CRM Implementation Checklist

Capture firmographic and behavioral criteria for the ICP plus the 3–5 segments sales and marketing will route against (enterprise vs. mid-market, new logo vs. expansion, vertical splits). Reuse the segment names already used in GA4 and the MAP — drift between systems is the root cause of broken attribution later.

Whiteboard the lifecycle: lead → MQL → SAL → SQL → opportunity → closed-won. Note who owns each transition and the SLA. PMM, demand-gen, and sales ops should all sign off on the definitions before any field gets configured.

List every system that will read or write contact / account data — web forms, MAP (HubSpot, Marketo, Pardot), CDP (Segment, mParticle), enrichment (ZoomInfo, Clearbit), event tools, Slack alerts. Note the system of record per field; conflicting source-of-truth is the #1 cause of CRM data rot.

Score Salesforce, HubSpot CRM, Pipedrive, and Dynamics against the must-have list: native MAP integration, custom-object support, API limits, role-based permissions, sandbox availability. Disqualify on missing must-haves before optimizing on price.

Required when EU contacts are in scope. Pull the vendor DPA, list sub-processors, confirm data-region (EU vs. US hosting), and route through legal. File the executed DPA in the privacy folder so audit requests don't trigger a scramble.

Build lifecycle stage, lead source, MQL date, SAL date, and SQL date as governed picklists — not free-text. Match the source / medium / campaign convention to the team's UTM doc so reports tie out to GA4 and the ad platforms.

Wire HubSpot / Marketo / Pardot into the CRM with two-way sync. Confirm sync direction per field (CRM-wins vs. MAP-wins), exclusion of sandbox addresses, and that consent / unsubscribe state replicates both ways within 10 minutes.

Implement scoring (demographic + behavioral) and round-robin or territory routing. Flag the MQL threshold and the SAL bounce-back path so a missed SLA escalates to the SDR manager rather than dying silently in a queue.

Capture utm_source, utm_medium, utm_campaign, utm_content, utm_term, gclid, fbclid, and the GA4 client_id on every form. Persist first-touch and last-touch separately. This is what unblocks multi-touch attribution downstream.

Submit a test form, watch it move through MQL → SAL → SQL with the routing, scoring, and Slack alerts firing. Test a re-engaged contact path and an unsubscribe path. Document the test cases so regression testing has a baseline.

Triggered when UAT did not pass cleanly. Triage defects by severity, fix in sandbox, re-run the failing scenarios, and re-validate consent and routing paths before promoting again.

Run role-based sessions: SDRs on lead views and dispositions, AEs on opportunity stages and forecasting, marketers on campaign objects and UTMs. Record sessions and post the SOP in Notion / Confluence for new hires.

Reconcile CRM pipeline by source against GA4 and the ad platforms. Variance over ~10% usually points to a UTM convention break or a duplicate conversion event firing. Walk the WBR with demand-gen and RevOps.

Run a duplicate report on email + domain, check required-field fill rates, and review the bounce / unsubscribe deltas since launch. Schedule an automated weekly hygiene job rather than doing this by hand each quarter.