B2B SaaS Finance: Cube vs Mosaic for ARR Waterfalls and Headcount

Mosaic pulls contract and invoice data from your billing system and CRM, then classifies every change as new, expansion, contraction, or churn using its own rule set. You configure exceptions (a downgrade that's really a plan migration, say) but the classification logic lives inside Mosaic, not in a formula you can open and edit line by line.

Cube takes the same raw data and lands it in a structured layer, but the waterfall itself is still a spreadsheet model your team builds and owns. That matters most when your revenue has edge cases a generic SaaS template doesn't anticipate: multi-year deals with built-in step-ups, seat-based pricing with a usage floor, or an early cohort still on grandfathered terms.

At the median private B2B SaaS company, expansion revenue makes up 40% of new ARR added in a given year¹, and that share tends to climb as a company scales past $50 million in ARR. If expansion is already your main growth lever, you'll spend real time in whichever tool you pick tuning how upgrades get classified, so test both on your messiest actual contracts before you decide.

A headcount plan that only tracks salary and start date misses the part that actually drives revenue: how long a new account executive takes to ramp to full quota capacity. Only 51% of SaaS account executives hit quota in a given year², which is exactly the kind of assumption a plan gets wrong when it assumes every hire ramps on schedule.

Mosaic's headcount planner links each sales hire's start date to a ramp curve and rolls that straight into the revenue forecast, so a board asking what happens if we delay two AE hires a quarter gets an answer in minutes. Cube handles the same question inside your workbook: you build the ramp curve yourself, which takes longer to set up but lets you model a ramp that doesn't fit a standard curve, like a hire who inherits an existing book of renewals.

Either way, the plan should separate quota-carrying hires from support roles, since lumping them together is how a board deck ends up overstating how fast new revenue actually shows up.

Burn multiple, net cash burned divided by net new ARR, is the number a board or investor will ask about before anything else in a down round or a bridge conversation. Investors generally treat a burn multiple under one as the mark of a capital-efficient company³, though the acceptable range loosens for earlier-stage companies still proving the model works.

Mosaic recalculates burn multiple automatically every time actuals load, and shows it next to Rule of 40 and runway in the same view, so a spike shows up before the month closes. Cube leaves you to build the calculation, which is more work up front but lets you decide exactly what counts as burn: does a one-time legal bill from a contract dispute belong in the number, or should it be excluded as non-recurring. Neither answer is wrong, but pick one and apply it consistently, because a burn multiple that changes definition quarter to quarter is worse than no burn multiple at all.

If your contracts are close to a standard subscription, one product, monthly or annual billing, a handful of tiers, Mosaic's out-of-the-box classification will probably match how you'd model it by hand anyway, and you save the setup time. The calculation gets harder to trust once you add usage-based overages, professional services bundled into the contract, or reseller margins that aren't simple percentage splits.

That's the point where Cube's spreadsheet foundation earns its complexity. You can build a formula that handles a customer paying a base fee plus metered usage that resets each renewal, something a configured metrics engine may force into an approximation. The tradeoff is that your model is only as good as the person who built it, and if that person leaves, the next analyst has to learn a bespoke workbook instead of a documented platform.

A reasonable test: pull your five most contractually unusual customers and try to model their ARR contribution by hand first. If that's straightforward, a templated tool will likely keep up. If it takes real judgment calls, you'll want the flexibility to encode those judgment calls yourself.

Run these checks before you commit to either tool:

• Check whether your contracts are close to a standard subscription: one product, monthly or annual billing, and a handful of tiers that a template can classify.
• Look for usage-based overages, bundled professional services, or reseller margins in your contracts, since these strain a generic classification of new, expansion, and churn.
• Ask how each tool treats a mid-month upgrade or a downgrade that is really a plan migration, and whether you can edit that rule yourself.
• Confirm your team can open and audit the formula behind the ARR waterfall line by line, rather than trusting a rule set hidden inside the tool.

Neither tool fixes a retention problem. If gross churn is climbing or net revenue retention has been sliding for three quarters, a better dashboard just gives you a clearer view of a shrinking number, it doesn't make the number bigger. Get the retention and pricing conversation right first, then choose the modeling tool that fits how your revenue actually behaves.

If you want a second opinion on the assumptions before you commit to either platform, MeetMyCFO's AI CFO, Frank, can walk through your current ARR waterfall and flag where the classification logic is likely to disagree with a configured tool like Mosaic.