Seed Round Milestone Thresholds That Unlock Series A Conversations
Founders need $2M–$5M ARR with strong growth and retention to start Series A conversations in 2025.
The median time from Seed close to Series A close now runs around 774 days, close to 26 months. That's stretched out from roughly 24 months not long ago, and from about 18 months back in 2020. Founders still building a fundraising calendar around an 18-month cycle are planning against a market that closed years ago, and no amount of hustle in month 14 fixes a timeline problem that started on day one.
Seed rounds have gotten bigger to compensate, which sounds like relief but actually raises the bar. Median Seed pre-money valuation reached $16 million in 2025, up 18% from the year before, with a median cash raise of $4 million. That extra capital buys more runway, but the Series A step-up has compressed to match it: the median step-up from Seed to Series A sat at 2.8x in Q2 2024, down sharply from 4.9x in Q2 2022. A smaller step-up means the A has to be earned with real metrics, not priced on narrative. Momentum pricing, where investors paid up on trajectory alone, isn't underwriting deals the way it did a few years ago.
The correct move is to work backward from 774 days, not forward from the Seed close, and most founders still do the opposite. Figure out which milestones need to be locked by month 18, so credible investor conversations can start by month 24. That's not a case for rushing the raise. It's a case for building the milestone calendar on day one of the Seed, not eighteen months in, once the runway clock is already loud in the room.
The ARR thresholds Series A investors actually use in 2025
The old benchmarks that defined a Series A conversation in earlier years don't hold anymore. Competitive B2B SaaS raises in 2025 cluster around $2 million to $5 million in ARR. In monthly terms, that's somewhere between $167,000 and $417,000 in MRR, which matters more to founders tracking month to month than to anyone waiting on the annual number to catch up.
Founders benchmarking against 2021, when some Series A rounds got funded at far lower ARR thresholds paired with steep growth, are benchmarking against a dead comp. That scenario is gone for non-AI companies, full stop, and pretending otherwise just wastes a fundraising cycle. AI-native startups are the real exception: the category commands a documented valuation premium, and that shows up directly in pricing, with Series A rounds for AI companies closing at a median 38% above non-AI deals in 2025. Consumer and marketplace businesses run on a different set of metrics entirely, covered further down.
Treating the ARR number as the finish line is the miscalculation almost everyone at this stage makes. It's the entry fee, not the outcome. Investors read the ARR figure alongside growth rate, retention, and unit economics, and a company that clears the ARR threshold but fails on any of those adjacent metrics gets a hard pass, not a maybe.
Growth rate as a multiplier on the ARR number
B2B SaaS investors want strong month-over-month growth, which annualizes to roughly 2x to 3x, or very high year-over-year growth, alongside that $2 million to $3 million ARR floor. The old T2D3 framework (triple, triple, double, double, double) still functions as the reference trajectory investors mentally check a company against, even when nobody says the acronym out loud in the meeting.
Position on the ARR ladder matters less than the slope underneath it, and founders who chase the number instead of the slope are optimizing for the wrong variable. A company sitting at a low ARR base but growing strongly year over year is telling a stronger story than one at $2 million ARR growing 30% a year: the first is on a curve that reaches the threshold fast, the second has stalled at a level that used to be a green light. Growth rate decides how fast a company closes that gap. A business with strong monthly growth from a $500,000 base can close the gap to threshold ARR relatively quickly, while a slower growth rate can stretch that timeline considerably. That math should shape runway decisions directly, not sit in a spreadsheet nobody reopens after the board meeting.
Deceleration worries investors more than a smaller headline number does. Investors are pattern-matching on trajectory, not position on a chart, so a flattening growth curve can concern them as much as a lower headline ARR number with momentum still intact. Track the rolling three-month growth rate with the same discipline applied to the ARR total, and have a real answer ready for any inflection, up or down.
Why net revenue retention has become a near-binary filter
Net revenue retention at or above breakeven is table stakes for even having the Series A conversation. Below 100%, meaning the existing customer base is shrinking in revenue terms, the conversation tends to end regardless of what the top-line ARR says. Competitive NRR sits meaningfully above breakeven, and anything at 120% or higher signals real expansion revenue: customers deepening their use of the product over time rather than merely sticking around.
The number shows up directly in pricing, too. Meaningful improvement in NRR feeds directly into the valuation model investors are running. NRR isn't a health check tucked into the diligence memo, it's an input investors plug straight into the valuation model. Logo retention, the simpler measure of how many customers stick around regardless of revenue, should sit at a strong level by Series A or B, a sharp jump from the lower logo retention typical at Seed.
Retention outranks growth rate in investor thinking for a specific reason: a company posting strong top-line growth while churning a large share of its revenue base is running a leaky bucket. The growth is real, but the floor underneath it is cracked, and investors have learned to look past the top-line number to the retention curve holding it up. Cohort curves that flatten and stabilize over time signal a durable base. A curve still falling steeply well into the cohort's life signals a product that hasn't found real fit yet.
If NRR sits below 100% heading into a raise, the move is to delay, diagnose the churn, and fix it, not paper over it with a bigger ARR number. Investors have seen that particular sleight of hand enough times to spot it on sight.
Unit economics thresholds that tell investors the model can scale
The working numbers: an LTV to CAC ratio above 3x, and a CAC payback period under 18 months. Past 18 months, payback period functions as close to a hard stop in the current market, because it tells investors that scaling acquisition will burn cash faster than the business can replace it.
That tolerance used to be looser. In the zero-rate years, investors funded long payback periods because capital was cheap enough to subsidize the wait. The repriced cost of capital took that patience away, and a payback period that would have cleared in 2021 now gets flagged on sight. Investors also want early evidence of a repeatable go-to-market motion beyond founder-led deals: proof that a sales hire could run the playbook without the founder in the room. That's a Series A requirement in its own right now, not a bonus layered on top of a working product.
Gross margin above 70% is the floor SaaS investors use to judge whether unit economics compound favorably at scale. Margins under that require a real explanation, not a wave of the hand about future efficiencies that haven't shown up yet. Founders should also be able to break CAC down by channel, because a blended number hides whether one channel is efficient and another is quietly destroying value. Not having that breakdown ready reads as a lack of operational rigor, no matter how clean the blended figure looks on the summary slide.
What Series A investors look for when ARR doesn't apply
Consumer products get judged on a different scale entirely. A meaningfully high DAU to MAU ratio functions as the engagement threshold that substitutes for revenue retention, and cohort curves that flatten rather than trend toward zero play the same role NRR plays for SaaS. Evidence of organic growth, word of mouth, referral loops, a viral coefficient approaching or crossing 1, tells investors the product has real pull instead of a paid-acquisition habit propping up the growth chart.
Marketplace businesses run on GMV scale and take-rate economics instead of ARR. The real question is whether the marketplace is getting more liquid over time and whether transaction economics improve as it scales, not whether raw volume is going up. Early-stage AI-native companies carry a documented premium too: Seed-stage AI companies closed at a 42% valuation premium over non-AI peers in 2025. But investors are getting sharper about telling apart AI products with genuine retention from AI-wrapper products that draw high initial curiosity and then bleed users just as fast, and that discernment is only going to get sharper from here.
Across every one of these models, the underlying question doesn't change even when the metric does: has the product crossed from interesting to relied upon? A consumer company with strong engagement but no articulated path to monetization makes a tougher Series A pitch than one that has at least tested a preliminary monetization layer, even a small, early one.
Team composition as a threshold, not a soft factor
Founders who say they need Series A capital to go hire a technical co-founder rarely get past a first conversation, according to CRV. For a technical product, the technical co-founder belongs on the team before the raise starts, not as a line item inside the funding plan. Pitching it as a future hire is a tell that reads badly to anyone who's sat on the other side of the table, and it should be treated as disqualifying, not merely suboptimal.
At Series A, investors are underwriting whether the founding CEO can run a company of 50 to 100-plus employees, not just whether they can ship a product. The question shifts from "can you build this?" to "can you hire, delegate, and operate at a level of complexity you haven't faced yet?" Nobody expects a complete C-suite this early. What investors want is the start of functional leadership: early signs the founder has begun building a team around themselves instead of holding every function personally, past the point where that's sustainable.
That expectation carries real weight because Series A is typically the first round where the lead investor takes a board seat, which changes governance dynamics for years afterward. Investors are underwriting the team they'll govern alongside, not just the metrics on the pitch deck. A founder who can credibly describe a path to a billion-dollar revenue opportunity, and show the operating capacity to chase it, is a different bet than one with identical ARR and a team that's visibly hit its ceiling. Map planned hires against the 774-day window, and make the critical executive hires early enough that those people have a track record to point to, rather than announcing them mid-raise as a promise nobody can yet verify.
How the bridge round phenomenon signals that most founders are missing the thresholds
Bridge rounds have grown into a large share of Seed-stage financing, mostly because companies burn through Seed capital without reaching Series A-ready metrics. Of the companies that raised a meaningful Seed round in 2023, only 24% had progressed to a further round by the time of a May 2026 report. For the 2024 cohort, the figure drops to 16%.
Read plainly, those numbers mean most Seed companies are either bridging, quietly winding down, or settling into self-sustaining businesses with no real growth trajectory: exactly the outcome Series A-track founders are trying to avoid. A bridge round isn't automatically a failure. It's the right tool when a company is genuinely three to six months from a real milestone. Far more often, though, it's a sign the original Seed-to-A plan underestimated how long the actual path would take, and founders are reluctant to admit that on the fundraising deck.
Step-up compression raises the stakes further. At a 2.8x median step-up from Seed to A, a bridge that buys extra months without moving the key metrics doesn't improve the valuation story at all. It just delays the reckoning. Investors draw a clean line here: a bridge that buys time to hit a specific, already-visible threshold is fundable, but a bridge that buys time and hopes the metrics improve on their own is a warning sign, and founders need to be honest with themselves about which one they're actually asking for.
Building a milestone roadmap that works backward from Series A readiness
Most founders build forward from the Seed close and hope the metrics arrive in time. That's the single biggest planning error at this stage, and it's backwards in a literal sense. Start instead from the Series A threshold the company is actually targeting, whether that's $3 million ARR with 110%-plus NRR for a SaaS business, or the engagement and retention equivalent for a consumer product, and work backward to what has to be true by month 12 and month 18 to make that number credible by month 24.
Set growth-rate checkpoints on a rolling three-month basis, not on an annual number that's already six months stale by the time it lands in a board deck. Treat NRR as a metric to defend from the first paying cohort onward, not something to patch up the quarter before a raise. Hire the technical co-founder or the first sales lead early enough that they've built a track record investors can actually evaluate, instead of bringing them on mid-raise as a promise about future capability that nobody can price.
The 774-day median and the 2.8x step-up aren't obstacles worth arguing with. They're the operating conditions of the current market, and founders who plan against them, deliberately and early, are the ones who convert Seed capital into a real Series A conversation instead of a bridge round that just delays an answer everyone already knows.
