it routes to franchise development + franchisor counsel for review.

How does Completions report on this without fabricating KPI commitments?

Pre-engagement baseline is established in the first 30 days. Reporting cycles cover the six workstreams: Measure coverage (per-location POS data completeness + per-location seasonal-normalization adherence + per-location mobility-data-vendor connection + per-vendor data-license posture freshness + per-loyalty-program customer-cohort data integration completeness), Compare quality (per-incident matched-control-set construction documentation + matching methodology version + per-covariate balance check + per-matched-set version pointer freshness), Attribute quality (per-incident contributing-factor decomposition completeness + per-factor confidence interval + per-causality-vs-correlation disambiguation methodology adherence + per-AI-Attribute LLM-classification accuracy when LLM-assisted), Decide quality (per-incident FDD Item 12 + Item 19 cross-reference completeness + per-incident counsel-review routing completeness + per-incident recommended-action rationale capture + per-incident audit-trail entry completeness), 5-anchor compliance posture freshness (FTC Franchise Rule + FDD Item 12 + Item 19 FPR substantiation + per-state FIL across 14 named states + per-state franchise relationship laws + FASB ASC 606 + per-franchisee inter-company accounting posture + per-vendor mobility-data + competitor-discovery data licensing posture + FTC v X-Mode/Outlogic 2024 + FTC v Mobilewalla 2024 posture + CCPA + CPRA + state-comprehensive-privacy + GDPR + NIST AI RMF + ISO 42001 + EU AI Act posture), audit-trail completeness (per-incident canonical decision record + per-matched-set construction record + per-Attribute decomposition record).

Engage Completions

Multi-unit franchise + multi-location retail operators opening new locations within trade-area distance of existing ones face a recurring cannibalization-detection question whose answer determines FDD Item 12 + Item 19 exposure. Completions architects the workflow as a 4-skill bundle layered above the existing Placer.ai + SafeGraph + Toast + Square + Salesforce Loyalty + Punchh + Snowflake + BigQuery + Looker + Tableau ecosystem. Start with the Tier 1 AI Readiness Assessment ($10k, 2-3 weeks), build with the Tier 2 Setup Sprint ($25-50k, 4-8 weeks), or engage Tier 3 Fractional CMO with AI Swarm ($15-25k per month, 6-month minimum).