Data Cloud Readiness Checklist
Seven areas to check before you configure your first data stream, so your Data Cloud project starts with clear use cases, known identifiers and owned data.
Data Cloud (now branded Data 360) projects rarely stall on configuration. They stall because nobody agreed which use case comes first, the customer identifiers in each source do not line up, or the team discovers halfway through that consent data lives in a spreadsheet. This checklist helps you find those gaps before the build starts, when they are cheap to fix.
Who should complete it
Work through it with three people in the room: someone who owns the business outcome (often marketing or service), someone who knows the source systems, and someone accountable for privacy and data governance. Answer each question with evidence, not intent. "We think the e-commerce platform has a customer ID" is a no until someone has looked at an export.
The seven readiness areas
- 01
Use cases are specific and ranked
Write the first two or three use cases as outcomes with an audience, a channel and a trigger, such as 'send a replenishment reminder to customers whose last purchase of product X was 60 days ago'. Vague goals like 'a single view of the customer' cannot be scoped or tested.
- 02
Sources are inventoried
List every system that holds data the first use cases need: CRM, commerce, web or app events, service, loyalty. For each, record owner, volume, refresh frequency, and whether a connector exists or a file or API feed is needed.
- 03
Identifiers are understood
Identity resolution depends on shared keys such as email, phone, customer number or device ID. Note which identifiers each source holds, their format, and how often they are missing or shared between people.
- 04
Data quality has been sampled
Pull a sample from each source and check for duplicates, invalid emails, inconsistent phone formats, placeholder values and orphaned records. Poor quality here becomes over-merged or under-merged profiles later.
- 05
Consent and purpose are mapped
Know which consent and preference data exists, where it is the system of record, and which purposes it covers. Activation should respect it from day one.
- 06
Access, licensing and environments are clear
Confirm what your contract includes, who administers the org, whether you need separate data spaces, and how consumption is metered. Check current details with your Salesforce account team and Salesforce Help.
- 07
The team and run model are named
Someone must own mappings, monitor job failures and approve new segments after go-live. Name those people before the build, not after.
Source inventory example
| Source | Key identifiers | Refresh need | Readiness notes |
|---|---|---|---|
| Sales Cloud | Contact ID, email | Near real time | Duplicate contacts known; dedupe rules needed |
| Commerce platform | Customer number, email | Hourly | Guest checkouts have email only |
| Web events | Cookie or device ID, email when logged in | Streaming | Tagging plan not yet agreed |
| Loyalty database | Member ID, phone | Daily file | Phone stored in mixed formats |
| Marketing Cloud Engagement | Subscriber Key | Daily | Subscriber Key equals CRM Contact ID |
Scoring your readiness
For each of the seven areas, mark it ready, partly ready or not ready. A practical rule of thumb:
- If use cases or identifiers are not ready, pause and fix those first. Everything else depends on them.
- If data quality or consent is only partly ready, you can usually start design, but plan clean-up work into the first phase.
- If access or the run model is not ready, the build can begin, but go-live should be gated on both.
Questions to answer about identity
Identity is where most readiness reviews uncover surprises. Before you design match rules, answer these:
- Which identifier is the most trusted in each source, and is it ever reused?
- How many records share an email address, and why (households, shared work inboxes, test data)?
- Are phone numbers stored with country codes?
- Do you need to keep business contacts and consumers apart?
- Which source should win when two sources disagree about a name or address?
The answers feed straight into match and reconciliation rules. Our articles on data quality checks before identity resolution and mapping custom sources to the Customer 360 Data Model go into the detail.
Questions we hear
Do we need perfect data before starting?
Can we start with just CRM data?
Who should own Data Cloud after go-live?
How long does a readiness review take?
Next steps
If your review shows mostly green, move on to source design using our guide to choosing data stream ingestion patterns. If it shows significant gaps, a short discovery phase with our Data Cloud consulting team can help you sequence the fixes. For a broader view of your Salesforce estate, see the Salesforce readiness assessment.
Not sure your data is ready?
We can run a structured readiness review with your source owners and give you a sequenced plan for the first Data Cloud use cases.
