Owned, partner, and enriched data working together
By Abhishek Shukla, Vice President, Product & Platform Management, Epsilon India
If you’ve been following the MarTech evolution, this story would sound familiar:
A global CPG company spends years rebuilding its data strategy. It launches a loyalty program, invests in capturing zero-party data, builds out a CDP, and implements measurement across every owned digital touchpoint. By the end, it has more first-party data than at any point in its history. However, when the CMO asks the question: of the consumers we want to reach this quarter, how many can we actually identify? The answer is roughly 8%.
This doesn’t reflect a failure in MarTech implementation but exposes a structural reality. The CPG company sells most of its products through retailers and e-commerce marketplaces. The customers who buy its products interact with the retailer’s checkout, not the brand’s. No amount of first-party data investment can change that reality.
This is the limitation the industry has been slowest to acknowledge. Since the early 2020s, the rallying cry across marketing has been to “build your first-party data”. However, for many categories, including CPG, hospitality, and financial services, first-party data alone reaches only the customers you’ve already engaged. The much larger group that buys through partners, or shops in-market without authenticating themselves, remains invisible to your systems.
Brands that succeed in this evolution are not the ones with the most first-party data, but the ones treating first-party data as a foundation, not the destination.
The structural limits
First-party data has constraints worth naming.
By definition, it only sees the customers you’ve already won. It tells you a great deal about your existing customers and almost nothing about prospects or lapsed customers. For acquisition and prospecting, first-party data is structurally narrow.
It’s a behavioral record, not a complete person. A loyalty program tells you what someone has bought from you, but not what else they buy, where they travel, or what other brands they consider. Even excellent first-party data captures a sliver of the consumer’s life.
Moreover, for many industries, the customer relationship is mediated. CPG brands sell through retailers. Hospitality brands sell through booking platforms. In such cases, the partner owns the transactional moment that matters most, not the brand.
The three layers that actually work
The strategies that produce measurable results combine three data layers, with identity as the connective tissue between them.
First-party data (the foundation): Owned data remains the anchor. This is the most reliable signal you have, captured with consent and grounded in actual relationships. Every other data type compounds value when joined to a strong first-party base.
Second-party data (partner data, shared safely): Second-party data is another organization’s first-party data, shared through a structured partnership. According to IDC’s FutureScape 2026, by 2028, 60% of enterprises will collaborate on data through private exchanges or data clean rooms. This is the infrastructure that makes second-party collaboration safe, governed, and operational.
Enriched data (privacy-compliant context): This includes the deterministic, consented, identity-anchored datasets that enrich what brands already know. Household composition, lifestyle indicators, in-market signals sourced through compliant providers, and joined to first-party data through verified identity.
When used together, the three layers compound. A CPG brand running a retail media campaign on a partner’s network can layer its first-party CRM (loyalty members) with second-party retailer data (recent purchase behavior) and enriched data (household composition, in-market signals), all within a clean room, never exposing raw PII. In a Snowflake case study, Booking.com and Snap, collaborating through Snowflake’s clean room infrastructure, increased their confidence in campaign measurement from under 20% to roughly 99%. That kind of jump is hard to imagine with any single data source.
What makes it work: identity and clean rooms
Combining first-party, second-party, and enriched data only works if you can confirm you’re talking about the same person across sources. Without a deterministic identity that’s anchored in stable signals like verified name and address, validated through transactions, partner data doesn’t reconcile to your customer. Identity is what turns three data layers into one coherent customer view rather than three disconnected datasets.
A clean room makes this reconciliation possible. A clean room is the legal and technical envelope that enables two parties to combine data without exposing raw PII to either side. This is what makes second-party collaboration something a privacy officer can sign off on.
The questions worth asking
For a marketer thinking about expanding beyond first-party data, here are some useful evaluation questions:
What share of your target audience are you currently invisible to in your own data, and which partners hold the visibility you’re missing? Where would a second-party data collaboration unlock prospecting reach that owned data can’t? And most importantly, does your identity layer let you confidently join data across these sources?
The bottom line
First-party data is not the destination. It is the foundation. A complete strategy combines owned, partner, and enriched data, coordinated through identity, governed through clean rooms, and measured through outcomes that none of the three could produce alone. The brands still treating first-party as sufficient aren’t wrong about its importance, but they’re underestimating how much further the strategy can go.
