Series A in 2025: The $3M ARR Bar & What It Takes

The Series A bar has quietly doubled. Carta's Peter Walker has the receipts: median ARR at Series A jumped from $1.3M in 2021 to nearly $3M in 2024, while the time to get there grew from 1.6 to 2.1 years. If you're still planning on the old playbook, you're already behind.

Key Facts

• Median ARR at Series A (2024): ~$3M, up from ~$1.3M in 2021 (Peter Walker, citing Silicon Valley Bank data)
• Median seed-to-A timeline: 2.1 years, up from 1.6 years (Peter Walker, Carta)
• Founders reaching Series A within 2 years: Only ~20% of 2023 seed cohort (Peter Walker, Carta)
• Founder ownership after first priced round: 56% (down from ~90% at founding) (Peter Walker, Carta)
• Median team size at Series A: 13 people in 2025, down from 20-22 in 2022 (Peter Walker, Carta)

The $3M ARR Bar: How Series A Requirements Have Shifted

The median ARR required to raise a Series A has risen from roughly $1.3M in 2021 to nearly $3M in 2024, based on Silicon Valley Bank data cited by Peter Walker. At the 75th percentile, that bar is now close to $7M — nearly double what it was three years ago.

If you built your fundraising plan around old benchmarks, you're operating with outdated intelligence. Peter Walker walked through SVB data showing the dramatic shift in what Series A investors expect. 'It used to be in 2021 that the median amount of ARR for a Series A company is just above $1 million bucks,' he explained. 'In 2024, that was nearly $3 million.'

That's not a modest adjustment — it's a near-tripling of the revenue bar in three years. And the math is uncomfortable. Most seed rounds are $3M or less. Getting from $300K–$500K ARR at seed close to $3M ARR on a single raise means achieving a near 1:1 revenue-to-capital ratio early — an extremely high bar. As Pablo Srugo put it: 'You've got to go from like $300K or $500K... to $3 million. You have to spend this $3 million on a one-to-one ratio if you're going to hit that. And that's just median.'

Walker's explanation: investor Overton windows have shifted. With companies like Lovable reportedly hitting $80M ARR in eight months, 'great' has been redefined. Investors aren't expecting every company to match that, but 'pushing great up means the Overton window of startups also moves up. You have to swim even faster just to maintain your current place.'

"If you're walking into a conversation with a Series A VC and you're walking in there with $1.5 million bucks of ARR, two things better be true. One, you better know that and adjust. And two, you better be growing really quickly." — Peter Walker

"You've got to go from like $300K or $500K to $3 million with $3 million bucks. That's a one-to-one ratio and that's just median." — Pablo Srugo

Seed-to-Series A Timelines: Plan for 3+ Years, Not 18 Months

The standard VC advice of raising every 18–24 months is now dangerously outdated. Carta data shows the median seed-to-A time is 2.1 years, up from 1.6. At the 75th percentile — meaning one in four founders — that wait is 3.5 years or longer.

The '18 to 24 months between rounds' rule was solid advice for years. It no longer is. According to Carta cap-table data, the median time between seed and Series A has grown from 1.6 years to 2.1 years. That's the median. For a large share of founders, it's far longer.

Srugo flagged the real planning implication: 'The latest data would show that if you raise a seed right now, you have a one in four chance of it taking three and a half years or longer to raise an A. Which, frankly, I can tell you no one is planning for. No one.' The average seed round simply isn't sized to support 3.5 years of operations.

Walker added that bridge rounds aren't a reliable escape hatch: 'I think bridge round behavior... gets people into trouble.' The Carta data on bridge round conversion rates reinforces this — companies that bridge to their A have significantly lower odds of getting there than those that go direct. The takeaway: raise enough at seed to survive longer than you expect, and don't count on a bridge to save you.

"You have a one in four chance of it taking three and a half years or longer to raise an A. Which, frankly, I can tell you no one is planning for." — Pablo Srugo

"The standard VC advice of you're going to fundraise every 18 to 24 months was pretty good advice for five or six years. Now that is maybe bad advice." — Peter Walker

• Median seed-to-A time: 2.1 years (up from 1.6 years)
• 75th percentile wait: 3.5+ years before reaching Series A
• Only ~20% of 2023 seed cohort reached Series A within two years
• Bridge rounds carry materially lower conversion odds than direct raises
• The old '18–24 month' planning window is now considered bad advice

AI vs. Non-AI: The Widest Funding Bifurcation in Venture History

Seed valuations for AI startups are running 40–50% higher than non-AI peers, while overall deal volume has roughly halved over three years. Fewer deals are getting done at higher prices — almost entirely concentrated in AI-native companies with the right credentials.

The numbers tell a stark story. Median seed valuations peaked at $14M pre-money in early 2022, declined, and are now climbing back past $15.6M — heading toward $16M in Q2 2025. But that average masks a massive split. 'There's an in crowd and an out crowd,' Walker explained. 'And the bifurcation between those two things has never probably been wider than it is today.'

AI startups — especially founders from AI labs or building in hot AI verticals — are commanding 40–50% valuation premiums at seed. Meanwhile, sectors like edtech, DTC retail, and food and beverage are raising smaller rounds at lower valuations with far less investor attention. Walker is direct about the consequences for non-AI founders: 'If you are not in that cohort, and that cohort is pretty small, it is going to take longer than you think to fundraise.'

Srugo's advice for founders on the right side of this divide: move now. 'You want to be a seller when everyone's buying... raise that money. You don't know how long that insane heat is going to last.' For those outside the AI spotlight, the path is longer but not impossible — it just requires adjusting expectations and proving out more revenue before re-engaging the market.

"Spun off from an AI lab? Congrats, here's a billion dollars. That is what's going on. There's a people in the sun that are AI native or somehow have the right credibility stamps, and then there's everybody else." — Peter Walker

"If you're not in that AI category, you really have to adjust to that reality. It takes way longer to raise an A, there's half as many deals getting done and most of those are not getting done in companies like you." — Pablo Srugo

Founder Dilution: What Your Cap Table Actually Looks Like

After a first priced seed round, founders own a median of 56% of their company — down from roughly 90% at founding. By Series D, the founding team's median ownership drops to approximately 10%. Every round costs more than most founders expect.

Dilution is a slow tax that compounds with every raise. Walker broke down exactly where the equity goes after a priced seed round: roughly 20% to the seed investors, 10–12% allocated to the employee option pool, and 11% to pre-seed angels and SAFE holders. 'So from 90% to 56% after their first priced round. That's a lot of dilution, people.'

Walker also flagged why dilution percentages barely move round-to-round: investors almost always take around 20%, not because that's what founders need to raise, but because of how funds are structured. 'The numbers that you're coming to in negotiation with your investors are driven at least as much, if not more, by the investor's ownership needs.' Understanding this gives founders negotiating leverage if they have demand.

By Series D, the founding team owns roughly 10%. Srugo's framing cuts through the noise: 'Raising money is not value creation. You raise money, you don't create value.' The implication — paper unicorn valuations built on successive large rounds can leave founders with less real wealth than the headline number suggests.

"The founding team, after their first priced round, own on median 56% of their own business. So from 90% to 56% after their first price round. That's a lot of dilution, people." — Peter Walker

"Raising money is not value creation. You raise money, you don't create value." — Pablo Srugo

Leaner Teams, Longer Runways: The New Startup Operating Model

Median team size at Series A has dropped from roughly 20–22 people in 2022 to just 13 in 2025. Founders are waiting longer before making first hires, scrutinizing every headcount decision, and doing more with less — a structural shift accelerated by AI tools and board pressure.

The headlines about lean startups are real — and the data backs them up. Walker shared that companies raising Series A in 2025 have a median of 13 employees, down from around 20–22 in 2022. New hires across Carta dropped from roughly 70,000 in January 2022 to around 26,000 in January 2025. 'Now every single hire is being scrutinized,' Walker noted. 'When somebody proposes a new hire, the first thing that is asked is, well, can we do that with AI?'

Walker pointed to Elon Musk's Twitter acquisition as a cultural inflection point: 'I kind of think people are underrating the example of Elon in this case... When did this all change? It was Elon buying Twitter and firing like 80% of the people, and then it still worked. To me, that was a wake-up call moment for a lot of people across tech.' Boards began asking hard questions, and founders who had answers gained credibility.

Srugo sees smaller teams as structurally advantageous for early-stage success: 'One of your edges at early stage is the lack of overhead. That is one of the edges and I think many founders give it up first.' Walker added the mathematical reality — the complexity gap between 15 and 20 people isn't five, it scales factorially across all additional relationships. Lean isn't just efficient; it may be a genuine competitive moat.

"Teams are smaller. Everyone's talking about it. The headlines are true. If you are building in AI startups, the expectation is that you're going to try to do more with less." — Peter Walker

"One of your edges at early stage is the lack of overhead. Don't give that one thing up." — Pablo Srugo

Series A Benchmarks: 2021 vs. 2024–2025