Founders rarely lose a raise on "bad metrics." They lose it by leading with the wrong metric for their stage — which invites the wrong question — or by presenting a number in a form that reads as manufactured. This guide covers both: which metric to lead with, and the conservative bar each clears.
For every metric there's a decision at each stage: lead with it (put it in the headline), support with it (show it, but it's not the star), or omit it (showing it at this stage only invites a question you can't yet answer). Here's the map for a B2B-SaaS raise:
| Metric | Pre-seed | Seed | Series A | Series B |
|---|---|---|---|---|
| ARR | context | LEAD | LEAD | LEAD |
| MoM MRR growth | support | LEAD | support | omit |
| YoY growth (×) | omit | support | LEAD | LEAD |
| Gross margin | omit | support | LEAD | LEAD |
| NDR | omit | omit | LEAD | LEAD |
| Burn multiple | omit | context | LEAD | LEAD |
| Logo retention | omit | support | support | support |
| CAC payback | omit | context | support | support |
| Design partners / LOIs | LEAD | support | omit | omit |
The single biggest shift is early to late: at pre-seed your design partners / LOIs are the traction — ARR is context only — while by Series A the story is carried by ARR, YoY growth, NDR and the burn multiple. A seed deck leading with NDR looks like it's answering a question no one asked yet; a Series A deck that buries NDR looks like it's hiding something.
These are deliberately conservative "bar to clear" bands, not top-decile vanity numbers, drawn from public benchmark reporting (Bessemer State of the Cloud, ICONIQ Growth, OpenView/High Alpha SaaS Benchmarks, KeyBanc, Carta). Verify against the source before you quote one to an investor — markets move.
| Metric | Stage | Clears | Strong |
|---|---|---|---|
| ARR | Seed | $100K | $1.0M |
| ARR | Series A | $1.5M | $3.0M |
| ARR | Series B | $5.0M | $10M |
| MoM MRR growth | Seed | 10% | 20% |
| YoY growth | Series A | 2× | 3× |
| Gross margin (SaaS) | Series A | 65% | 75% |
| NDR | Series A | 100% | 110% |
| NDR | Series B | 110% | 120% |
| Burn multiple (lower is better) | Series A | ≤2× | ≤1× |
| CAC payback (lower is better) | Series A | ≤18mo | ≤12mo |
| Logo retention | Series A | 80% | 90% |
The number is half the battle; the form is the other half. A few of the presentation rules the engine applies:
State it as committed/contracted ARR (not run-rated one-offs), with current MoM growth beside it. Never annualize a single good month.
Show as consecutive MoM % over the last 6+ months, not a single spike. At seed, sustained ~15%+ MoM is the headline number if ARR is still small.
State as a clean multiple (e.g. "3× YoY") from a stated base. A multiple off a tiny base must show the absolute base too, or it reads as manufactured.
Show the % with the COGS line beneath it. If below software-grade (AI inference, payments, hardware), pair it with the explicit path up as you scale.
Present as a trailing-12-month cohort number, defined (expansion + churn, existing logos only). Above 100% is the whole point — it means you grow with zero new logos.
Net burn ÷ net new ARR, read next to your growth rate, never alone. In the efficiency era this is the number the IC memo cites verbatim.
Fully-loaded CAC recovered from gross-margin dollars, in months. State the loading — an unloaded number reads as a dodge.
Never lead with headline GMV. Pair it with take-rate so net revenue is visible; GMV alone invites the "thin marketplace" question.
The report grades each metric you provide 🟢/🟡/🔴 against your stage and model, tells you which to lead with, and gives the exact form to present it in.
Request the report — $490 Try the free calculator