Two brands, one Data Cloud org: data spaces vs. a unified customer view
Northwind Outdoors runs two retail brands, Trailhead (camping) and Velocita (cycling), from one Salesforce CRM and one Data Cloud org. Loyalty members can shop both brands; about 38% of the 1.8M loyalty profiles have purchases in both. Trailhead's marketing team is contractually barred from seeing Velocita order data, and Velocita's team has the same restriction in reverse. The data team proposes one data space per brand, each ingesting the shared CRM connector plus its own ecommerce stream. The CMO also wants a quarterly cross-brand customer value report that counts a loyalty member once across both brands. The first attempt created two data spaces, but the cross-brand report now shows 1.1M members in Trailhead and 900K in Velocita with no overlap, and the CMO says the numbers are obviously wrong.
1. The CMO wants one cross-brand member count, while each brand's legal team forbids the other from seeing its order detail. What does a data-space-per-brand design actually isolate, and what will it not solve? Recommend an architecture and justify the trade-off.
Read the worked answer
A data space is the isolation boundary in Data Cloud. Data streams, DLOs/DMOs, identity resolution rulesets, unified profiles, segments and activations are scoped to a space. The same person appearing in both spaces receives two different Unified Individual IDs. There is no cross-space identity resolution and no cross-space segment join.
So the proposal satisfies the legal separation requirement but cannot produce a person-level cross-brand count from unified profiles. A third analytics data space fed by a curated, minimal extract is the pragmatic answer: loyalty_id, brand, value band, and no order detail. An ELT job, approved by legal, writes only those fields. That space can count distinct loyalty_id across brands without exposing Velocita orders to Trailhead users.
If person-level cross-brand activation were required, the only clean option would be one data space with brand-scoped processes and strict access controls. That trades legal isolation for identity continuity. Because the stated need is reporting, not cross-brand activation, choose two brand spaces plus a small, deliberately limited analytics space.
2. If the company instead used a single data space, what controls would you put in place to reduce the risk of brand data leakage, and what residual risk remains?
Read the worked answer
In a single space, identity resolution can unify a member across brands, which is attractive. The controls would be: separate brand-specific segments and activations; separate data actions and target credentials; field minimization so each brand only ingests what it needs; masking or excluding sensitive order fields at ingestion; tightly scoped permission sets; and a documented process that brand users do not build ad hoc queries against the shared DMOs.
The residual risk is important. Data Cloud permission sets govern feature and object access, not row-level brand filtering within a DMO. A user with broad query access could see rows belonging to the other brand. That is a governance and contractual risk, not just a UI issue. A single space is therefore appropriate only if legal accepts that the brands share a trust boundary. If the contract requires hard separation, data spaces are the safer control even though they fragment identity.
3. What explains the 1.1M plus 900K split, and what would you tell the CMO?
Read the worked answer
The two counts add to 2.0M, which is larger than the 1.8M loyalty population because shared members are counted once in each space. Each space has its own identity resolution and its own Unified Individual records, so the overlap is invisible. The numbers are not necessarily wrong for each brand; they are wrong to add together.
Explain it to the CMO as a scope issue: Trailhead has 1.1M unified profiles in its space, Velocita has 900K in its space, and the overlap is roughly the 38% cross-shop population. The deduplicated enterprise count must come from the analytics space that receives the minimal loyalty extract. Set the expectation that the brand spaces are operational systems for marketing execution, while the analytics space is the system of record for cross-brand counts.