What data do you need before automating marketing workflows?

CRM fields: populated, consistent, and defined

The CRM fields an automation reads must be populated on most records, entered consistently, and defined in one place everyone accepts. A field that's blank on 40% of records, or filled by three teams using three conventions, will silently corrupt every workflow built on it. Audit the five fields your target workflow depends on before anything else.

Campaign and event tracking: complete enough to trust

Automation that reacts to campaign performance needs tracking that captures events reliably, attributes them consistently, and doesn't have gaps nobody talks about. If your team already argues about whether the dashboard numbers are right, an automation reading those numbers inherits the argument and acts on it.

Customer identity: one customer, one record

Any automation that personalizes, suppresses, or scores needs to know that the person in the CRM, the email platform, and the ad audience are the same person. That's an identity resolution problem, and it's the most common silent blocker in marketing automation. You don't need a full CDP to start, but you do need an honest answer about how often the same customer exists as three different records.

How to check without a six-figure data project

You can test all three foundations on a sample in days, not quarters. Pull 50 records and count populated fields. Trace one campaign's events end to end. Search one loyal customer's email across your systems and count the duplicates. The point of checking first is that the answer changes the plan: sometimes the right 90-day move is a data fix that makes every later automation cheaper.

This kind of check is part of the deep dive, and whether your data can support the top 3 moves is something the free first call surfaces first.