Stefan Knight/Tech Startup Multiples (MOICs)
Venture CapitalAI Created

Tech Startup Multiples (MOICs)

Stefan Knight · Antler · March 2026

First: the honest baseline

Before we dive into the glamour of 50x AI returns, here's the unglamorous truth about the median VC fund:

  • The median VC fund returns 1.72x TVPI — barely above returning capital
  • Only top-quartile funds break 3.2x
  • DPI (cash actually returned to LPs) remains stubbornly low: only 37% of 2019-vintage funds had distributed anything meaningful by Q1 2025
  • The exit drought of 2022–2024 means capital is locked up, not realised

The difference between a good fund and a median one isn't being right more often. It's backing the one company that returns the entire fund.

This is why the stage of entry matters so much. The return profile changes dramatically:

Stage Target MOIC Realistic MOIC Failure Rate
Seed 50–100x 5–15x ~39%
Series A 10–15x 3–8x ~20%
Series B 5–10x 2–5x ~12%
Series C/D 3–5x 1.5–3x <10%
Late / Pre-IPO 1.5–2.5x 1.2–2x <5%

The sector-by-sector breakdown

1. Enterprise SaaS — The Reliable Workhorse

SaaS remains the most data-rich sector and the most acquisitive. Acquisition multiples have settled into a new normal of 3–15x ARR for private companies — down from the 15–25x ARR frothy peak of 2021, but with meaningful premiums available for AI-integrated businesses.

  • Private SaaS median: 2.9–3.8x ARR at acquisition (2024–2025)
  • Top-tier deals (high NRR, AI-adjacent): 8–15x ARR
  • Notable benchmarks: Slack acquired at 26x ARR; GitHub at 23x ARR; Adobe's (failed) Figma bid at 50x ARR
  • Implied MOIC for early investors: 5–20x depending on entry stage and company quality

NRR above 120% is the single most powerful predictor of premium multiples in SaaS. It compounds the revenue base and signals product-market fit that buyers pay up for.

2. Artificial Intelligence — A Separate Universe

AI is not just a sector anymore — it's a valuation regime unto itself. The average revenue multiple across 90+ AI M&A deals sits at 25.8x. Foundational model companies operate in another stratosphere entirely: Anthropic at ~36x ARR, Databricks at 27.9x ARR on a $62B valuation (since risen to $134B).

  • Public AI companies: 25–35x revenue
  • Top private AI deals: 15–30x ARR
  • Foundational model outliers: 36–250x ARR
  • Wiz (Google acquisition, $32B): Estimated 30–50x MOIC for early investors
  • AI accounted for 53% of global VC funding in H1 2025 — concentration risk is extreme

The critical caveat: The top 10 AI companies are absorbing disproportionate capital and delivering exceptional returns. The vast majority of AI startups are burning through GPU infrastructure costs with no clear monetisation path. Being in AI is not enough — being in the right AI company is everything.

3. Cybersecurity — The Sleeper Strong Performer

Total disclosed cybersecurity M&A topped $84B in 2025. VC financing hit $20.7B across 820 deals — up 52% year-on-year. But the real story is the actual MOIC data from exits:

Company Acquirer Capital Raised MOIC
Talon Cyber Security Palo Alto Networks $143M 4.4x
Dig Security Palo Alto Networks Undisclosed 7.8x
Spera Security Okta Undisclosed 13.0x

Identity and cloud security are the strongest sub-sectors right now — Spera's 13x MOIC in identity security is a signal, not an outlier.

4. Fintech — A Tale of Two Worlds

Global fintech investment collapsed from $240B in 2021 to $96B in 2024 — a 60% contraction. Rate sensitivity, regulatory scrutiny, and the implosion of BNPL and crypto verticals wiped out significant value. But not all fintech is equal.

  • B2B banking infrastructure / API-first: 5–12x MOIC for early investors — high strategic value to acquirers
  • Consumer lending / BNPL: Largely impaired. Most 2021–2022 investors seeing <1–2x
  • Payments: Revenue growth slowed 12% → 4% in 2024, compressing multiples to 3–5x revenue
  • Wealthtech and Insurtech: Mixed — 4–7x for AI-augmented players, 2–3x for legacy

If you backed B2B fintech infrastructure before 2020, you're likely sitting on a strong outcome. If you backed consumer BNPL in 2021, you're learning an expensive lesson about rate sensitivity.

5. Healthtech — Consolidation, Not Hypergrowth

Digital health hit a decade-low of 118 M&A deals in 2024. The telehealth boom is well and truly over. What's emerging is a more durable, PE-driven consolidation phase:

  • AI-enabled clinical workflow tools: 6–8x revenue at acquisition
  • Revenue cycle management, EHR automation: Consistent buyer demand, 5–7x revenue
  • Consumer wellness apps: 3–4x revenue — churn-sensitive and post-COVID normalised
  • PE investment in healthtech rose 50% in 2024 as financial buyers replace VC-led exits

Early VC investors in pre-2020 clinical workflow automation are achieving 5–15x MOIC. Pandemic-era telehealth investors are mostly looking at 0.5–2x.

6. Consumer Tech & Marketplaces — Power Law Extremes

Consumer tech is where VC power-law dynamics are most violent. A handful of platforms generate 100–500x MOIC for early backers. The long tail returns near zero. Median marketplace multiple sits at a modest 2.2x revenue.

  • Premium marketplaces (Airbnb-class): Up to 9.5x revenue
  • Standard two-sided marketplaces: 1.9–2.2x revenue
  • Consumer subscription apps: 2–6x ARR (churn is everything)
  • Ad-supported consumer apps: 1–4x revenue — most compressed segment

Buyers now demand a clear path to profitability before acquiring. The era of "grow at all costs" consumer app funding is over.

7. Cleantech — Patient Capital Required

Climate tech raised a record $2.1T globally in 2024 (including infrastructure capex). But pure VC-style returns require patience:

  • Software-first grid / climate analytics: 4–8x revenue, 5–15x MOIC for early investors
  • Hardware / energy storage: 1–3x revenue, long holds of 8–12 years
  • 89% of 2025 cleantech exits were acquisitions (not IPOs)
  • IRA and EU Green Deal create a policy floor — but also policy risk

Business model beats sector

Here's the insight that gets lost in sector-focused analysis: business model is more predictive of MOIC than sector. An AI company with a services-heavy revenue model will underperform a SaaS company in a "less exciting" sector.

Business Model Acq. Multiple Early VC MOIC Range Key Risk
Recurring B2B SaaS 3–15x ARR 5–20x AI displacement
AI / Foundation Models 15–50x ARR 15–50x+ Concentration
Usage-based / API 5–12x ARR 8–20x Volume dependency
Marketplace (take-rate) 2–9x rev 3–50x Winner-takes-most
Transactional / Payments 3–5x rev 3–10x Margin compression
Consumer subscription 2–6x ARR 2–15x Churn sensitivity
Ad-supported consumer 1–4x rev 1–8x AI content disruption
Deep tech / hardware 1–5x rev 2–8x Long cycles
Tech-enabled services 0.5–2x rev 1–4x No SaaS premium

The compression story: 2021 → 2025

This is the uncomfortable chapter. The 2021 vintage produced the most impaired cohort of VC-backed companies in modern history.

Sector 2021 Peak Multiple 2025 Multiple Impact on 2021 Investors
SaaS (non-AI) 15–25x ARR 3–6x ARR 0.3–0.5x MOIC at current prices
Fintech (consumer) 8–12x rev 2–4x rev Significant impairment
AI-native SaaS 15–20x ARR 15–30x ARR Still 5–10x+ for strong cos
Cybersecurity (AI) 10–15x rev 8–15x rev Relatively resilient

A 2021 Series B investor in a non-AI SaaS company that raised at 20x ARR is looking at a 0.3–0.5x MOIC at current market prices. Without an AI re-rating story, those assets won't recover without a significant macro re-rating. This is the buried lede in most VC portfolio reviews right now.

Five things I take from this

1. B2B SaaS with genuine AI integration is the sweet spot. It combines recurring revenue predictability with AI valuation premiums. This is where the risk-adjusted MOIC is strongest for 2025–2027 exits.

2. Cybersecurity is underrated as a return generator. Spera at 13x MOIC, Dig at 7.8x — identity and cloud security is delivering AI-tier returns with B2B SaaS predictability. Less talked about than AI, better unit economics.

3. Not all AI is equal — and most of it will disappoint. AI accounts for 53% of global VC in H1 2025. The top 10 companies are generating extraordinary returns. Most of the rest are burning cash on GPU costs with commoditising models. Selecting in AI is harder than any other sector right now.

4. Stage of entry is non-negotiable for MOIC. Seed-stage capital is the only place where 10x+ portfolio MOIC is structurally achievable. Later-stage investors compress returns to 2–4x on a blended basis. The math only works at the beginning.

5. The 2021 vintage is a write-off for most non-AI assets. The companies that didn't develop an AI story are unlikely to achieve the multiples required to return 2021 entry prices. Founders and investors in those businesses need a strategy, not hope.

The bottom line

Private tech returns in 2026 look nothing like 2021. The market has bifurcated — not just between sectors, but within them. An AI-native B2B SaaS company and a non-AI B2B SaaS company in the same market will trade at completely different multiples, attract different buyers, and return fundamentally different outcomes to investors.

For founders: the AI question is no longer optional. For investors: entry stage and business model matter more than ever. For LPs: patience and selectivity are the only edges left.

The best investments of the next five years will be made by people who understand that 'AI' is not a sector. It's a capability — and the companies that build it into genuine distribution will be worth multiples of those that don't.


Stefan Knight · Antler · March 2026

Sources: Cambridge Associates, SaaS Capital, Finro, Finerva, Momentum Cyber, Rock Health, KPMG Fintech Pulse, Alvarez & Marsal, Aventis Advisors, Sightline Climate