Do SaaS-era metrics and benchmarks apply to AI application companies?

Victor Lazarte · Apr 14, 2025

Rule-based revenue investing (e.g. $10M ARR means you're a category winner) worked in the SaaS era but no longer holds in AI

Much AI revenue is experimental, and it is now trivial to wrap a thin workflow around ChatGPT and sell it into a vertical — that generates millions in revenue but no enterprise value, because the workflow becomes less valuable as models improve

Scope: revenue growth still counts for something, just not what it used to

26:44 20VC: Benchmark's Victor Lazarte on Why Portfolio Construction is BS| Why SaaS Spreadsheet Investing is Dead | Why China is a Stabilising Force for the US | Three Traits All the Best Founders Have & The Lie All Big Tech Companies Have Been Telling

Everett Randle · Nov 10, 2025

The venture industry needs a new taxonomy and metric set for AI app companies, because they differ from SaaS companies in a dozen ways yet are still being forced into SaaS frameworks

AI app companies are meaningfully different from SaaS companies, so applying SaaS-era metrics to them misjudges them

22:27 20VC: Benchmark's Newest General Partner Ev Randle on Why Margins Matter Less in AI | Why Mega Funds Will Not Produce Good Returns | OpenAI vs Anthropic: What Happens and Who Wins Coding | Investing Lessons from Peter Thiel and Mamoon Hamid

Maor Shlomo · Nov 24, 2025

Getting users to a deployed, functioning application is now a saturated benchmark; the meaningful question is how many apps reach actual day-to-day usage

Agents have been battle-tested enough on querying databases and categorized user flows that most users reach working apps

Scope: says he doesn't have the answer to the usage number

33:47 20VC: Base44's Maor Shlomo on How Vibe Coding Will Kill SaaS and Salesforce | Why it is BS that Vibe Coding Platforms Do Not Have Defensibility and Bad Margins | Why He Worries About Google, Not Replit and Lovable | Why Long Anthropic, Not OpenAI?

Maor Shlomo · Nov 24, 2025

The relevant quality metric has shifted from bugs to intent-following, because LLMs are now smart enough and the infrastructure battle-tested enough that broken apps are rare

When they started the issue was models breaking apps; now the question is whether the agent makes the right change without messing up other features

34:37 20VC: Base44's Maor Shlomo on How Vibe Coding Will Kill SaaS and Salesforce | Why it is BS that Vibe Coding Platforms Do Not Have Defensibility and Bad Margins | Why He Worries About Google, Not Replit and Lovable | Why Long Anthropic, Not OpenAI?

Harry Stebbings · Mar 19, 2025 · hedged

The industry may have misled a generation of pre-AI enterprise companies about revenue scaling: the triple-triple-double-double benchmark will no longer be enough to raise later rounds now that AI companies scale far faster

AI-era companies growing from 2 to 100 million in a year reset the bar that later-stage investors use

Scope: framed as a worry rather than a settled view

18:48 20VC: The 10 Question Framework a $217BN Manager Uses to Make Investment Decisions | Lessons from Turning Down Stripe, Coinbase and Losing Money on Northvault | The Bull Case for Bytedance | How Anduril Could Be a $200BN Company with Peter Singlehurst

Harry Stebbings · Mar 21, 2025 · hedged

A generation of traditional SaaS companies will fail to raise funding because they can't match the new revenue expectations set by companies going to $50-100M in a year or two.

Expectations moved from triple-triple-double-double to the growth curves of outliers like Mercor and Midjourney, so ordinary enterprise companies now look inadequate.

Scope: framed as a worry

43:22 20VC: Selling Drift for $1.2BN is the Biggest Failure: What No One Tells You About Selling Your Company | Why Incumbents Are Slower & Worse Than Ever | Why the Most Valuable Companies in a World of AI Will Not Have More Than 100 People with Elias Torres

Harry Stebbings · May 5, 2025 · hedged

A generation of Series A and B companies doubling or 2.5x-ing will look uninteresting because investors will stack-rank them against AI companies with unparalleled revenue scaling

Companies like Lovable and Bolt are scaling revenue at rates unlike anything before, resetting the comparison set

Scope: framed as a worry he's asking Bucky to confirm or refute

28:40 20VC Exclusive: Why Mega Platforms Will Win in VC | Why You Cannot Do VC If You Do Not Do Pre-Seed | Why Market Sizing is BS | Where Will Foundation Models Build/Buy Apps vs Where Will They Not with Bucky Moore

Bucky Moore · May 5, 2025

The bar for what 'great' growth looks like has genuinely risen, and investors are right to reprioritize toward companies showing the new momentum profile

Prioritizing and ruthlessly reprioritizing toward the best ideas in front of you is core to the VC job, and pent-up demand for intelligence across areas of work is so high that growth rates and market pull are like nothing we've seen

Scope: the durability of that revenue is still unproven

29:21 20VC Exclusive: Why Mega Platforms Will Win in VC | Why You Cannot Do VC If You Do Not Do Pre-Seed | Why Market Sizing is BS | Where Will Foundation Models Build/Buy Apps vs Where Will They Not with Bucky Moore

Jake Saper · Mar 10, 2025

Retention cohorts for the current wave of fast-growing AI companies will disappoint on average, with a minority of outliers doing better than expected

Scope: these companies are still too young to have the data

44:17 20VC: Lessons from Investing $2BN and Returning $8BN in Cash | Why Most Venture Partnerships are Broken | We Sold Salesforce Early and Lost Out on Billions | Are The Best Deals Always Expensive and Competitive with Jake Saper @ Emergence Capital

Harry Stebbings · Mar 28, 2025

Nearly all AI software companies have poor gross dollar retention, though there are exceptions like Lovable at 85%, which is good for the category

Lovable's 85% is better than ChatGPT's

Scope: not every AI company

35:40 20VC: Why Traditional VC is Broken: How VCs Learned Nothing from 2021 | Why LPs are More Important than Founders & Advice to Emerging Managers | Bull Case for Bytedance & Why TikTok's Ban Doesn't Matter with Mitchell Green, Lead Edge Capital

Harry Stebbings · May 23, 2025

By traditional SaaS standards, the retention numbers celebrated in AI companies are bad — 86% retention means churning 14% a month

Traditional enterprise retention benchmarks were very much higher

56:18 20VC: ElevenLabs Head of Growth on Why You Do Not Need PMs | The 7-Part Launch Playbook That Gets 700K+ Views Per Product | The Truth About CAC, Payback & Performance Marketing in AI with Luke Harries

Immad Akhund · May 12, 2025

$5B of revenue is more attainable than it used to be because companies are now scaling to $100M in revenue in two to three years.

22:48 20VC Exclusive: Mercury Founder Launches First $26M Fund | Why Founders Should Take the Highest Price | Why Serial Entrepreneurs are Better | Why AI Is So Overhyped | The Future of Venture Capital with Immad Akhund

Byron Deeter · Aug 25, 2025

Growth benchmarks have been reset by AI — the best companies now go zero to $100M in about 1.5 years, versus the seven-year journey to $100M that defined the cloud era — though triple triple double double is still a very good business

Bessemer's state of the AI report quantified this, identifying 'supernova' profiles inside AI galaxies, against the old state of the cloud chart of seven-year centaurs

Scope: the traditional arc is still a pretty good business if you ride and scale it

25:21 20VC: Do Margins Matter in AI? | Is Defensibility Gone For Good? | Is Vertical SaaS Dead in a World of AI | What SaaS Rules Are BS and No Longer Apply in a World of AI | The Future of Venture: Why Chanel vs Walmart is BS with Byron Deeter

David Cahn · Oct 27, 2025

The right frame is the 'zero to $100M club' — the best AI companies get to $100M revenue very quickly, and fast growth is the best available indicator of smashing product market fit

Unlike the early internet, everyone is now online and everyone wants to buy AI, so genuinely good products get adopted extremely fast

Scope: a company doesn't have to be at $100M in revenue; he'd happily invest at $2M ARR if PMF is smashing; companies don't always have to grow this fast

41:15 20VC: Sequoia's David Cahn on The Winners and Losers in AI | The $0-$100M Revenue Club: Is Triple, Triple, Double, Double Dead? | The Future of Defence: Who Wins and Who Loses | How to Analyse Margins and Growth Rates in a World of AI

Philipp Freise · Jun 30, 2025

The handful of AI companies scaling from zero to $100M in a year are exceptional winners and will not be the norm for what comes afterwards

Scope: explicitly framed as a prediction

19:05 20VC: Inside KKR's Monster $8BN European Fund | The $500M Turkey Gamble That Went Wrong | Do Andreessen & General Catalyst Scare KKR? | Will AI Kill the PE Model? | Can The PE Model Survive without IPOs and Where is the Liquidity with Philip Freise

Byron Deeter · Aug 25, 2025

Consumer-like growth curves are now happening in enterprise businesses, with a rare subset going zero to $100M ARR in about 1.5 years ('supernovas') and a larger group doing it over roughly four years ('shooting stars')

Adoption curves can pull through great products at previously inconceivable rates and speeds

Scope: a rare class of company — only a small subset even of companies Bessemer backs; four-year 'shooting star' profile is the fatter part of the curve

26:24 20VC: Do Margins Matter in AI? | Is Defensibility Gone For Good? | Is Vertical SaaS Dead in a World of AI | What SaaS Rules Are BS and No Longer Apply in a World of AI | The Future of Venture: Why Chanel vs Walmart is BS with Byron Deeter

Harry Stebbings · Sep 20, 2024

AI consumer subscription products show unusually sharp revenue scalability, ramping much faster than businesses that need AEs, SDRs, BDRs and customer success teams

21:32 20Growth: The 7 Core Levers to Win at Consumer Subscription: Growth Loops, CAC + LTV Benchmarks, Pricing, Packaging, Notifications, Discounts, Paywalls | The Breakdown with Phil Carter

Harry Stebbings · Mar 19, 2025

Revenue scalability at the AI application layer is unlike anything previously seen, with companies adding $3-5M a week, and prices for them are exorbitant

Scope: accepts the three-layer model with commoditisation concentrated in the LLM middle layer

17:45 20VC: The 10 Question Framework a $217BN Manager Uses to Make Investment Decisions | Lessons from Turning Down Stripe, Coinbase and Losing Money on Northvault | The Bull Case for Bytedance | How Anduril Could Be a $200BN Company with Peter Singlehurst

Bucky Moore · May 5, 2025

Investors must abandon the canonical 'works in the first twelve months or not' timeline, because the most interesting software companies now require solving deeply technical problems that can take multiple years

The low-hanging fruit of buildable software businesses has been picked, so the best companies involve never-before-solved technical problems; there is growing evidence of companies like Figma and Clay taking five or six years before takeoff, and something that takes years to get right is often more defensible and hard to replicate

Scope: Clay example given with 'if I'm not mistaken'

27:30 20VC Exclusive: Why Mega Platforms Will Win in VC | Why You Cannot Do VC If You Do Not Do Pre-Seed | Why Market Sizing is BS | Where Will Foundation Models Build/Buy Apps vs Where Will They Not with Bucky Moore

David Cahn · Oct 27, 2025

The narrative that good companies raise seed, A and B within twelve months and grow revenue immediately is false and not how most successful companies work

His two most exciting investments, Clay and Juicebox, each spent three to four years 'in the wilderness' figuring out the product before growth

Scope: based on his own portfolio examples

42:57 20VC: Sequoia's David Cahn on The Winners and Losers in AI | The $0-$100M Revenue Club: Is Triple, Triple, Double, Double Dead? | The Future of Defence: Who Wins and Who Loses | How to Analyse Margins and Growth Rates in a World of AI

Harry Stebbings · Oct 27, 2025

How long a company takes to reach $1M in revenue doesn't matter; what matters is how long it takes to go from $1M to $50M

42:33 20VC: Sequoia's David Cahn on The Winners and Losers in AI | The $0-$100M Revenue Club: Is Triple, Triple, Double, Double Dead? | The Future of Defence: Who Wins and Who Loses | How to Analyse Margins and Growth Rates in a World of AI

David Cahn · Oct 27, 2025

Time from $1M to $50M is a historically strong leading indicator of company outcomes

The data he has looked at suggests it is a historically good algorithm

42:41 20VC: Sequoia's David Cahn on The Winners and Losers in AI | The $0-$100M Revenue Club: Is Triple, Triple, Double, Double Dead? | The Future of Defence: Who Wins and Who Loses | How to Analyse Margins and Growth Rates in a World of AI

Jake Saper · Mar 10, 2025 · hedged

The replacement benchmark for great software companies should be 'quadruple one-twenty' — roughly quadrupling revenue year over year while holding net dollar retention at 120% or above

Fast growth is now achievable, but retention is the thing that hasn't come home to roost — most of these businesses have no two- or three-year retention data yet, so durability has to be built into the benchmark

Scope: offered as a proposed formulation ('if I were to posit'); retention data for these companies does not yet exist

38:52 20VC: Lessons from Investing $2BN and Returning $8BN in Cash | Why Most Venture Partnerships are Broken | We Sold Salesforce Early and Lost Out on Billions | Are The Best Deals Always Expensive and Competitive with Jake Saper @ Emergence Capital

Jake Saper · Mar 10, 2025

AI-era companies that prove 120%+ net dollar retention while sustaining current growth rates will be generational companies

Scope: conditional on retention actually holding up

40:07 20VC: Lessons from Investing $2BN and Returning $8BN in Cash | Why Most Venture Partnerships are Broken | We Sold Salesforce Early and Lost Out on Billions | Are The Best Deals Always Expensive and Competitive with Jake Saper @ Emergence Capital

Max Junestrand · Jan 26, 2026

NRR is not currently a meaningful metric for Legora

So much of the growth comes from new contracts rather than renewals of 2024 contracts — they added $7M ARR in a single day in December 2025, more than 2023 and 2024 combined — so real retention numbers won't be known until 2026

28:21 Hypergrowth makes retention metrics uninformative for now

Luke Harries · May 23, 2025

Entire waves of AI companies will now cross the traditional $100M ARR IPO threshold

B2B and enterprise companies that solve real problems and retain customers will hold that cohort for a while, so a company already at $10M of revenue is on track

55:51 An entire wave of ai companies will cross the 100m arr threshold not just exceptions

Harry Stebbings · Apr 4, 2025 · hedged

Speed to revenue milestones no longer works as a quality heuristic in AI, because multiple products reach $10M ARR in months and will probably be dead in two years

AI has compressed revenue ramps far beyond the old eighteen-months-to-$10M gold standard, breaking the signal

9:30 Fast revenue ramps no longer signal quality since many die within two years

Mati Staniszewski · Sep 8, 2025

Speed of ARR growth does not matter as a metric

66:40 Arr growth speed is not a meaningful metric independent of horizon

Jake Saper · Mar 10, 2025 · hedged

The retention outliers will be the businesses that find a sticky wedge that lets them endure

44:23 Outliers that find a sticky wedge outperform the median on retention