The 4–8 week timeline, week by week
Those phases overlap in practice. Migration prep starts during discovery because data cleanup is the long pole, and training material gets written while testing runs.
| Phase | Typical duration | What actually happens |
|---|---|---|
| Discovery & design | Week 1 | Process mapping, data model, pipeline and lifecycle definitions, integration inventory |
| Build | Weeks 2–3 | Properties, pipelines, workflows, forms, templates, permissions, reporting scaffold |
| Migration | Weeks 3–4 | Extract, clean, map, staged import, field-level reconciliation report |
| Testing | Week 5 | End-to-end record journeys, automation dry runs, integration sync checks |
| Training & adoption | Weeks 6–8 | Role-based training, documentation handover, hypercare and fix cycle |
What actually decides your timeline
The Hub combination matters less than people expect. Marketing Hub plus Sales Hub on clean data is faster than Sales Hub alone on a decade-old Salesforce instance with 40 custom objects.
- Data volume and quality — 70,000 records that are clean move faster than 5,000 that aren't.
- Number of integrations, and whether any are custom-built rather than off-the-shelf.
- Decision latency — the single biggest hidden cost. If sign-off on a lifecycle definition takes nine days, you lose nine days.
- Whether you're migrating from another CRM or starting fresh.
- How many teams use the portal. One sales team is a build; sales, marketing, service and finance is a program.
When it becomes a three-to-six month project
Enterprise migrations with custom objects, regulated data-handling requirements, ERP integrations or multi-region portals genuinely take longer. That's fine — the failure is pretending it's an 8-week job and then change-ordering the client into a corner.
The other long-timeline cause is the rebuild: a portal that was implemented badly, has been running for two years, and now has live workflows nobody dares switch off. Untangling that safely takes longer than starting clean. See our take on the
audit-first approach if that sounds like your portal.
How to compress the timeline honestly
- 01Name one internal decision-maker with authority to sign off on the data model. Not a committee.
- 02Start data cleanup in week zero, before the engagement formally begins.
- 03Freeze scope on phase one. Custom objects and the ABM program can be phase two.
- 04Book the training sessions in the calendar on day one, so they don't slip.
- 05Go live with a reporting scaffold, not perfect dashboards — dashboards get refined once real data flows.
Common mistakes
- Treating go-live as the finish line instead of the start of adoption.
- Migrating dirty data because cleaning it 'would delay launch' — you've just moved the delay downstream.
- Letting the implementation run without a named internal owner.
- Building every workflow before a single record has moved through the pipeline.
- Skipping the reconciliation report and finding out three months later that 4% of records never arrived.