Customer data platform
Why it matters
Without unified data, teams act blind. Marketing emails a lead who spoke to sales yesterday. A customer who just raised a support ticket gets a pitch for the feature that failed them. With a shared profile, messaging responds to behaviour: who read the security docs, who visited pricing twice, who has gone quiet. It also saves busywork, because nobody has to check three systems by hand before a call.
How to apply it
- Start with the question, such as which customers are at risk, then connect only the sources that answer it.
- Decide which events matter, such as page visits, form fills, demo requests and usage drops, instead of capturing everything.
- Set data rules early: who can see what, which fields are personal data and how long records are kept.
- Check the matching. One person showing up as three profiles is worse than no profile.
- Send the merged profile back out to the tools that act on it. A profile nobody uses is only a storage cost.
What it is
Most companies hold customer facts in several places. The website has page visits, the CRM has deals, the support tool has tickets, the product has usage, the email tool has opens. None of them sees the whole person. A CDP connects to all of them, matches records that belong together and builds one profile per person or account. It then keeps that profile current and shares it with the other tools.
Matching is the hard part. The same person may appear as an email address in one tool, a user ID in another and a cookie in a third. A CDP uses those shared identifiers to join them, which is often called identity resolution.
Common mistakes
- Buying a platform before knowing which decision the unified profile should improve.
- Connecting every source at once, so the data is wide but nobody trusts any of it.
- Skipping a check on how records are matched, which creates duplicate or wrongly merged profiles.
- Collecting personal data without a clear purpose, a lawful basis or a retention period.
- Treating it as a marketing tool only. Support, product and success data are often the most useful inputs.
- Leaving the profile in the platform without sending it back to the tools where people work.
