Serious A (Part II): A Tale of Two Series A’s

By, Zaz Floreani
Carta, PitchBook, KPMG, and Cooley data all point to the same story: fewer Series A rounds, slightly higher prices, and a widening gap between AI-driven growth and everything else. 2025 may be shaping out to be a year of a “two-track” Series A market.

“Welp,” I said to myself after hanging up from a regular update call with a fellow investor at a massive coastal multistage fund. I couldn’t stop scratching my head after our conversation.

When I asked how things were going at their fund, this investor casually shared three Series A deals they’d either lost or bowed out of due to high valuations. Two of the three were pre-revenue. All were “AI deals.” The one deal that did have real sales had a growth curve that would make most investors blink.
Meanwhile, we have three startups in market raising Series A rounds—each with stellar metrics, revenue run rates well above the old $1M ARR benchmark, and real AI integrated into their products—and their processes are moving at a standard clip.

It feels like we’re living in two parallel Series A universes: the AI acceleration track, where velocity and vision drive rapid raises, and the “A is the new B” track, where discipline and proof needed are higher than ever

A Tale of Two Series A’s
For most startups, Series A has always been the milestone of proof—you had repeatable sales, a growing customer base, and a team ready to scale. But in 2025, the rise of AI-native companies has created a second track that moves at lightning speed.

Some of these companies are achieving growth and adoption rates we simply haven’t seen before. Others are still raising on pedigree and potential. The truth, as usual, lies somewhere in between: many AI deals are genuinely earning their valuations, while others are benefiting from the halo of momentum.

Just as in my July post, the latest data from Carta and PitchBook shows that overall Series A activity remains about the same, with AI/ML accounting for 60% of 2025 venture deal value. Meanwhile, non-AI SaaS and marketplace founders face the same climb—one that still rewards progress and discipline, but requires more patience to convert it into capital.

So yes, there are two markets—but both are functioning. Capital hasn’t disappeared; it’s just selective.

Devil Is in the Data
While Carta hasn’t yet released its Q3 data, which typically does an excellent job of isolating Series A data (vs other stages), the story through Q3 from other sources is remarkably consistent with where we left off this summer. There’s been no massive post-summer bounce.

According to Crunchbase, in North America, the total dollars invested in Series A and B deals together rose slightly while the number of deals declined slightly.

Carta’s Q2 numbers show median valuations up -  median Series A valuation reached $47.9 million—a new high on their platform. But that number could be misleading as it’s likely being driven by AI deals. KPMG’s Q3 report showed a similar median pre-money valuation of $46.8 million, while Cooley’s Q3 data (with a smaller dataset) actually saw the median valuation in Q3 tick slightly down from Q2—though interestingly, Seed and Series B valuations both rose.

In short: fewer Series A rounds are getting done, but at higher prices—a dynamic that looks more like a tale of scarcity than strength.

The Series A Time Continuum
The time between Seed and Series A continues to stretch. Carta’s Q2 report shows the median interval hit 616 days (~20 months), up about two months from two years ago.
Fewer founders are making it to the A, and those who do are older, more mature companies than in prior cycles.

​Cooley’s data adds another signal of investor caution: Pay-to-play provisions remain elevated at 10.1% of deals. So founders should be aware of the typical terms that the potential investors they are talking to offer.

The Split Reality

The bifurcation in the market is mainly along the lines of AI or non-AI, but that also correlates with location.  Or as Peter Walker of Carta put it:  it is true the venture in SF functions differently AND that media attention for companies in SF is much higher…The 90th percentile for the Bay Area fundraises is more than double the "other US cities' figure. More money = more attention = distortion of what the "market" truly is right now.

Maybe obvious, but worth saying regardless - you can be supremely ambitious and not live in San Francisco. You can build an incredible startup and not raise a 90th percentile round. You can create massive impact and not be in the golden AI circle.

Overall the Series A landscape hasn’t worsened—it’s evolved. Non-coastal founders who can show fast, efficient growth (many of them AI-enabled) are finding capital more accessible than headlines suggest it just might be taking longer than usual.

What Can You Control
For founders outside the AI super-funding world, the playbook is less about survival and more about precision.

• Execute, execute, execute:  hit the milestones you need to in order to ensure you can clear the new, higher bar for non-AI companies at Series A.
• Manage burn deliberately: Treat cash discipline as a strategy, not a sacrifice. It buys you options.
• Build early relationships: Don’t wait until you’re “ready to raise” to meet investors. Start now. Build your shortlist, share updates, and let people track your progress over time.

Fundraising is still a relationship business. The best rounds—AI or not—are going to the founders who’ve earned investors’ conviction long before they get a term sheet.