Cube vs Mosaic for Omnichannel Retail: Store, Wholesale, and DTC P&L

A retail footprint with a dozen or more locations has real performance variation store to store, driven by rent, local labor cost, and foot traffic that a regional or company-wide average completely obscures. Two stores with identical revenue can have very different profitability once their actual rent and staffing costs are applied individually.

Cube and Jirav both let you build store-level P&L as its own layer in the model, with each location's rent, labor, and local marketing cost tracked against its own revenue rather than allocated from a company average. Mosaic's dashboards are built around company-wide SaaS metrics and don't have a native concept of a physical location's individual P&L.

When a popular style sells faster than expected, the company has to decide whether the remaining units go to stores, get held back for wholesale commitments already promised to retail partners, or get sold through the higher-margin DTC channel, and that allocation decision has real margin consequences depending on which choice gets made.

Build an explicit inventory allocation rule into the model, prioritizing by margin or by contractual wholesale commitment, rather than letting allocation happen ad hoc at the warehouse level with no visibility into the margin tradeoff being made in real time.

Wholesale revenue trades a lower per-unit price for guaranteed volume and no inventory or markdown risk once the product ships to the retail partner, while DTC and owned retail carry the full margin but also the full inventory risk. Blending all three into one gross margin number hides how a shift toward wholesale volume, chasing guaranteed orders during a soft DTC quarter, say, changes overall profitability even when unit economics haven't moved in any one channel.

Track gross margin by channel as a standing report, not a special analysis run once a year, so a channel-mix shift shows up in the regular numbers leadership already looks at.

The seasonal buy commitment has to account for store inventory, wholesale account commitments, and DTC demand simultaneously, and the lead time on that buy, often placed months ahead of the season, means the channel split has to be forecast well before real sales data from any channel comes in.

Build the buy plan with an explicit channel allocation assumption stated up front, and true it up against actual channel performance as the season progresses. Cube's and Jirav's flexibility to model this multi-channel buy split is a better fit than Mosaic's more standardized approach.

Cube's spreadsheet-native model can go as deep on channel-level detail as you're willing to build, store-by-store, account-by-account for wholesale, campaign-by-campaign for DTC, limited mainly by your own time to maintain it. Jirav offers similar flexibility through driver-based templates that get you to a working multi-channel model faster, with somewhat less granularity than a fully custom Cube build.

Mosaic's strength is automated, low-maintenance reporting, which works well if the brand is willing to accept a company-wide or two-channel view rather than the full store-by-store and account-by-account breakdown the other tools can provide.

• If store-level and wholesale-account-level detail matters to how you actually run the business, Cube's depth is worth the setup and maintenance time.
• If you want multi-channel modeling without a long build, Jirav's driver-based templates get you a working version faster, with less granularity than a full custom build.
• If a simpler, lower-maintenance company-wide view is acceptable, Mosaic's automated reporting reduces ongoing work, with the tradeoff of less channel-level depth.

Whichever tool you choose, resist the urge to report one blended retail number to leadership. The channel mix shift is usually the real story, and a blended number is exactly what hides it.