Engage swarm · Email-orchestration agent · Per-location list- segmentation skill · Build pillar · Published August 30, 2026
How to build per-location list segmentation that does not go stale
A segment built on Friday using last-90-day RFM math is partially stale by Tuesday. Members made new purchases over the weekend; attended events triggered behavioral cohort moves; geofence entries fired during Saturday foot-traffic events; demographic data refreshed when the Census ACS dropped a new release. The Define + Refresh + Gate + Audit skill bundle on the email-orchestration agent sits above Klaviyo + Mailchimp + Iterable + Braze + Customer.io + Salesforce Marketing Cloud + ActiveCampaign + Bloomreach + HubSpot ESPs + Segment + mParticle + Tealium + Snowplow + RudderStack CDPs + Hightouch + Census + Polytomic reverse-ETL and writes a centralized segment-of-truth with staleness detection, per-criterion refresh cadence, suppression propagation, and named regulatory anchors in every audit record — ECOA Regulation B disparate-impact, EU AI Act Article 22, Robinson- Patman, cannabis state board, HIPAA marketing authorization, TCPA + FCC 24-18 + 10DLC + CASL, CCPA right to opt out of automated decision-making, and the rest of the stack.
The 4-skill bundle on the email-orchestration agent
Define
Centralized segment-of-truth above per-vendor segment fragmentation. Catalogs 14 standing segmentation strategies (RFM + CLTV-based + persona archetype + propensity + lifecycle stage + engagement + geofence + demographic + behavioral + lifecycle-event + NPS / CSAT cohort + subscription tier + loyalty tier + cross-banner / cross-location). Per-segment semver versioning + effective-date + sunset-date + rollback + per-criterion lineage. Reverse-ETL via Hightouch + Census + Polytomic + Hevo + Workato + Tray.io + n8n writes per- destination shape translation across Klaviyo + Iterable + Braze + Customer.io + Mailchimp + Meta + TikTok + Google Customer Match + LinkedIn + Snap + Pinterest + Reddit. Per- banner namespace + cross-banner reconciliation.
Refresh
Staleness detection (time-since-recompute + behavioral-change detection via Bayesian changepoint PELT / binary segmentation / CUSUM / EWMA + demographic-update detection across Census ACS / Claritas PRIZM / Experian Mosaic / Esri Tapestry + master-record-update propagation + stale-segment warning at N / 2N / 3N day thresholds). Per-criterion freshness (RFM monetary on 90-day rolling + persona on annual + geofence on real-time + propensity on model-version-aware + NPS on quarterly + LTV on monthly + subscription tier on status- change + loyalty tier on tier-transition + engagement on hourly + behavioral cohort on real-time). Refresh cadence (real- time + hourly + daily + weekly + monthly + on-event + on- criterion-change). FBC feedback loop writes back into Define.
Gate
Five anchors before any segment writes downstream. ECOA Regulation B disparate-impact + Fair Housing + FCRA (segment cannot use or proxy for protected class). CAN-SPAM + TCPA + FCC 24-18 + 10DLC + CTIA + CASL 36-month + Federal DNC + state DNC + TSR. CCPA + CPRA right to opt out of automated decision- making + 17-state + WA My Health My Data + Texas SCOPE + GDPR Article 6 + 7 + 9 + 17 + 22 + LGPD + DPDP + PIPEDA + Quebec Law 25 + COPPA. HIPAA 45 CFR 164.508 + 164.514 + FINRA Rule 2210 + 3110 + SEC Rule 206(4)-1 + state bar + state professional licensing + FDA DSHEA + OPDP + DEA + 12-state cannabis board matrix + alcohol TABC + CalABC + DISCUS tied- house + tobacco FDA + state lottery + Robinson-Patman + FDD Item 12 + 17 + 19. Illinois BIPA + Texas CUBI + Massachusetts AG v Copley Advertising 2017 + Tennessee ELVIS Act + EU AI Act Article 22 + 26 + 50 + 13 + 14 + 15 + Digital Services Act + Digital Markets Act + ADA Title III + WCAG 2.2 AA + EAA EN 301 549 + NIST AI RMF + ISO 42001 + per-vendor LLM zero- retention. Policy-as-code via OPA Rego + AWS Cedar + Casbin + Cerbos + Oso + Styra DAS + Permit.io.
Audit
Per-segment WORM record at every Define commit and every Refresh recompute. Storage: AWS S3 Object Lock + Azure Blob immutable + Google Cloud Storage Bucket Lock + Wasabi WORM. Retention stacks (longest applicable wins): 7-year FTC + 7- year IRS + 7-year HIPAA medical record + 7-year state-bar record + 6-year SEC + 3-year FINRA 4511 + 3-year FINRA Rule 3110 social-media supervision archive + 36-month CASL suppression + 3-year Illinois BIPA biometric retention + GDPR Article 30 + SOC 2 CC7 / CC8. End-to-end replay rewinds Define + Refresh + Gate + suppression propagation + reverse- ETL trace with confidence tier and explainability at every stage.
The real vendor ecosystem this sits above
ESPs + CDPs + reverse-ETL
Klaviyo + Mailchimp + Iterable + Braze + Customer.io + Salesforce Marketing Cloud + ActiveCampaign + Bloomreach + HubSpot + Drip + Brevo + Constant Contact + Cordial + Emarsys + Selligent Marigold + Twilio Engage ESPs; Segment + mParticle + Tealium + Snowplow + RudderStack CDPs; Hightouch + Census + Polytomic + Hevo + Workato + Tray.io + n8n reverse-ETL.
Upstream data sources
Per-location master record (sibling build-pillar); runtime- readable behavioral cohorts (sibling); per-member LTV math (sibling); POS (Square + Toast + Clover + Lightspeed Retail + NCR Aloha); loyalty (Smile.io + LoyaltyLion + Yotpo + Annex Cloud + Punchh + Paytronix); ecommerce (Shopify + BigCommerce + WooCommerce + Magento); CRM (Salesforce + HubSpot + Dynamics 365 + Zoho + Pipedrive); GBP / GSC impressions; foot traffic (Placer.ai + SafeGraph + Foursquare); reservation (OpenTable + Resy + Tock + SevenRooms); booking (Square Appointments + Vagaro + MindBody + Booker); geofence (OneSignal + Bluedot + Radar + Foursquare Pilgrim + Reveal Mobile).
Policy-as-code + WORM + sibling skills
OPA Rego + AWS Cedar + Casbin + Cerbos + Oso + Styra DAS + Permit.io policy-as-code expresses every Gate rule. AWS S3 Object Lock + Azure Blob immutable + Google Cloud Storage Bucket Lock + Wasabi compliance WORM holds the per-segment audit substrate. Sibling skills on the same email- orchestration agent: per-location dynamic content (segment definitions drive per-render dynamic-content live-read at send-time); lifecycle flow architecture; send-time optimization; push / SMS / direct-mail channel extension.
The 6-workstream reporting cycle
Numeric uplift commitments are not made up-front. The engagement ships a pre-engagement baseline across six workstreams; the cycle tracks delta against that baseline. Reporting is the substrate, not the promise.
- Define coverage. Per-strategy coverage across the 14 standing segmentation strategies; per-segment lineage completeness; per-destination shape-translation coverage across the 12 standing reverse-ETL destinations; per-banner namespace + cross-banner reconciliation completeness.
- Refresh quality. Per-criterion freshness adherence; staleness-detection latency; behavioral-change detection sensitivity; per-segment-class refresh-cadence adherence; FBC feedback-loop recalibration cadence.
- Gate quality. Per-anchor evaluation completeness (ECOA + Fair Housing + FCRA + CAN-SPAM + TCPA + FCC 24-18 + 10DLC + CASL + CCPA opt-out + GDPR Article 22 + HIPAA + FINRA + Robinson-Patman + cannabis state board + Illinois BIPA + EU AI Act Article 22 / 50); per-anchor pass / fail / route-to-counsel distribution; ECOA disparate-impact audit cadence; cannabis state-board posture by state.
- Audit quality. Per-segment WORM record completeness; retention-window coverage (longest of 7-year FTC + 7-year IRS + 7-year HIPAA + 7-year state-bar + 6-year SEC + 3-year FINRA 4511 + 3-year FINRA Rule 3110 + 36-month CASL + 3-year Illinois BIPA + GDPR Article 30 + SOC 2 CC7 / CC8); end-to-end replay success rate.
- Compliance posture. Suppression-propagation completeness across bounce + spam + unsubscribe + state DNC + Federal DNC + GDPR Article 17 + CCPA opt-out + HIPAA opt-out + CASL 36-month; CCPA right-to-opt-out-of-automated-decision- making coverage; GDPR Article 22 transparency stamp coverage; ECOA disparate-impact analysis cadence; cannabis state-board posture by state; EU AI Act Article 50 disclosure coverage on AI-driven segment behavior.
- Audit-trail completeness. Per-anchor regulatory citation completeness; sibling-handoff pointer completeness into the email-orchestration bundle (per-location dynamic content + lifecycle flow architecture + send-time optimization + channel-extension siblings) and into the customer-graph agent (deterministic + probabilistic hybrid identity resolution + cross-touchpoint identity resolution + behavioral cohorts + versioned customer history DSAR).
Frequently asked questions
What is per-location list segmentation that does not go stale — and what is the staleness-by-Tuesday problem?
A segment built on Friday using last-90-day RFM math is partially stale by the following Tuesday. Members made new purchases over the weekend; attended events triggered behavioral cohort moves; geofence entries fired during Saturday foot-traffic events; demographic data refreshed when the Census ACS dropped a new release. The four-skill bundle on the email-orchestration agent — Define, Refresh, Gate, Audit — sits above your existing Klaviyo + Mailchimp + Iterable + Braze + Customer.io + Salesforce Marketing Cloud + ActiveCampaign + Bloomreach + HubSpot + Drip + Brevo + Constant Contact + Cordial + Emarsys + Selligent + Twilio Engage ESP surface and your Segment + mParticle + Tealium + Snowplow + RudderStack + Hightouch + Census + Polytomic + Hevo CDP / reverse-ETL substrate, and writes a per-location segment-of-truth with staleness detection, refresh cadence per criterion, suppression propagation, and named regulatory citations preserved in the audit trail.
Why do Klaviyo + Mailchimp + Iterable + Braze + Customer.io + Salesforce Marketing Cloud break at multi-location segmentation scale?
Each ESP ships a per-tenant flat-segment primitive — the marketer authors "Engaged 30d + LTV greater than $200 + Located in Tampa DMA" in Klaviyo, the segment recomputes when used in a flow, and the platform treats it as freshly valid until the operator manually refreshes. None coordinates segment-of-truth across the 40+ upstream data sources (per-location master record + behavioral cohorts + per-member LTV + POS + loyalty + ecommerce + CRM + GBP / GSC + foot traffic + reservation / booking + geofence). None applies a per-criterion freshness model (RFM monetary on 90-day rolling vs persona archetype on annual refresh vs geofence on real-time vs propensity on model-version-aware). None enforces ECOA Regulation B disparate-impact analysis when segment construction proxies for protected class via ZIP code or surname or grocery-basket content. None coordinates suppression across CAN-SPAM master-suppression + state DNC + Federal DNC + GDPR Article 17 erasure + CCPA opt-out + HIPAA opt-out + CASL 36-month + cross-platform propagation. The four-skill bundle Define + Refresh + Gate + Audit sits above the ESP and CDP surface — it does not replace it. Define establishes the segment-of-truth with versioning + lineage. Refresh runs staleness detection + per-criterion cadence. Gate enforces the regulatory anchors. Audit writes a per-segment WORM record.
What does Define do — segment-of-truth + per-criterion lineage + reverse-ETL destination translation?
Define establishes a centralized segment-of-truth that sits above the per-vendor segment fragmentation (Klaviyo segment vs Iterable list vs Braze segment vs Mailchimp tag vs Customer.io segment vs Salesforce Marketing Cloud data extension vs HubSpot list). It catalogs the 14 standing segmentation strategies — RFM Recency / Frequency / Monetary; CLTV-based per-LTV-quintile or decile; persona archetype (Claritas PRIZM + Experian Mosaic + Esri Tapestry + Spatial.ai PersonaLive + Acxiom Personicx); propensity scoring (purchase / churn / reactivation / upgrade / downgrade); lifecycle stage (acquisition / activation / repeat / retention / reactivation / offboarding); engagement (active / dormant / at-risk / churned); geofence (per-store trade-area + per-DMA + per-ZIP + per-drive-time isochrone); demographic; behavioral; lifecycle-event; NPS / CSAT cohort; subscription tier; loyalty tier; cross-banner / cross-location. Each segment carries per-semver versioning + effective-date + sunset-date + rollback + per-criterion lineage (per-source-record trace + per-feature derivation + per-rule history + per-model version + per-data-quality snapshot). Reverse-ETL via Hightouch + Census + Polytomic + Hevo + Workato + Tray.io + n8n writes the segment-of-truth into each downstream destination with per-destination shape translation (Klaviyo segment + Iterable list + Braze segment + Customer.io segment + Mailchimp tag + Meta Custom Audience + TikTok Custom Audience + Google Customer Match + LinkedIn Matched Audience + Snap Custom Audience + Pinterest Tag Audience + Reddit Custom Audience). Per-banner segment-namespace and cross-banner / cross-location reconciliation prevent drift.
What does Refresh do — staleness detection + per-criterion cadence + FBC feedback loop?
Refresh runs staleness detection across three coordinated subsystems and applies per-criterion cadence. Staleness detection: time-since-last-recompute against per-segment-class N-hour threshold; behavioral-change detection (per-N-percent cohort shift + Z-score threshold + Bayesian changepoint via PELT or binary segmentation + CUSUM + EWMA); demographic-update detection (Census ACS new release + Claritas PRIZM update + Experian Mosaic update + Esri Tapestry update + Spatial.ai PersonaLive update + Acxiom Personicx update); master-record-update propagation from the customer-change-event-emission sibling skill; stale-segment warning (amber at N days, red at 2N days, suppress-from-send at 3N days). Per-criterion freshness: RFM monetary on 90-day rolling; persona on annual; geofence on real-time; propensity on model-version-aware; NPS on quarterly; LTV on monthly; subscription tier on status-change; loyalty tier on tier-transition; engagement on hourly; behavioral cohort on real-time. Refresh cadence: real-time (geofence + behavioral cohort + event-driven), hourly (engagement + propensity score), daily (lifecycle stage + RFM + active vs dormant), weekly (LTV quintile + persona archetype), monthly (demographic baseline + Census ACS), on-event (purchase + return + churn + birthday + anniversary + graduation + relocation + marriage + divorce + new baby), on-criterion-change. FBC feedback loop (per-realized vs predicted segment stability + per-realized vs predicted staleness rate + pattern learning + multi-arm bandit regret + recalibration) writes back into Define.
What does Gate do — ECOA Regulation B + EU AI Act Article 22 + Robinson-Patman + cannabis state board + HIPAA + CAN-SPAM + TCPA + 10DLC + CASL anchors?
Gate evaluates five operationally distinctive anchors before any segment writes downstream. Anchor 1 (the most operationally distinctive): ECOA Regulation B disparate-impact + Fair Housing Act disparate-impact + FCRA when segment affects credit decisioning — segment construction cannot use or proxy for protected class (race + color + religion + national origin + sex including sexual orientation and gender identity + marital status + age + receipt of public assistance + good-faith exercise of Consumer Credit Protection Act rights); disparate-impact analysis required when ZIP code, surname, grocery-basket content, streaming preferences, or other proxies correlate with protected class. Anchor 2 (outbound channel + suppression discipline): CAN-SPAM Act (10-day opt-out honor window + accurate header information + clear unsubscribe + physical postal address); TCPA 47 USC 227 + FCC Declaratory Ruling FCC 24-18 March 2024 + prior-express-written-consent + 8am–9pm recipient-time-zone quiet hours + revocation honored on any reasonable channel; 10DLC A2P Campaign Registry + CTIA Messaging Principles + Common Short Code Carrier; RCS business messaging; CASL 36-month suppression; Federal DNC + state DNC + Telemarketing Sales Rule. Anchor 3 (automated-decision + privacy): CCPA + CPRA right to opt out of automated decision-making + 17-state comprehensive privacy + Washington My Health My Data Act 2024 + Texas SCOPE Act 2024 + GDPR Article 6 lawful basis + Article 7 consent + Article 9 special categories + Article 17 right to erasure + Article 22 automated decision profiling (when segment drives automated offer-eligibility + tier assignment + price personalization + credit decisioning) + LGPD + DPDP + PIPEDA + Quebec Law 25 + COPPA 15 USC 6501 for under-13 segments. Anchor 4 (vertical-specific): HIPAA marketing authorization 45 CFR 164.508 + 164.514 de-identification + no PHI in segment export; FINRA Rule 2210 + FINRA Rule 3110 social-media supervision + SEC Rule 206(4)-1 Marketing Rule + SEC Reg S-K + state bar advertising + state professional licensing when financial / legal; FDA DSHEA + FDA OPDP + DEA Schedule II-V; 12-state cannabis board matrix (Massachusetts CCC + California DCC + Colorado MED + Oregon OLCC + Washington WSLCB + Nevada CCB + Illinois CCSL + New Jersey CRC + New York OCM + Arizona DHS + Alaska AMCO + Maine OMP — most boards limit or prohibit cannabis-product-affinity segmentation and segment-driven offer-eligibility); alcohol TABC + CalABC + SLA + DISCUS Code tied-house; tobacco FDA prohibition; state lottery; Robinson-Patman Act when segment drives price discrimination; FDD Item 12 territorial-protection per FTC Franchise Rule 16 CFR 436 + FDD Item 17 + Item 19. Anchor 5: Illinois BIPA + Texas CUBI when biometric proxies; Massachusetts AG v Copley Advertising April 2017 when geofence segment touches healthcare facilities; Tennessee ELVIS Act 2024 + 11-state deepfake-law matrix; EU AI Act Article 22 automated decisions + Article 26 deployer obligations + Article 50 transparency for AI-generated content + Article 13 + 14 + 15; Digital Services Act Article 30 + Article 26 trader identity; Digital Markets Act; ADA Title III + WCAG 2.2 AA + EAA EN 301 549; NIST AI RMF + ISO 42001 + ISO 27001 + SOC 2 Type II; per-vendor LLM zero-retention verified per call. Policy-as-code expression via OPA Rego + AWS Cedar + Casbin + Cerbos + Oso + Styra DAS + Permit.io.
What does Audit do — per-segment WORM + reverse-ETL trace + end-to-end replay?
Audit writes a per-segment WORM record at every Define commit and every Refresh recompute: per-segment ID + per-banner pointer + per-location pointer + per-canonical list-data-source snapshot + per-vendor source-contribution snapshot (Klaviyo / Mailchimp / Iterable / Braze / Customer.io / Salesforce Marketing Cloud / ActiveCampaign / Bloomreach / HubSpot / Drip / Brevo / Cordial / Emarsys / Selligent / Twilio Engage / Segment / mParticle / Tealium / Snowplow / RudderStack / Hightouch / Census / Polytomic) + segmentation-strategy snapshot + staleness-detection snapshot (time-since-recompute + behavioral-change detection + demographic-update + master-record-update propagation + stale-segment warning state) + per-criterion freshness snapshot + refresh-cadence decision + FBC feedback-loop snapshot + segment-of-truth + segment versioning + segment lineage + reverse-ETL trace (per-destination shape translation across Klaviyo + Iterable + Braze + Customer.io + Mailchimp + Meta Custom Audience + TikTok + Google Customer Match + LinkedIn + Snap + Pinterest + Reddit) + cross-banner / cross-location reconciliation + per-banner namespace + per-anchor Gate decision with evidence + suppression propagation (bounce + spam + unsubscribe + state DNC + Federal DNC + GDPR Article 17 + CCPA opt-out + HIPAA opt-out + CASL 36-month) + sibling-handoff pointers. Storage on AWS S3 Object Lock + Azure Blob immutable + Google Cloud Storage Bucket Lock + Wasabi compliance WORM. Retention stacks (longest applicable wins): 7-year FTC substantiation + 7-year IRS + 7-year HIPAA medical record + 7-year state-bar record + 6-year SEC + 3-year FINRA 4511 + 3-year FINRA Rule 3110 social-media supervision archive + 36-month CASL suppression record + 3-year Illinois BIPA biometric retention schedule + GDPR Article 30 records of processing + SOC 2 CC7 / CC8. End-to-end replay rewinds Define + Refresh + Gate + suppression propagation + reverse-ETL trace with confidence tier and explainability at every stage. Sibling handoffs flow into the per-location dynamic content sibling skill on the same agent (segment definitions drive per-render dynamic-content live-read at send-time), lifecycle flow architecture, send-time optimization, push / SMS / direct-mail channel extension, the per-member LTV math build-pillar, the master record build-pillar, the runtime-readable behavioral cohorts build-pillar, the customer change-event emission build-pillar, the versioned customer history DSAR build-pillar, the cross-touchpoint and deterministic + probabilistic hybrid identity-resolution build-pillars, and the routing build-pillars (five-destination routing + borderline routing + multi-stream severity routing + routing-audit-trail + FBC-override-learning).
Engage Completions on the email-orchestration bundle
The Define + Refresh + Gate + Audit four-skill bundle ships as the orchestration layer above your existing ESP + CDP + reverse-ETL + POS + loyalty + ecommerce + CRM + GBP + foot-traffic + reservation + booking + geofence surface. ECOA Regulation B disparate-impact + CAN-SPAM + TCPA + FCC 24-18 + 10DLC + CASL + CCPA opt-out + GDPR Article 22 + HIPAA marketing authorization + FINRA + Robinson- Patman + cannabis state board + Illinois BIPA + Massachusetts AG v Copley Advertising 2017 + EU AI Act + NIST AI RMF anchors are preserved in every per-segment audit record. Tier 1 AI Readiness Assessment scopes the bundle in two to three weeks; Tier 3 Fractional CMO with AI Swarm operates the bundle end-to-end.
Related reading
- Per-location list segmentation (parent commercial pillar — buyer-outcome framing)
- Per-location dynamic content for multi-location communications (sibling on the same email-orchestration agent — consumes segment definitions at send-time)
- Real-time customer change-event emission (upstream substrate that feeds Refresh)
- Versioned customer history for DSAR compliance (Audit substrate for GDPR Article 17 right-to-erasure propagation)