Implementation Plan: Edge-to-Activation
Purpose: Week-by-week phases, milestones, and dependencies for Edge-to-Activation engagements. Used in SOW scope sections, kickoff decks, and to populate Linear milestones. Service line: Data Subservice: Activation & attribution (legacy offering bucket:
activation-attribution/) Delivery model: fixed-scope (typical Phase 0 + Phase 1) / retainer (Phase 2) Last updated: 2026-04-08
At a Glance
| Phase | Name | Duration | Key output | Gate |
|---|---|---|---|---|
| Phase 0 | Signal Recovery Audit | ~1 week | Edge capture live; discrepancy report; pilot scope agreed | Client sign-off on findings |
| Phase 1 | Pilot / Attribution | ~4–6 weeks | Edge → warehouse → reverse ETL/CDP on pilot paths; testing complete; validation | Client sign-off on Phase 1; agreement to Phase 2 if retainer |
| Phase 2 | Rollout & Maintenance (optional) | Ongoing (retainer) | Full channel rollout; monitoring; runbooks; sustained fixes | Per retainer SOW |
Total duration (Phase 0 + Phase 1, typical): ~5–8 weeks (often compressed when Phase 0 and Phase 1 are combined per SOW — see SOP SOW copy block).
Effort (rough): Managing Lead 16–24h · Senior Engineer / Data 80–120h · PM 12–20h (varies by traffic complexity and warehouse stack)
Prereqs Before Kickoff
- SOW signed; phases in scope confirmed (Phase 0, Phase 1, Phase 2 retainer)
- CDN/Edge access path agreed (Cloudflare, Fastly, or equivalent) with rights to deploy capture logic
- Main domain and relevant subdomains routed through the edge network (e.g. Cloudflare orange cloud) where required — see Implementation Playbook
- Data warehouse access; Edge dataset will be created in the same region/location as existing client datasets (BigQuery/Snowflake — location is immovable after creation)
- Analytics / ad-platform access for reconciliation (GA4, Segment, ad accounts as applicable)
- Client engineering contact for Worker/route conflicts, reverse ETL/CDP, and pixel coexistence
Phase 0 — Signal Recovery Audit
Deploy Edge capture in parallel with the existing stack; quantify signal loss and attribution gaps; align scope for Phase 1 pilot.
Duration: ~1 week (Level 1 in SOP scope table)
DRI: Delivery Lead + Senior Engineer
Week-by-week
| Week | Activities | Deliverable |
|---|---|---|
| Week 1 | Access verification (edge, warehouse, analytics); deploy Edge capture (Worker or equivalent); baseline traffic and conversions; draft discrepancy views | Edge events landing in warehouse; first-pass discrepancy narrative |
Milestone: Signal recovery audit complete
- Output: Discrepancy report (Edge vs client-layer); attribution gap by source/channel; documented pilot scope
- Gate: Client sign-off on findings and Phase 1 scope
- Dependencies: Prereqs above; no blocking Worker/route conflicts unresolved (chain or merge plan if another edge script owns routes)
Phase 1 — Pilot / Attribution
End-to-end pilot: modular Edge implementation, warehouse schema, phased testing (dev → single page → subdirectory → full pilot rollout), reverse ETL or CDP to activation tools, documentation and handoff.
Duration: ~4–6 weeks (Level 2 in SOP; calendar often maps to weeks 2–5 after Phase 0, with testing phases not compressed)
DRI: Senior Engineer + Delivery Lead
Week-by-week (pattern)
| Week | Activities | Deliverable |
|---|---|---|
| Week 1–2 | Edge Worker (or equivalent) modularized and deployed; BigQuery/Snowflake schema and ingestion; dev testing | Events reliably in warehouse; unit/integration tests green |
| Week 3 | Single-page production test; data validation | Signed-off single-page validation |
| Week 4 | Subdirectory or segment rollout; reconciliation vs pixels | Gap analysis aligned with expectations |
| Week 5–6 | Full pilot scope rollout; reverse ETL/CDP wired for pilot paths; runbooks and client enablement | Phase 1 acceptance criteria met |
Adjust cadence per Implementation Playbook — do not skip or compress dev → single-page → subdirectory → full rollout gates without explicit client risk acceptance.
Milestone: Phase 1 pilot complete
- Output: Pilot channels covered; testing framework executed; attribution validation / discrepancy tooling usable by client; documentation and handoff
- Gate: Client sign-off on Phase 1; if Phase 2 retainer is sold, commercial and scope alignment for ongoing work
- Dependencies: Phase 0 sign-off (if Phase 0 ran separately); warehouse and activation tool access stable
Phase 2 — Rollout & Maintenance (retainer)
Expand beyond pilot channels; ongoing maintenance (publishers, pixels, edge changes); reporting integration; defined support and escalation.
Duration: Ongoing (typical commercial minimum 3–6 months — see SOP)
DRI: Senior Engineer + Delivery Lead (rotating with client team over time)
Cadence (typical)
| Cadence | Activities | Deliverable |
|---|---|---|
| Weekly / biweekly | Prioritized rollout backlog; channel-by-channel verification; discrepancy monitoring | Updated runbooks; steady-state health |
| Monthly | Exec-ready summary of coverage, gaps, and major changes | Optional monthly review deck or doc |
Milestone: Retainer operating model in place
- Output: Channel rollout backlog agreed; escalation path documented; reporting hooks into client GTM/analytics rhythms
- Gate: Defined SLA and commercial alignment (separate retainer SOW)
SOW Copy Block (Pasteable)
Timeline: Phase 0 ~1 week; Phase 1 ~4–6 weeks; Phase 2 ongoing retainer when purchased. Phase 0 and Phase 1 are often combined in a single statement of work — align language with the SOP SOW copy block.
Phase 0 — Signal Recovery Audit: Edge capture deployed in parallel with existing tracking; baseline and discrepancy report; attribution gap by source; alignment on pilot scope.
Phase 1 — Pilot: Edge → warehouse → reverse ETL/CDP on agreed pilot surfaces; phased testing and validation; documentation and enablement.
Phase 2 — Rollout & Maintenance: Expand to additional channels; sustained maintenance and monitoring; runbooks and support (retainer).
Client responsibilities: Timely access to edge, warehouse, analytics, and ad platforms; engineering time for route/Worker conflicts and activation tooling; review of deliverables within agreed SLA.
SOW ↔ project plan (per client)
When to use: After the SOW is signed and before Linear tickets are cut.
Canonical template: Copy
sow-project-plan-template.mdintoknowledge/clients/{client}/resources/sow-project-plan.md(or client Notion) per engagement.Atomic chain (offering → execution): PLAYBOOK_DEVELOPMENT_PLAN.md §8.2 — Service line → Subservice → Offering → Phase → Deliverable → Ticket.
Retro / Update Log
| Date | Engagement | What changed |
|---|---|---|
| 2026-04-08 | — | Initial canonical implementation plan for offering folder (aligned with SOP, implementation playbook, Linear template). |
Links
- Offer (pitch): offer.md
- SOP (delivery): sop.md
- Demo: demo.md
- Linear template: linear-template.md
- Implementation Playbook (technical): edge-to-activation-implementation-playbook.md
- Edge-to-Activation Linear skill:
.cursor/skills/edge-to-activation-tickets/SKILL.md(repo root) - Rate card: RATE_CARD.md
- Playbook development plan: PLAYBOOK_DEVELOPMENT_PLAN.md