Funding & M&A Digest — September 13, 2026: The Rollup Thesis Gets Its Second Proof Point

Five weeks. That’s the gap between Bending Spoons closing Airtable at $1.285B enterprise value and signing a definitive agreement for Miro at $1.355B EV. Two category-defining B2B brands, both absorbed by the same Milan-based rollup, inside roughly 35 days.

If you were waiting for confirmation that the “SaaSpocalypse” thesis was more than a bearish Twitter thread, this is it. One data point is noise. Two, sequenced this tightly, at almost identical price bands, is a clearing price.

What Sub-$2B Actually Means

Both Airtable and Miro were once mentioned in the same breath as Figma and Notion. Both carried multi-billion private marks at peak. Both had enormous brand equity, genuine product love, and tens of millions of users.

Neither had durable free cash flow.

That’s the variable that matters now. The market has re-priced high-brand, low-FCF SaaS at sub-$2B EV regardless of logo recognition. Bending Spoons’ playbook is unglamorous and effective: acquire a beloved product, strip discretionary spend, retain the cash-generating core, and run it for margin rather than growth. It works precisely because the public and late-stage private markets have stopped paying for growth that doesn’t convert to cash.

The uncomfortable read-through for founders: if you’re a Series D company growing at the 2026 median of 26% with negative free cash flow, you are not a takeover target — you are inventory.

The Benchmarks That Make This Inevitable

The M&A behaviour isn’t irrational. It’s a rational response to a benchmark set that has quietly deteriorated for four straight years:

• Median growth rate: 26%, down from 30% in 2022.
• Net revenue retention: 101% — top decile at 113%, bottom decile at 98%. Median NRR barely clearing 100% means the median SaaS company no longer grows from its installed base at all.
• CAC up 14% year over year, with average CAC payback at 14 months versus 5–7 months for elite operators.
• LTV:CAC “healthy” bar has moved from 3:1 to 4:1. What used to be respectable is now the minimum viable threshold.
• B2B monthly churn averaging 3.5% — implying roughly a 28-month customer lifetime. Top performers hold under 2%, buying themselves a ~50-month lifetime.

Stack those together and you get a company that grows slower, retains less, pays more to acquire, and takes longer to recoup. That is not a $10B narrative. That is a $1.3B asset with a good logo.

Where the Money Is Actually Going

Meanwhile, the primary market is alive — just extremely selective. Today’s flow:

• Lio — $30M Series A, a16z-led, AI procurement
• Catena Labs — $30M Series A
• Fresha — $80M from KKR at a $1B valuation
• CasaPerks — $15.8M seed
• GEEIQ — $6.8M
• Circolife — $4.5M pre-Series A (Cooling-as-a-Service, India)

The pattern is unmistakable: agentic workflow automation inside unsexy back-office functions. Procurement. Payments infrastructure. Scheduling. Not another chat interface bolted onto a foundation model.

SVB’s data showed 65% of US enterprise software VC went to AI startups in 2025. But the composition has shifted hard. VCs are now publicly listing the AI SaaS characteristics they will not fund — thin wrappers, no proprietary workflow data, no switching costs. The wrapper era closed sometime in the last two quarters, and the market didn’t send a memo.

The Margin Problem Nobody Wants to Underwrite

Here’s the number that should anchor every AI-native board deck this quarter: AI-native subscription gross margin runs 55–70%, against 75% median SaaS and 80–85% for mature enterprise software.

A 20-point margin gap means an AI-native company needs roughly 1.3x the revenue to generate the same contribution profit as a traditional SaaS peer. That’s not a rounding error — it’s a structurally different business requiring structurally more capital to reach the same absolute gross profit.

The only path out is inference deflation, and that’s arriving fast. Microsoft’s MAI-Transcribe-2 just undercut OpenAI, Google and ElevenLabs on both price and latency. Cognition shipped SWE-2 at 64% below Fable 5.1. Model-layer deflation is now roughly quarterly.

So your FY27 AI COGS assumption is wrong — and wrong in your favour. But note the symmetry: the same deflation that fixes your margin also collapses your pricing power. Cheap inference is not a moat. It’s a commodity input your competitors get too.

🇮🇳 India: Liquidity Without an IPO

The most instructive Indian SaaS story today isn’t a funding round. It’s Whatfix’s $58M employee and investor buyback at roughly a $900M valuation.

This matters structurally. Indian SaaS has spent a decade telling ESOP holders that liquidity arrives at IPO. Whatfix just demonstrated a third path: generate enough cash to buy your own paper back and reset the retention clock without a listing. Expect secondaries and structured buybacks to become the default ESOP retention tool across Indian SaaS in FY27 — particularly for companies in the $50M–150M ARR band that are too profitable to need capital and too small to justify a public listing.

The counterpoint: the public window is genuinely open. RentoMojo’s IPO was subscribed 72.87x, with the QIB portion at 177.3x. Domestic institutional appetite for Indian subscription businesses is at a structural high