How to migrate marketing automation to HubSpot without losing pipeline data
QualityPro Hub operations team
Short answer: you protect pipeline data in a marketing automation migration by auditing and cleaning before you move, mapping every field, lifecycle stage and integration, rebuilding only the programs you still need, running the old and new systems in parallel for a period, and checking that leadership reports match before you switch the old platform off.
Who this is for: marketing operations, revenue operations and IT leaders moving from Marketo Engage, Account Engagement (Pardot), Eloqua or another platform to HubSpot, often alongside a CRM change. The same approach applies to migrations into Salesforce, Adobe Experience Cloud or Customer.io.
Why do migrations lose pipeline data?
Data rarely disappears in one big failure. It leaks through small gaps: a field that didn't map, a lifecycle stage that changed meaning, a sync rule that stopped updating the CRM, a form that kept posting to the old system, or attribution history left behind because nobody owned it. Each gap is minor. Together they break the pipeline reports leadership relies on.
Phase 1: What should you audit before moving anything?
- Database: total records, duplicates, unsubscribes and consent status, bounced and inactive contacts.
- Fields: every field in use, who uses it, which reports depend on it. Retire what nobody uses.
- Lifecycle and lead status: definitions of each stage and the rules that move records between them.
- Scoring and routing: current models, thresholds and assignment rules.
- Programs: active nurtures, recurring emails, webinars, forms, landing pages and their owners.
- Integrations: CRM sync, CDP, webinar and event tools, enrichment, ads platforms, data warehouse.
- Reporting: the dashboards leadership uses every week, and the numbers they show today. This is your baseline.
Phase 2: How do you map and clean the data?
- Build a field map: old field, new property, data type, owner, and whether it syncs to the CRM.
- Rewrite lifecycle stage definitions in plain language and get sales and marketing leadership to sign off.
- Deduplicate and normalize key fields (country, state, industry, job level) before import.
- Decide what history you need to keep: campaign membership, first and last touch, original source, and activity for open opportunities at minimum.
- Confirm consent and subscription status migrate correctly. Getting this wrong is a compliance risk, not just a data problem.
Phase 3: What should you rebuild, and what should you leave behind?
A migration is the best chance you'll get to simplify. Rebuild the programs that drive pipeline today. Archive the rest with documentation. Rebuild scoring and routing from the definitions you just agreed on, not by copying old logic line by line. Recreate forms on the new platform and update every page that embeds them, including microsites and event pages that are easy to miss.
Phase 4: How do you cut over safely?
- Run a test sync with a sample of records and check every mapped field end to end.
- Run both systems in parallel for an agreed period, with clear rules about which system owns which programs.
- Compare leadership reports in both systems: new leads, MQLs, pipeline by source and conversion rates. Investigate every difference above an agreed tolerance.
- Switch forms, integrations and sending domains on a planned date, and freeze changes in the old system.
- Keep read-only access or an export of the old platform for historical questions.
Migration checklist
| Area | Done when |
|---|---|
| Baseline reports | Weekly leadership numbers captured from the old system |
| Field map | Every active field mapped or retired, with an owner |
| Lifecycle | Stage definitions signed off by sales and marketing |
| Consent | Subscription and consent status verified after import |
| CRM sync | Test records sync both ways with correct values |
| Forms | Every form on every site points to the new platform |
| Programs | Pipeline-driving programs rebuilt and tested |
| Attribution | Original source and campaign history preserved |
| Reports | New reports match the baseline within tolerance |
| Hand-off | Documentation, runbooks and named owners in place |
Frequently asked questions
How long does a marketing automation migration take?
It depends on database size, the number of programs and integrations, and whether the CRM is changing at the same time. Plan for the audit and cleanup to take as long as the build itself.
Should we migrate the CRM and marketing automation at the same time?
It can save time when both are changing anyway, but it doubles the risk. If you do both, sequence the data model and sync design first and keep one team accountable for both.
Does the same approach work for Salesforce, Adobe Experience Cloud or Customer.io?
Yes. The phases are the same whatever the destination: audit, map and clean, rebuild, run in parallel, then cut over. Moving to Adobe Experience Cloud (Marketo Engage, Adobe Journey Optimizer and Real-Time CDP) usually adds profile, identity and audience design in the CDP. Moving to Customer.io usually adds event tracking and product data, since it's built around customer behavior.
What gets forgotten most often?
Forms on secondary sites, consent status, attribution history and the teams who need training on the new system.
Planning a migration? We own migrations end to end across HubSpot, Salesforce, Adobe Experience Cloud and Customer.io, as one or several workstreams alongside your team or an existing partner. Request a call.