Cube vs Mosaic for B2B Marketplaces: GMV, Take Rate, and Float

A marketplace growing GMV thirty percent quarter over quarter while its take rate compresses from four percent to three percent can show flat or even declining real revenue despite a headline number that looks like strong growth. Any report that leads with GMV without take rate right beside it is telling an incomplete story to whoever's reading it, investors, lenders, or the leadership team itself.

Build every revenue report with GMV, take rate, and resulting commission revenue as three connected lines, not one blended growth number. Cube and Jirav both make this straightforward since take rate is just a calculated ratio you can track by segment; Mosaic's metrics are built around a fixed subscription price, not a variable commission rate.

A marketplace often has to lower its take rate to win larger transactions or attract high-volume repeat buyers and sellers, since a smaller percentage of a much bigger deal can still mean more absolute commission dollars. If you only track a single blended take rate, you can't tell whether compression is a deliberate, healthy strategy for landing bigger accounts or an unplanned erosion across the whole platform.

Track take rate by transaction size band and by category, so a deliberate discount on your largest accounts doesn't get confused with a broader pricing weakness. This level of segmentation is where Cube's or Jirav's spreadsheet flexibility earns its keep over Mosaic's more standardized reporting.

A marketplace that collects payment from a buyer and holds it briefly before remitting to the seller, even for a few days, generates float, cash the company holds temporarily that isn't actually its revenue. Counting that float as available cash without tracking the corresponding payable to sellers overstates the company's real liquidity, and a finance team that spends against that inflated number can find itself short when a large remittance batch goes out.

Model float and the seller payable it corresponds to as a matched pair, not as free cash the company can spend. Cube and Jirav both let you build this as an explicit balance-sheet-style line in the forecast rather than letting it disappear into a general cash number, which matters more as transaction volume grows and the float balance becomes large enough to tempt a finance team into treating it as working capital.

Trust and safety, dispute resolution, and account management headcount typically scale with transaction count and buyer-seller relationship complexity, not with GMV, since a smaller number of very large transactions can require just as much support staff as a larger number of small ones. Planning headcount against GMV growth alone risks understaffing support functions during a period when deal sizes are growing but transaction count isn't.

Build a headcount trigger tied to transaction volume and active buyer-seller pair count, separate from the GMV-driven revenue forecast. Both Cube and Jirav handle this as a straightforward, separate driver in the model.

• If take-rate segmentation and float tracking are core to how you manage the business, Cube's or Jirav's flexibility to model both explicitly will serve you better than Mosaic's subscription-oriented defaults.
• If the marketplace also runs a smaller subscription or listing-fee revenue stream alongside commission revenue, Mosaic's recurring-revenue tools can add value for that specific line.
• If you want a working GMV-to-revenue model fast without a long spreadsheet build, Jirav's driver-based templates get you there quicker.

Whichever tool you choose, never let a board deck or investor update lead with GMV alone. Take rate is the number that tells you whether growth is actually translating into real revenue.