Can AI-native startups beat incumbents that already own the data and distribution?

Harry Stebbings · Sep 27, 2023

AI differs fundamentally from prior platform shifts because the incumbent set is stronger than it has ever been

The winners in cloud (Google, Amazon, Microsoft) were the same companies that won before it, and mobile's incumbents were weak players like Motorola and Nokia — unlike today

47:13 20VC: "How Being a Founder Almost Killed Me"; We Have Lied to a Generation of Founders | The Hardest Truths About Being a Founder Revealed | Why AI Co-Pilot is BS, Seat Pricing is Over & User Interfaces are Stupid with Christian Lanng

Harry Stebbings · Oct 6, 2023

This generation of software incumbents is stronger and faster than in past platform shifts, has integrated AI well, and has already killed a generation of startups in between

Companies like Notion and Adobe abandoned existing product strategy and roadmaps to move fast on AI, and unlike the mobile shift — where the incumbents were an accidental pager maker and a failing division — today's incumbents are FANG-scale

41:39 20VC: Why Great Companies are Defined by How Many Things They Say No To, Why Being First Does Not Matter & Why Market Over Traction or Team is the Most Important Thing with Guillermo Rauch, Founder & CEO @ Vercel

Scott Belsky · Dec 20, 2023 · hedged

AI is a platform shift that favors incumbents who know their customers, unlike the cloud and mobile shifts which genuinely favored startups

Cloud and mobile required different developer stacks, go-to-market and platform relationships, whereas AI can be layered onto an existing product — a single prompt bar in Photoshop reimagines forty years of accumulated tools without building a net-new company from zero

Scope: 'I'm not actually sure'; credits Adobe's own execution too; every platform shift is different

32:12 20VC Roundtable: Spotify, Adobe & Linkedin CPOs on How AI Changes The Future of Product, Why AI is Now the Product, How TikTok Changed Product, Why Cost is the Biggest Barrier to LLM Usage & Why Incumbents Can Adopt AI Faster Than Any Prior Innovation Cyc

Harry Stebbings · Mar 20, 2024

Today's incumbents are more threatening to startup innovation than any previous incumbent set, making it hard to know where to place bets

They now have data advantages tying products together, enormous free cash flow (Microsoft at ~$350M/day), and unprecedented speed under strong leadership — unlike past incumbents who were laggards

Scope: framed as an investor's worry

46:45 20Product: Why Process is Killing Your Product Team and How to Remove it | Three Product Decisions Every Team Needs to Make | Why the Best Companies Build Movements and Lessons from Shopify and Atlassian on How to Do It Right with Jean-Michel Lemieux

Harry Stebbings · May 8, 2024

Incumbents have never been as strong as they are today, which makes it genuinely hard to see how AI value accrues to startups

Microsoft throws off around $350M of cash a day and Amazon's latest results were exceptionally strong

52:44 20VC: GV's Tom Hulme on Why Investing in Foundation Models is like Investing in "Power Stations", The Conventional Wisdom in VC that is BS & Lessons from a 24x Angel Track Record, 255x on Robinhood and Making Billions on Uber

Eric Vishria · Sep 25, 2024

Today's incumbents are paranoid and on top of things, so incumbent response has to be factored into any AI investment even though they will still move slower than startups

They are not dopes and don't have their heads in the sand, unlike incumbents in earlier platform shifts

Scope: incumbents will still be slower than the average startup

22:21 20VC: Benchmark's Eric Vishria on Where is the Value in AI: Chips, Models or Apps | Why Nvidia Will Not Be The Only Game in Town | The Commoditisation of Foundation Models | Which AI Apps Have Sustaining Value vs Hype and Short Term Revenue

Sridhar Ramaswamy · Feb 10, 2025

Mobile was a major platform shift in which the incumbents actually did well, contrary to the usual disruption narrative

Facebook pivoted from mobile web to apps, Google ported search and took mobile monetization from 10% of desktop to 100%, and Amazon wasn't disrupted; mobile produced new companies like Uber and Lyft but the incumbent generation had already smartened up

35:53 20VC: Why Model Providers Will Kill Many Startups Moving into the Application Layer | Why Deepseek is not a Threat to OpenAI & Why OpenAI Beats Anthropic | Apps vs Models vs Infrastructure: Where is Value in AI with Sridhar Ramaswamy, Snowflake CEO

Byron Deeter · Aug 25, 2025

The AI wave favors incumbents far more than the cloud transition did, which is a genuine reason for challengers to be scared

Cloud one had both a business model dislocation (license to subscription) and a delivery dislocation (on-prem to multi-tenant cloud); AI is only an extension of cloud, so incumbents keep their platform, data and distribution advantages

Scope: applies to fast-moving incumbents

15:00 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 Frankel · Aug 8, 2026

Google is a net winner in the AI shift and is arguably in pole position, though it is having to fight hard

Google comes from that environment and its ability to search and apply AI with context is incredible

Scope: they're having to fight like crazy for it

58:06 20VC: The AI Boom Will Create Enormous Roadkill: Who Wins & Loses | Why Founders Should Never Take Multi-Stage Money at Seed | Why Triple, Triple, Double, Double is Good Enough

Jerry Murdock · Aug 22, 2026

Meta, Google and Microsoft are all long-term holds and will remain trillion-dollar companies for at least a decade, because billion-user consumer bases act as a stabilizer that buys them time to catch up on AI

Consumers are slow to accept change, so a mass customer base is a buffer; all three have failed on coding agents but the scale gives them room to fix it

Scope: as an economic buyer holding the stock, not a claim they stay as important as today

56:04 20VC: The AI Bubble Will Burst: Half the Neoclouds Will Die | China: Should We Ban Chip Exports & Be Fearful of Chinese Open-Source | Mag7: Who Dies and Who Thrives: Why Meta is Meh and Microsoft is Mega

Vince Hankes · May 3, 2023 · hedged

It is too early to call whether incumbents or startups capture the value in this AI wave; the better lens is which new categories get created, where incumbency doesn't matter

Big technology cycles historically create new categories rather than incumbents extending into them — the .com era produced Google and PayPal, social produced Facebook and Twitter, mobile produced DoorDash and Uber rather than mobile CRM from Salesforce

Scope: acknowledges it may sound like a cop-out

25:53 20VC: The OpenAI Memo: Why Invest? Is it too Late to Catch OpenAI? Are OpenAI's Models Truly Defensible? Does the Value in AI Accrue to Incumbemts or Startups - Application Layer/Infrastructure? What Happens with Regulation? with Vince Hankes @ Thrive

Christian Lanng · Sep 27, 2023

Building an AI-first startup today is as pivotal and revolutionary as building the first cloud companies was fifteen years ago, and a new generation of AI-first-architecture startups will emerge over the next five to ten years

Cloud allowed software to be made in a fundamentally different way from on-prem and changed the enterprise world; the same architectural shift is happening now

Scope: not talking about ChatGPT-style consumer products; nobody knows yet what these companies will look like

46:30 20VC: "How Being a Founder Almost Killed Me"; We Have Lied to a Generation of Founders | The Hardest Truths About Being a Founder Revealed | Why AI Co-Pilot is BS, Seat Pricing is Over & User Interfaces are Stupid with Christian Lanng

Guillermo Rauch · Oct 6, 2023

Incumbents with the right AI strategy, foresight and investment still usually get disrupted at platform shifts, because the new way of doing things is so radically different they don't initially recognize the alternative as a competitor

Growth and user attention migrate to the new interface, so the old product keeps existing but stops growing — like desktop software after mobile, or AM/FM radio; only a rare incumbent like Meta manages to start over on the new platform

Scope: some companies do execute the shift well; acknowledges incumbents are often underestimated

42:22 20VC: Why Great Companies are Defined by How Many Things They Say No To, Why Being First Does Not Matter & Why Market Over Traction or Team is the Most Important Thing with Guillermo Rauch, Founder & CEO @ Vercel

Bastian Lehmann · Apr 8, 2024

Seemingly unassailable market dominance is routinely overturned, as with BlackBerry owning the smartphone market and Grubhub owning US food delivery

47:05 20VC: Postmates Founder Basti Lehmann on How the Uber Deal Went Down and How a $2.65BN Deal Turned into $5BN, Why Great VCs Add No Value and VC Value Add is BS Marketing & Why The Biggest Companies in History Will be Born Today and Replace Incumbents

Anton Osika · Aug 18, 2025

Many incumbent enterprises that can't adopt AI will be disrupted by cheaper, much better AI-native alternatives built from the ground up

If incumbents aren't enabled, someone else will be; companies built ground-up for AI plus exposure to customers and legal requirements can move much faster on customer experience, as visible in banking where old banks move slowly

Scope: Scale of the shift is uncertain and varies across enterprise segments; Incumbents retain benefits like trust from having been around a long time

38:37 20VC: Lovable CEO Anton Osika on $120M in ARR in 7 Months | The Honest Truth About Defensibility and Unit Economics for AI Startups | The State of Foundation Models: Long Grok, Short OpenAI, Why | Replit vs Lovable vs Bolt: What Happens

Alex Rampell · Jan 12, 2026

Any large company that doesn't have software at its core will be eaten by a competitor that does

Software-native entrants reverse-engineer into whatever product or service the incumbent provides

6:58 20VC: a16z's $15BN Fundraise with Alex Rampell | The Best Companies Have Hostages Not Customers | The Best Founders Materialise Capital, Customers and Labour | Mid-Sized Funds with Die and The Future of Venture Capital

Clay Bavor · Jul 4, 2026

The advent of a new technology is the moment when the deck is reshuffled in favor of small companies, making it the right time to start one

New technology resets incumbent advantages; in late 2022 it was clear language models were going to be a thing

5:42 20VC: Open Models vs Frontier Models: Who Actually Wins? | The $100,000 Token Budget Every Engineer Will Need | Why Forward-Deployed Engineers Are the Future of Enterprise AI with Clay Bavor, Co-Founder of Sierra

Fred Turner · Jul 18, 2026

Provider network contracting — built up over 50 to 100 years by incumbents — has been one of the biggest sources of staying power for large health plans, and agents can now collapse that moat.

There are ~1.2 million providers and ~60-70k contracts to negotiate; incumbents accumulated these slowly over decades, so a new entrant previously could not catch up, but an email agent can run the full research-outreach-negotiate-redline-sign workflow.

Scope: agent is email-only; larger hospital systems and doctor groups still require phone and in-person contact

50:02 20VC: $5BN in Revenue, 7 to 7,000 Employees in 9 Months, 206,000 Tests in a Single Day: The Craziest Story in Startups: Curative with Fred Turner

Harry Stebbings · Jun 5, 2023

The most attractive feature of a market like this is that incumbents on the supply side can't respond because doing so would cannibalize their own business

Scope: framed from a tech investor's rather than healthcare investor's perspective

38:31 20VC: The Largest Venture Backed D2C Consumer Exit; PillPack: $0-$300M Revenues in 5 Years & The Biggest Lessons Scaling the B2B Business to $300M in 2.5 Years with TJ Parker, Co-Founder @ PillPack

Alex Lebrun · Jun 19, 2023

Incumbents can use AI to stay competitive in their existing markets but will not invent the disruptive products that would kill their own business

They mostly sprinkle AI dust on existing products as enhanced features rather than rethinking the paradigm, and like Kodak with the digital camera they won't release what cannibalises their main revenue

Scope: some incumbents do move fast, so calling them simply slow is unfair

18:51 20VC: Why No Models Today Will Be Used in a Year, Why Open Will Always Beat Closed in AI, Why Proprietary Data is Less Important Than Ever And Why EU AI Regulation is a Disaster with Alex Lebrun, Founder & CEO @ Nabla

Noam Shazeer · Aug 31, 2023

A startup can move far faster than a big company and can launch products big companies won't

Large companies move too slowly because they worry about compromising their existing products

Scope: based on his own experience at Google

19:32 20VC: Spending $2M to Train a Single AI Model: What Matters More; Model Size or Data Size | Hallucinations: Feature or Bug | Will Everyone Have an AI Friend in the Future & Raising $150M from a16z with Noam Shazeer, Co-Founder & CEO @ Character.ai

Jeff Seibert · Nov 22, 2023

In outdated, massive industries like accounting, a new upstart will do AI-native workflow automation better than incumbents will

Incumbents like Intuit cannot be trusted to adopt the technology in a way that disrupts their own product lines

31:17 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits

Aaron Levie · May 22, 2024 · hedged

The route by which startups beat incumbents in AI is incumbent blind spots plus classic innovator's dilemma — the AI approach looks too unlike the incumbent's seat-based business model to register as a threat until it is too late

A seat-based customer support vendor sees AI agents as incompatible with how its customers buy, and the software needed to manage AI agents differs from what it would naturally build, so it doesn't respond until it's disrupted

Scope: "that is sort of not gonna happen exactly like that" — presented as the one disruptive angle rather than the likely outcome

24:04 20VC: Box's Aaron Levie on Predictions for the Next Wave of AI: Will Foundation Models Be Commoditised | How the Business Model of SaaS Changes Forever | Startups vs Incumbents: Who Wins | App vs Infrastructure Layer: Where is the Value?

Eric Vishria · Sep 25, 2024

For an existing vertical SaaS company to make the AI platform shift it has to be willing to burn down its existing business in short order, and many entrepreneurs will not do that

Platform shifts create a variant of the innovator's dilemma for the incumbent product

Scope: not the innovator's dilemma exactly as Christensen articulated it; concedes the data-advantage path Harry describes is possible

14:18 20VC: Benchmark's Eric Vishria on Where is the Value in AI: Chips, Models or Apps | Why Nvidia Will Not Be The Only Game in Town | The Commoditisation of Foundation Models | Which AI Apps Have Sustaining Value vs Hype and Short Term Revenue

Aravind Srinivas · Jun 15, 2026 · hedged

One could argue that Perplexity changed google.com more than any product manager at Google ever has

Nobody inside Google wanted to ship an answer engine or tinker with the interface generating $250B a year, yet AI mode now replicates Perplexity's font, citations, inline hyperlinks and suggested follow-ups

6:45 20VC: Micron Will Be More Valuable Than Meta | How Export Controls Helped Not Hurt China | Power is the Bottleneck to AI | Why Dario Has Done a Disservice to AI with his Labour Replacement Messaging with Aravind Srinivas, Founder @ Perplexity

Harry Stebbings · Apr 21, 2023

Speed of execution is the biggest determinant of whether a company achieves product-market fit

It is the clearest difference he observes between companies that achieve PMF and those that don't

39:07 20VC: Who Wins in AI; Startup vs Incumbent, Infrastructure vs Application Layer, Bundled vs Unbundled Providers | From 150 LP Meetings to Closing $230M for Fund I; The Fundraising Process, What Worked, What Didn't and Lessons Learned with Tomasz Tunguz

Sarah Guo · Apr 28, 2023

Speed is classically the only real advantage startups have over incumbents, and it matters more than ever now because the environment is moving at warp speed and large organizations cannot move that fast.

Some years a decade happens, and that is happening right now; it's hard to make a large organization move at that pace.

Scope: she frames it as a discouraging answer offered in the name of intellectual honesty

19:15 20VC: In AI Who Wins? Startups or Incumbents? What Happens to Wealth Inequality? Why Will $10BN+ Companies Only Have 10 People | Why Defensibility in Startups is BS & Speed is Everything? Why Large Groups Worsen Decision-Making with Sarah Guo

Ishan Mukherjee · Apr 4, 2025

Startups beat incumbents in AI sales software by competing on time to value — landing something that delivers ROI immediately rather than over quarters of system-integrator implementation

Incumbents have data and distribution but their implementations take quarters and system integrators to prove ROI; a startup's shipping machine can compress that and then earn the right to expand

46:55 20Sales: How the Best Sales Teams Use AI to Win Enterprise Deals | Sales Teams Will Be Dramatically Smaller | How to Ramps Sales Reps Way Faster | Why Unpaid Design Partners are BS | Why this Generation of Sales is Soft with Ishan Mukherjee @ Rox

Varun Mohan · Jun 2, 2025

Speed still decides outcomes even against incumbents with distribution: a startup with good insights that out-executes strategically will win the category

Google won in the early 2000s despite Microsoft already being a behemoth — that is how startups ever win important categories

24:10 20VC: Windsurf Founder on Will Model Companies Own the App Layer | Why Moats Do Not Exist in a World of AI | Why the Notion of Single Person $BN Companies is BS | Lovable vs Bolt & Cursor vs Windsurf: How Does it All End with Varun Mohan

Zach Lloyd · Oct 17, 2025 · speculative

Large companies ship AI slowly because of internal process gates like cross-org coordination and localization requirements, which is why startups win on speed

He can picture the PM meetings where a Gemini integration is blocked because it isn't translated into 70 languages

Scope: explicitly imagining the internal conversations rather than reporting them

57:30 20VC: The Startup Adding $1M ARR Every Week | Competing Against OpenAI's Codex and Claude Code: Who Wins | Why Gemini is Failing and GPT-5 Is Winning | Do Margins Matter in a World of AI | The Ugly Truth About AI Coding with Zach Lloyd, Warp

Harry Stebbings · Dec 1, 2025

Incumbents will not adapt; we are in a ten-to-twenty-year decline of incumbents as value transfers from them to startups that eat their lunch

Incumbents are unable and unwilling to adopt new tools because of data problems, permissioning, and abhorrent internal technology buying processes — as seen in European banks

19:21 20VC: Scale, Surge, Turing, Mercor: Who Wins & Who Loses in Data Labelling | Is Revenue in Data Labelling Real or GMV? | Why 99% of Knowledge Work Will Go and What Happens Then? | Why SaaS is Dead in a World of AI with Jonathan Siddharth @ Turing

Harry Stebbings · Jan 5, 2026

Revolut is a strong long-term bet because incumbent banks are structurally incapable of responding

Their structures, comp bands and policies prevent them from competing, and no one at a 100-year-old bank has real ownership — a hired CEO paid millions can simply leave with no downside

15:50 20VC: $0-$260M in Revenue in Three Years: How We Did It | You Need to Work Weekends to Win — Most Founders Aren't Ambitious Enough | The Revolut Playbook: Speed, Urgency, Extreme Ownership, and Zero Excuses with Alan Chang @ Fuse Energy

Fred Turner · Jul 18, 2026

Some of the biggest health insurers will fail to keep up on margins with AI-agent-native competitors, because the blocker is operational and people-ops politics rather than technical capability.

Capturing the margin improvement would require laying off tens of thousands of people, which shrinks internal fiefdoms, so incumbents will do it slowly over roughly ten years while newer players compete more effectively in the meantime.

Scope: applies to 'some' of the biggest insurers, not all; some incumbents do have the technical expertise; the change will happen eventually, just slowly

48:20 20VC: $5BN in Revenue, 7 to 7,000 Employees in 9 Months, 206,000 Tests in a Single Day: The Craziest Story in Startups: Curative with Fred Turner

Jake Saper · Mar 10, 2025

Jake has changed his mind: he no longer fears incumbents will capture most of the value from LLMs, because focused startups are outpacing them despite incumbents' data and distribution advantages.

Focus is the recurring lesson of startups — a company solving one narrow problem, like Unify in go-to-market, moves far faster and builds a product customers want more than Salesforce can.

Scope: 'still early days in this game and things could change'; biggest mind-change of the past twelve months

53:35 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

Zach Lloyd · Oct 17, 2025

Cursor beat Copilot because its autocomplete was simply better and switching costs into it were low, and it then executed fast on enterprise

Cursor's suggestions were more accurate and developers care about that; the wedge was strong and cheap to adopt

Scope: he does not think Cursor's agent product is that good

29:17 20VC: The Startup Adding $1M ARR Every Week | Competing Against OpenAI's Codex and Claude Code: Who Wins | Why Gemini is Failing and GPT-5 Is Winning | Do Margins Matter in a World of AI | The Ugly Truth About AI Coding with Zach Lloyd, Warp

Harry Stebbings · May 16, 2026

A focused company will beat a diversified incumbent in a market because it's their main course and only the incumbent's dessert

You care much more about a market when it's all you do

34:23 20VC: Lessons from Jensen Huang on "Founder Mode" | How to Know if OpenAI or Anthropic Will Kill your Company | How USV Liking Music Made Them $1BN on an Investment | The Five Year Desert to Product Market Fit & a $5.3BN Valuation with Shiv Rao @ Abridge

Alex Atallah · Aug 10, 2026

Most companies building routers/gateways are doing it because it's fashionable rather than to win, which leaves their gateways many months behind fully focused competitors

Copying puts you in a 'playing to exist' mindset serving your existing customer base, whereas a company 100% focused on routing builds a better product, as shown in internal benchmarks

Scope: about companies for whom routing is a side quest

14:48 20VC: Will OpenRouter Sell for $10BN to Stripe? | Why Chinese Open Models Are Beating America—and What Happens Next | Why Enterprises Are More Fearful of Anthropic and OpenAI Than China | Is the Routing Layer Becoming a Commodity with Alex Atallah

Guillermo Rauch · Oct 6, 2023

A challenger competing with a mature incumbent like AWS should not replicate the incumbent's sprawling product surface but bring simplification to the market

AWS's complexity reflects where they are now, not the approach that a new entrant should copy

21:34 20VC: Why Great Companies are Defined by How Many Things They Say No To, Why Being First Does Not Matter & Why Market Over Traction or Team is the Most Important Thing with Guillermo Rauch, Founder & CEO @ Vercel

Guillermo Rauch · Oct 6, 2023

New AI entrants can beat incumbents precisely by building far less product surface, since AI can deliver what previously required exposing many options and utilities

Incumbents' historical approach was additive — more menus, icons, options, and adding them got people promoted — so an AI-native product that removes all of that is a disruption they are emotionally and organizationally invested against

Scope: applies to software categories where AI can replace explicit option surfaces

39:30 20VC: Why Great Companies are Defined by How Many Things They Say No To, Why Being First Does Not Matter & Why Market Over Traction or Team is the Most Important Thing with Guillermo Rauch, Founder & CEO @ Vercel

Shiv Rao · May 16, 2026

Founders should focus mostly on their own product and use counter-positioning — building in ways the incumbent can't copy without damaging its existing business

Big companies have soft spots; if competing would hurt their current business, they're prohibited from responding directly and have to watch you

34:49 20VC: Lessons from Jensen Huang on "Founder Mode" | How to Know if OpenAI or Anthropic Will Kill your Company | How USV Liking Music Made Them $1BN on an Investment | The Five Year Desert to Product Market Fit & a $5.3BN Valuation with Shiv Rao @ Abridge

Harry Stebbings · May 8, 2024

Vertical AI providers claiming a 5% better UI than incumbents have no real advantage

Marginal UI improvement 'butters no parsnips' against incumbent distribution

53:44 20VC: GV's Tom Hulme on Why Investing in Foundation Models is like Investing in "Power Stations", The Conventional Wisdom in VC that is BS & Lessons from a 24x Angel Track Record, 255x on Robinhood and Making Billions on Uber

Andrew Feldman · Oct 6, 2025

Competing against a market-share leader means every dollar of revenue defaults to the incumbent unless you are ten times better, more aggressive and more creative

This has been true for him both against NVIDIA now and against Cisco for fifteen years prior

Scope: framed from his own competitive experience

59:49 20VC: Cerebras CEO on Why Raise $1BN and Delay the IPO | NVIDIA Showing Signs They Are Worried About Growth | Concentration of Value in Mag7: Will the AI Train Come to a Halt | Can the US Supply the Energy for AI with Andrew Feldman

Howie Liu · Aug 25, 2023 · hedged

AI-native startups win by opening up novel use cases and new user bases rather than by competing for the incumbents' existing use cases.

Newer companies can lean into the tech without supporting existing customer expectations or distribution models; products like Gamma and Tome are pursuing new ways of communicating visually rather than traditional slides, much as Prezi did.

Scope: incumbents like Google Slides and PowerPoint will for sure add gen-AI capabilities too

15:02 20Product: Enterprises are not Adopting AI Yet, When Will AI Break Into Enterprise, What are the Blockers, What Do Enterprises Need from AI & Why Services Companies Will Win in the Next 10 Years of AI Implementation with Howie Liu, Founder & CEO @ Airtabl

Anish Acharya · Feb 9, 2026

In this product cycle capable incumbents will win by making their existing category products better, while startups will own the native categories that did not exist before the cycle

That is the historical pattern of product cycles — Microsoft builds a better word processor, Google a better search engine, and we are already starting to see it

Scope: if history is any guide

11:47 20VC: Is SaaS Dead in a World of AI | Do Margins Matter Anymore | Is Triple, Triple, Double, Double Dead Today? | Who Wins the Dev Market: Cursor or Claude Code | Why We Are Not in an AI Bubble with Anish Acharya @ a16z

Sarah Guo · Apr 28, 2023

Company outcomes and market structure are unknowable at the very beginning; narratives about structural advantage belonging to incumbents or startups are just convincing stories

Markets are shaped by actors with agency, so what matters is who is playing rather than any intellectual narrative

Scope: framed as a belief change from earlier in her career

36:07 20VC: In AI Who Wins? Startups or Incumbents? What Happens to Wealth Inequality? Why Will $10BN+ Companies Only Have 10 People | Why Defensibility in Startups is BS & Speed is Everything? Why Large Groups Worsen Decision-Making with Sarah Guo

Mike Krieger · Mar 3, 2025

The AI wave can benefit both existing vertical SaaS incumbents and newly created startups; each faces a different challenge rather than one group clearly winning

Incumbents have established products, behaviors and relationships but must skate to where the puck is going without alienating existing customers, while startups lack data and relationships but can differentiate by painting the future and delivering value quickly

6:13 20VC: Anthropic CPO Mike Krieger: Where Will Value Be Created in a World of AI | Have Foundation Models Commoditized | When Do Model Providers Become Application Providers | What Anthropic Learned from Deepseek

Sebastian Siemiatkowski · Mar 1, 2021

Everyday retail banking is a trillion-dollar industry that has lost sight of the customer and survives on barriers to entry in capital, technology and regulatory protection, making it an ideal target for entrepreneurs

Incumbents have legacy systems and have been shielded by massive entry barriers, so customer focus, speed and new technology are a decisive advantage

Scope: about retail/day-to-day banking, explicitly not investment or corporate banking

32:15 20VC: Klarna Founder Sebastian Siemiatkowski on Scaling Europe's Most Valuable Private Tech Company, How To Motivate and Challenge Your Team Most Effectively & The Biggest Lessons From Working with Mike Moritz

David Schneider · Sep 11, 2024

Any market worth going after will have competition, but not all competitors are equal — incumbents harvesting revenue for the bottom line rather than serving customers can be decimated by a fresh product plus an aggressive go-to-market team.

At ServiceNow in 2011, BMC and HP Software were effectively already dead because they had stopped focusing on the customer while the market shifted to the cloud.

Scope: enterprise space; specific to incumbents in decline

9:15 20VC: Scaling ServiceNow to $5BN in ARR | Leadership Lessons from Doug Leone, Frank Slootman and Bill McDermott | VC Value Add: Is it Real and Why the Worst VCs are "Seagull VCs"

Reid Hoffman · Jun 10, 2024

AI remains a VC-investable class, but not by attacking incumbents where they are dominant — startups must find new territory rather than assault the fortress at the front door

Incumbents have network effects, moats, enterprise integration and customer lists in their core areas, so competing head-on (a new smartphone versus iPhone, or a frontier model to take on Google search) is hopeless; disruption is about diverting the river

Scope: excludes categories where incumbents hold dominant positions

17:07 20VC: Reid Hoffman on Foundation Models: Who Wins & How Do Incumbents Respond | The Inflection AI Deal: How it Went Down | Why Trump is a Threat to Democracy | The Future of TikTok | Lessons from Sam Altman, Brian Chesky and the OpenAI Board

Jake Saper · Mar 10, 2025

The answer to 'Google could just build this' is to start with a narrow problem Google won't solve as well and expand from there, even if the initial market looks far too small.

Veeva's entire global market for pharma CRM was $400M at investment — too small for a multi-billion-dollar business — yet Veeva is now worth ~$35B.

55:12 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 · Jun 19, 2023

Some incumbents — such as Notion, Adobe, and the travel expense management company — have moved very fast on AI

Notion and Adobe moved very fast, and Navan built its expense product solely on OpenAI very quickly

19:16 20VC: Why No Models Today Will Be Used in a Year, Why Open Will Always Beat Closed in AI, Why Proprietary Data is Less Important Than Ever And Why EU AI Regulation is a Disaster with Alex Lebrun, Founder & CEO @ Nabla

Harry Stebbings · Mar 10, 2025

Incumbents are shipping faster than ever, contradicting the received wisdom that they are slow.

Observed speed of shift at companies like Adobe.

54:51 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

Jesse Zhang · Sep 19, 2025

In AI, startups mostly start fresh rather than needing feature parity, because incumbents' existing customer base is baggage that forces backwards compatibility

Anything incumbents build must be compatible with all their locked-in customers, which slows them versus AI-native competitors

Scope: incumbents do know AI is here and are investing in it

16:36 20VC: Why 90% of Founders Build Startups Wrong | Why AI Growth Rates are Sustainable & Remote Work is BS and the AI Talent War | Competing with Brett Taylor and Sierra: Who Wins the Customer Service War with Jesse Zhang, Decagon

Jesse Zhang · Sep 19, 2025

None of the configuration work incumbents built for prior-generation chatbots and phone trees transfers to the AI approach, which levels the playing field

If you asked from scratch what the best way to bring AI to these teams is, the answer requires throwing away most or all of what's been built

17:24 20VC: Why 90% of Founders Build Startups Wrong | Why AI Growth Rates are Sustainable & Remote Work is BS and the AI Talent War | Competing with Brett Taylor and Sierra: Who Wins the Customer Service War with Jesse Zhang, Decagon

Mike Cannon-Brookes · Oct 13, 2025

This is a classical technology race in which incumbents like Atlassian — with high R&D investment, the same model access, 300,000 customers and tens of millions of end users — can win by infusing AI into their platforms faster than new entrants can build distribution

Existing businesses have identical access to the models plus an installed base, so the race is about speed of infusion versus startups' speed of arrival

37:13 Incumbents can win by infusing ai faster than startups build distribution

Harry Stebbings · Oct 13, 2025

The central competitive question is whether the startup can acquire distribution before the incumbent acquires innovation

39:23 Outcome hinges on a race between startup distribution and incumbent innovation speed

Fred Turner · Jul 18, 2026

The incumbent lab industry was structurally incapable of 10x-ing testing capacity because it is a low-margin business built around squeezing incremental efficiency, not rapid scaling

Quest and LabCorp are optimized to gain 1% margin by perfecting existing processes; the mindset, people and supply chain aren't built for 10x expansion, so you have to build from scratch

28:27 Incumbent efficiency mindset cannot scale fast

Clem Delangue · May 12, 2023 · hedged

Incumbents choosing the easy bundled AI solution creates a large opportunity for AI-native startups that train their own models to disrupt them.

AI-native startups that build and train their own models can build much better products than those that just consume APIs, as seen with startups building on Hugging Face.

16:32 Incumbents taking easy bundled ai route open the door for ai native challengers that train their own models

Clem Delangue · May 12, 2023 · hedged

Big companies struggle to build AI natively because it is a fundamentally different paradigm of building software, requiring scientists to work six months on an architecture before release, which slower large teams built on the old paradigm can't accommodate.

Based on what he observes in the field; large organizations move slower and started from a very different software paradigm.

24:18 Big companies struggle with ai native paradigm due to long uncertain rnd cycles

Christian Lanng · Sep 27, 2023

The strong-incumbent argument holds in consumer but not in enterprise, where the market is multipolar rather than winner-takes-all

Cloud platforms enabled SaaS to explode from ~1,000 to 35,000 companies and many unicorns; enterprise has layers of moats — compliance, regulatory, incompetence — that make decisions happen differently, so Thiel's power law applies to consumer but not enterprise

47:36 Enterprise is multipolar so incumbency binds less

Zach Lloyd · Oct 17, 2025

None of the big incumbents are doing a good job in the developer tooling / AI coding space

Microsoft screwed up Copilot, Amazon's coding product is only okay, Google has never been good at developer products, and Apple is always very slow — it's a big-company problem

59:08 Big tech incumbents consistently underperform in developer tooling

Michael Eisenberg · Jun 19, 2024

Unlike the internet, AI will be very hard for legacy companies to catch up on

With the internet you could keep your retail business and bolt on ecommerce or an intranet, but with AI if your data isn't set up right you simply can't get there — as with Buffett's insurance businesses discovering they had hundreds or thousands of databases that didn't talk to each other

12:17 Ai is harder for legacy companies to bolt on than the internet was

Jerry Murdock · Feb 28, 2026

Google's assets are better long term for the agent era, but the decisive question is who reaches a billion users first, and OpenAI converting ChatGPT into personal autonomous agents would be an insanely great business

Owning Gmail is a phenomenal asset for an autonomous agent business that OpenAI lacks; but OpenAI's 800 million chatbot users and a conversion to personal agents could win the race

53:09 Incumbent data assets win unless challenger reaches user scale first

Alex Rampell · Jan 12, 2026

A startup can win a software category if the rate of new company creation in the customer base is high enough, since new entrants are not hostages and will pick the best product

Stripe worked because Patrick and John bet their customers didn't exist yet; Mercury never took a customer from SVB before SVB failed but grew on newly created companies

23:24 Startups win only where new customer formation is fast

Aidan Gomez · Aug 19, 2024

Salesforce is not going to be displaced by AI-native competitors

Benioff is very cognizant of the threat and won't let it happen, and enterprise software is extremely sticky — it tends to stay for decades and is very hard to displace

41:45 Sticky enterprise incumbents with vigilant leadership resist ai native displacement

Harry Stebbings · Oct 6, 2023

A human-in-the-loop, assistive AI future favors incumbents with distribution and existing product suites, leaving startups with limited capital and weak moats nearly impossible to fund

Incumbents can service the assistive use case incredibly well off their existing distribution and product suite, which destroys the venture case for startups

39:09 Assistive ai future entrenches incumbent distribution advantage crowding out startups

Noam Shazeer · Aug 31, 2023

There will be many winners in this AI wave — big companies, startups, universities and individuals — rather than startups or incumbents winning

So much value is about to be created that there's room for multiple players, and hardware is progressing so fast that what takes a big company today will be doable in a university lab or a garage a few years later

20:04 Hardware progress democratizes capability so many different players big companies startups universities individuals win

Cliff Weitzman · May 9, 2026

Figma is a hold, not a sell, and will survive AI-native design competition

It's an antifragile company with a great team, excellent sales, real taste; Dylan Field built the original renderer and can rebuild again; in a fast-moving world betting on the Figma team beats betting on incumbents bad at innovating

31:01 Adaptive founder led incumbent team survives the shift

TJ Parker · Jun 5, 2023

Incumbents ignore a startup entirely until its volume becomes visible — PillPack got network access trivially as a small independent pharmacy and only faced termination notices after scaling from $10M to $70M run rate

PBMs process every transaction and see the volume, so at small scale they didn't know or care who PillPack was; growth is what triggered the response

38:38 Incumbents ignore startups until volume becomes visible then respond

Sarah Tavel · May 6, 2024

Incumbent services firms will not be the main beneficiaries of AI work-automation tools; a class of hungry up-and-comers will adopt the technology early, come in with different pricing and faster delivery, and take market share

Lessons from marketplaces: new technologies are best served for hungry up-and-comers who want to become the new incumbents, and their cost advantage passed to customers wins share

15:35 Hungry up and comers not incumbent services firms capture ai automation share

Bastian Lehmann · Apr 8, 2024

Incumbents capture most of the benefit in the current AI wave, but companies founded today will become larger in AI than today's largest companies, and they will start as overlooked toys

Incumbent advantage is obvious and not contrarian; transformative companies begin as things that look like toys and are not the most obvious application

46:17 Todays ai startups will surpass todays largest companies starting as overlooked toys

Mitchell Green · Mar 7, 2026

The biggest opportunity to disrupt incumbents today is in highly levered companies, in any sector, because leverage starves them of the cash flow needed to innovate

In 1999-2000 the retailers that went bust (Sears, Kmart, Montgomery Ward, Bed Bath and Beyond) carried huge leverage, while unlevered Walmart, Target, Home Depot and Lowe's bet the company on e-commerce and now dominate it

8:03 Levered incumbents are the disruptable ones

Saam Motamedi · Jul 15, 2024

Startups can now build a radically different CRM and become the next Salesforce

All three disruption conditions are now true: AI agents can materialize data views on the fly from inboxes and call recordings without Salesforce's data model, pricing is shifting, and the interface is becoming agent-based

20:45 Simultaneous shifts in data model pricing and interface let ai native challengers beat entrenched crm incumbents

Scott Belsky · Dec 20, 2023

The hard, differentiating part of AI product work is final-mile tuning that only the incumbent can do because it knows its customer and product intimately

Customers love products that are empathetic to their problems and meet them where they are, and only the company that knows the customer can finesse those moments

18:12 Final mile tuning from deep customer knowledge is the incumbents durable edge

Jesse Zhang · Sep 19, 2025 · hedged

Startups like Decagon and Sierra will out-execute Salesforce on product speed, but Salesforce's distribution advantage means counter-positioning against it is a constant strategic question

Startups execute faster on product; incumbents have far greater distribution

11:55 Product speed wins but incumbent distribution forces permanent counter positioning