Series A was won on cohort math and payback windows. Series B is won on whether that math scales.
If you raised your Series A on the four metric categories from the prior post in this series — growth & efficiency, retention & cohort behavior, unit economics at scale, and efficiency of growth — you walked out of the round with a model that worked at $3M ARR. Series B is the same game with bigger stakes: at $5–15M+ ARR, the partner across the table is no longer testing that you have unit economics. They're testing that the numbers still hold at $40M ARR. The shape of the answers changes. The underlying tests don't.
This is the second post in the metrics-stage playbook: the seven metrics Series B investors actually test, how each one evolves from the Series A version of the same question, and the prep you can do this week to walk into the meeting with answers the room respects.
It's tempting — and a trap — to walk in with seven clean numbers and treat them as seven isolated screens. Series B investors don't ask them that way. They ask them as categories, because what they're testing is whether the business still works as it scales. The seven metrics below are the ones that come up across every Series B diligence call. Prep each as a small story, with what the investor is testing, the trap to avoid, and a line you can actually deliver.
What headcount of annual recurring revenue you sit in, and how you got there credibly.
What they're testing: that the business has reached a band where the next round's math is even legible. At $5–15M+ ARR, the Series B investor is modeling forward to $40M and beyond, and they need you to be at a scale where that model isn't a fairy tale. The number itself matters less than how you got there — that the trajectory is repeatable, that the cohort math underneath it isn't a fluke, and that the ARR you're quoting isn't gross-of-churn dressed up as growth.
The trap founders fall into: quoting gross ARR without netting churn or contraction. A company at $12M ARR with 9% gross churn and a single upgrade cohort driving most of the new bookings looks very different from one at $12M ARR with 4% gross churn and broad-based expansion. Both can quote the same headline. Only one survives diligence.
The prep line you can deliver: "We're at $11.4M ARR on a net-of-churn-and-contraction basis. Gross ARR is $13.1M. The delta is normal — about 11% gross churn in the SMB segment and 4% in mid-market — and our net-new ARR over the trailing six quarters has been driven by expansion in the existing book, not by a single cohort window. Top-decile accounts run 38% of revenue, down from 51% twelve months ago." That's the language of a founder who has the math under the math.
Year-over-year ARR multiplier, with the cohort shape underneath.
What they're testing: that the growth curve is still steep, but that you've crossed out of the hyper-growth regime where revenue multiples compress and the bar shifts to efficiency. At Series B, 2–3x YoY is the band — fast enough to justify the round's price, slow enough that you're not burning harder than the math supports. Series A "3x+" was normal. Series B "3x+" reads as undisciplined unless you can defend the spend behind it.
The trap founders fall into: relying on T2D3 language (triple, triple, double, double, double — the YC-coined growth curve for SaaS) when you're already above $5M ARR. T2D3 stops being the right frame after the second year. At Series B, the right frame is "are we still compounding cleanly, and is our payback window still making the new growth good business." Quote the multiplier, the absolute ARR behind it, and the spend that produced it.
The prep line you can deliver: "YoY growth is 2.4x on $11.4M ARR, and net-new ARR has been compounding at about 18% Q-o-Q over the trailing four. The growth is in mid-market expansion, not top of funnel paid — net-new S&M spend was up 11% last quarter, and net-new ARR was up 22%, so efficiency is improving as we scale. We don't expect to stay above 2x forever; the right shape now is 2x compounding on a bigger base." That's the conversation Series B partners respect.
Net dollar retention, with the composition underneath.
What they're testing: whether your existing book of business is paying you more in year two than year one — and whether that pattern holds as the book gets bigger, more diverse, and more brutal in cohort shape. At Series A, 120+ NDR was the screen. At Series B, the band widens to 110–130% because investors know that any book at scale contains long-tail cohorts, and a 130 book looks different from a 130 book that hides a 95 cohort under it. They want the curve.
The trap founders fall into: NDR that looks product-led but is actually renewal-heavy or driven by a single upgrade cohort. Investors who write checks at this stage have seen too many "118% NDR" stories where the path to 118 was a price increase or a single upgrade window. Don't give them room to assume that. Lead with the composition — expansion vs price vs usage vs downgrade — and the cohort trajectory over the last four to six quarters.
The prep line you can deliver: "Net dollar retention is 122 across the trailing four quarters. Of that, 96 is expansion, 14 is price, and 12 is usage-driven. The 2024 cohort is at 119 in its second year; the 2025 cohort is at 131 in its first year. Gross retention is 91. Logo retention is 86, with concentration in the bottom decile of accounts that churned in months 8–14 — that's the cohort we rebuilt onboarding around in Q1." That decomposition, delivered cleanly, is the cue that you understand retention the way the investor does.
Lifetime value to customer acquisition cost, in the post-enterprise-buildup shape.
What they're testing: whether your payback model still works after you've spent the last two years investing in enterprise motion, customer success, and partnerships — all of which compress early LTV:CAC ratios by raising CAC and lengthening time-to-payback. The Series A ratio of 3:1+ came from a leaner model. The Series B version of the same number has to defend itself against the heavier cost structure you've built on purpose. Investors know you built it on purpose. They want the math to still work.
The trap founders fall into: quoting a blended view that hides poor payback in sales-led. A blended 3.4:1 can hide a 1.8:1 in sales-led and a 5.0:1 in self-serve. The blended number is fine if you can defend both halves of it; it's a trap if you can't. Have the channel-level cut ready, with year-one and year-two cohort splits, and a forward-looking view of where the blended number lands as sales-led payback improves.
The prep line you can deliver: "Blended LTV:CAC is 3.6:1 on a 36-month modeled LTV. Sales-led sits at 2.4:1 today, up from 1.8:1 a year ago, and we expect it to clear 3:1 within two quarters as the CS investments from last year normalize. Self-serve runs at 5.2:1 with payback under six months. The 2024 sales-led cohort is at 2.9:1; the 2025 cohort is at 2.6:1 with a similar trajectory." That's the answer to a Series B version of a Series A question.
Months for a sales-and-marketing dollar to come back as gross profit from the cohort it brought in.
What they're testing: how long it takes you to earn back the cost of acquiring a customer, and whether that duration is short enough that the round's capital can be deployed twice before raise dilution. At Series A, 14–18 months was normal. At Series B, the bar tightens — not because your cost structure improves, but because CAC rises as the category matures and competition for the same accounts intensifies. Under 18 months is still good; under 14 months is excellent; over 24 months is a problem.
The trap founders fall into: a CAC that only looks good because you've under-invested in CS for two years. Payback improves when you cut customer success to the bone — the math gets pretty until the cohorts that weren't supported churn in their second year, and your NDR falls out from underneath you. Investors can spot this if they're looking at payback and NDR together. Make sure your payback investment in CS is visible in the model, not absent from it.
The prep line you can deliver: "Sales-led blended payback is 17 months today, down from 22 a year ago. That improvement is driven by tighter ICP and a 30% reduction in bottom-quartile pipeline, not by cutting CS — CS spend is up 18% over the same period and is fully reflected in the cost stack. Mid-market payback is 13, enterprise is 21, and our self-serve motion is at 5. Gross-margin impact of CS investment is about 4 points and is included in the payback model." That's the version of payback the Series B room is listening for.
Net new ARR divided by the prior quarter's sales-and-marketing spend.
What they're testing: whether each incremental dollar of sales-and-marketing spend is producing over 70¢ of net new ARR. Magic number is the cleanest single-screen for whether your go-to-market motion is still scaling profitably, or whether you've hit the point where adding spend produces diminishing returns. A Series A magic number of 1.0+ was excellent. At Series B, the band shifts — 0.7 to 1.0 is good, >1.0 is hard to sustain, <0.5 means you've stopped earning on incremental spend.
The trap founders fall into: quoting it on a book of business that includes a long tail of low-quality cohorts. Magic number collapses if you grow into segments that don't pay back. A 0.9 magic number driven by a flat-out charge into a low-LTV segment in Q4 looks nothing like a 0.9 driven by deep expansion of an existing ICP. Have both the headline and the segment breakdown ready — specifically whether the last two quarters of growth came from expansion of segments that already worked, or from new segments that haven't been tested yet.
The prep line you can deliver: "Magic number is 0.84 on net new ARR / prior-quarter S&M spend. That number is segment-balanced — 0.91 in mid-market and 0.74 in enterprise — and both segments are inside the ICP we validated at Series A. We're not chasing volume in segments we haven't tested. The number has held in the 0.75–0.95 band across the last six quarters, which we read as a healthy steady-state for our motion." That's the language of a founder who hasn't let the magic number drift into low-quality growth.
Net burn divided by net new ARR, with the trend, not the snapshot.
What they're testing: how much capital you're consuming to produce each dollar of new ARR, and whether the trajectory is one an investor would call capital-efficient at Series B scale. Burn Multiple (from the David Sacks framework) compresses the conversation about growth and efficiency into a single ratio. At Series A, sub-2.0x was the goal. At Series B, sub-1.5x is good; trending down quarter over quarter is what makes it fundable.
The trap founders fall into: improving burn multiple by cutting growth spend rather than improving efficiency. Burn Multiple goes down when you cut S&M as fast as you cut net burn. That's not an efficiency story — it's a growth-restraint story, and investors who write checks at this stage can spot the difference. The healthy burn multiple comes from net-new ARR rising faster than net burn, not the opposite. Name the trend explicitly and what produced it.
The prep line you can deliver: "Burn multiple on the trailing four quarters is 1.3x, down from 1.9x a year ago. The improvement comes from net-new ARR up 41% over that period against net burn down only 9%. We didn't cut growth — we cut the bottom 15% of the sales motion that wasn't paying back, which is reflected in the magic number, not in S&M headcount. The trajectory we model forward keeps burn multiple in the 1.0–1.3x band through the Series B deployment." That's a Series B burn multiple answer that reads as healthy, not as managed.
If you raised Series A on the four categories from the prior post in this series — growth & efficiency, retention & cohort behavior, unit economics at scale, and burn multiple — the Series B version of each sits on the same spine. The shape of the answers changes. The underlying tests don't.
What changes is the bar, the scale at which the answer has to hold, and the cost structure the answer has to live inside. What doesn't change is the investor's question — they want to see that you understand the relationship between several numbers and what they imply about the business as it compounds. That's the same question at $3M ARR as it is at $15M ARR. The map below shows how each Series A category evolves at Series B, and what to bring into the room:
That's the bridge. The six Series B metrics up top aren't a different test — they're a harder version of the test you already passed. The prep is the same prep, with bigger numbers and a more demanding audience.
One week doesn't change any of these numbers. It changes how you present them, and presentation is most of the Series B meeting.
Series B founders who raise aren't the ones with the prettiest numbers — they're the ones who've done the work to understand what the numbers imply. That's the difference between a deck an investor files away and a deck they pass up to partnership. Prep this week and you walk in with answers the room respects, at a scale where the math still has to hold.
The Founder Academy Scaling Seed to Series B course walks through each of these metrics with founders who are in the room. Bring your numbers, leave with a model that still works at $40M ARR.
Explore Scaling Seed to Series B →Series B is won on whether the Series A math still holds at scale — at $5–15M+ ARR, with a 2–3x YoY multiplier, NRR in the 110–130% band, an LTV:CAC above 3, payback under 18 months, magic number above 0.7, and a burn multiple trending down underneath 1.5x. Each of those is the Series B version of a Series A question. Prep this week by decomposing ARR into gross versus net, cutting LTV:CAC and payback by channel and cohort year, projecting unit economics at 3x and 5x today's revenue, and naming what produced the trend on your burn multiple. The founders who raise at this stage are the ones who've done the work to understand what their numbers imply at the next scale.
Want to go deeper?
The playbook for scaling your startup from initial traction to institutional venture rounds. For founders who've raised seed and need to level up.
Explore Scaling from Seed to Series B — $249 →