[ANCHOR] Customer Segmentation
Product Requirements Document · ANCHOR v1.0 · Confluence source of truth: page 51234439381. Repo status: draft keeps the index lint-safe (anchors carry no stories); the live initiative status is ACTIVE.
This is an ANCHOR PRD. It frames the initiative-level problem, personas, north-star metrics, and cross-cutting decisions, and indexes the child PRDs that deliver it. Execution detail (constraints, stories, ACs, operator logic, screens) lives in the child PRDs linked in the Phase Index.
| Field | Value |
|---|
| PM | Zhelia Alifa |
| PRD Version | 1.0 |
| Status | ACTIVE |
| PRD Type | ANCHOR |
| Anchor | Yes — this IS the Anchor |
| Labels | epic:qontak-cdp | module:customers | feature:customer-segmentation |
| Last Updated | 2026-06-26 |
Phase / Component Index
| Component | Goal | PRD Link | Epic Key | Status |
|---|
| Basic Attributes (Strategic — foundation) | Rule-based dynamic segments on default & custom fields + basic events (created/updated at/by), consent, and loyalty attributes; saveable, auto-refreshed, reusable as Broadcast recipients. | prd-customer-segmentation-basic-attributes.md | TF-2544 | In Progress |
| Use-Case Activities | Extend the rule builder with cross-module association conditions (Conversations, Campaigns, Company, Deals, Tickets), advanced nested/exception builder, segment detail + performance, utilize-for-campaign, and the segmentation permission model. | prd-customer-segmentation-use-case-activities.md | TF-2549 … TF-2555 | Draft |
1. One-liner + Problem
One-liner: Dynamic, rule-based customer segmentation in CDP — classify contacts into auto-updating groups from customer attributes and cross-module activity, reusable for targeting and campaigns.
Problem: Qontak CDP users cannot create meaningful, auto-updating customer groups. Today they export contacts to spreadsheets for manual classification or apply static tags individually — hours per week per team, producing stale snapshots and making it impossible to run targeted bulk actions (campaigns, broadcasts) against the right audience. Broadcast recipients stay static and manual, CRM data in CDP (Deals, Tickets, Conversations, Companies) goes unused for audience selection, and CDP remains a data viewer rather than a customer-intelligence platform. The gap is felt most by Growth and Enterprise accounts managing hundreds-to-thousands of contacts across business units.
2. What Happens If We Don't Build This
- CDP stays a data viewer, not an intelligence platform — blocks the 2026 personalization roadmap (Customer Signal Score, Loyalty, Campaign Intelligence) that all build on segmentation.
- Broadcast keeps sending to broad, unqualified audiences (manual CSV selection) → poor campaign performance.
- No competitive parity with modern CDPs (HubSpot, Salesforce, Zoho, Intercom) — Enterprise accounts evaluate competitors.
- Qontak cannot upsell enterprise personalization features gated behind segmentation.
3. Target Users + Persona Context
| Persona | Role | Goal | Pain | Workaround |
|---|
| Primary — Sales / CS Team Lead | Team leader responsible for contact management and campaign targeting in CDP | Maintain up-to-date, rule-defined customer groups (e.g. "Active Leads", "Loyal Customers", "Churn Risk") without manual effort | No dynamic grouping — every list must be rebuilt whenever contact data changes (new deals, resolved tickets, lapsed conversations) | Export contacts to spreadsheet weekly, filter manually, re-upload tags — 2–4 hours per cycle per team lead |
| Secondary — Marketing / Campaign Manager | Runs Broadcast/Campaign using CDP contact lists as audience | Target broadcasts/campaigns to precisely defined audiences (e.g. "open deal > 30 days", "no conversation in 90 days") | Cannot define audience by attribute or CRM activity — must use static lists or send to all | Build audience in external tool (spreadsheet / Mailchimp), upload CSV — does not reflect real-time CDP data |
4. Success Metrics (Initiative-level)
| Category | Metric | Definition | Baseline | Target |
|---|
| Adoption | ⭐ Accounts creating ≥ 1 segment | % of active CIDs that create at least one segment within 30 days of GA | N/A — feature does not exist | ≥ 20–40% of active accounts |
| Efficiency / Impact | Broadcast using segments | % of broadcast recipient lists built from a customer segment (vs manual CSV) | N/A | Growing share quarter-over-quarter post-GA |
| Stickiness | CDP MAC | Monthly active companies using CDP (segmentation drives recurring use) | Current MAC | Increase post-GA |
5. Key Decisions + Alternatives Rejected
5a — Decisions Made
| Date | Decision | Rationale |
|---|
| 2026-Q2 | Rule-based (manual) segmentation for v1 — not AI/ML auto-segments | Foundational, deterministic, explainable; AI/ML is a later phase |
| 2026-Q2 | Daily scheduled re-evaluation (08:00), not real-time on every change | Acceptable freshness at far lower cost/complexity than per-change re-eval |
| 2026-Q2 | Web-only configuration (Qontak One web), not mobile | Config is a power-user, desktop task; mobile out of scope |
| 2026-Q2 | Package-gated capability (Plus / Ultimate / 360) | Monetization: basic fields on Plus; use-case activities on Ultimate/360 |
| 2026-Q2 | Split into two PRDs — Basic Attributes (foundation) + Use-Case Activities | Basic attributes ship first as the engine; association conditions extend it |
5b — Alternatives Rejected
| Alternative | Why Rejected | Date |
|---|
| AI/ML-suggested segments for v1 | Not foundational; needs the rule engine + event data first | 2026-Q2 |
| Real-time segment re-evaluation on every contact change | Cost/complexity far exceeds the freshness benefit at MVP | 2026-Q2 |
| Keep manual CSV / external tools | Stale, manual, no real-time CDP data, no competitive parity | 2026-Q2 |
| Sync segments directly into Broadcast/Campaign audience lists in v1 | Separate integration initiative; deferred to keep MVP scoped | 2026-Q2 |
6. Open Questions
| # | Type | Question | Owner | Deadline |
|---|
| 1 | Assumption | Association data (Deal, Ticket, Conversation, Company linked to a contact) is queryable by the segment evaluation engine without joins that degrade performance beyond acceptable limits — needs engineering validation on real data volume. | CDP Engineering | 2026-06-11 |
| 2 | Open Question | Should segment membership be exportable as a contact list (CSV/XLSX)? If yes, needs an export endpoint + quota governance. Working answer: until CDP has an export feature, segment can be exported via the export initiative. | PM (CDP) | 2026-06-04 |
| 3 | Open Question | Final package-gating matrix per tier (Plus / Ultimate / 360) for basic fields vs use-case activities — confirm with Billing. | PM (CDP) + Billing | 2026-Q2 |
PRD CHANGELOG
| Version | Date | By | Section | Type | Summary |
|---|
| 1.0 | 2026-06-26 | Claude | All | CREATED | ANCHOR PRD created for the Customer Segmentation initiative; indexes the two child PRDs (Basic Attributes + Use-Case Activities); initiative-level problem, personas, success metrics, decisions, and open questions. |