ExpertsFM20VC: Why Now is the Time for the Application Layer | Why OpenAI & Anthropic Won't Win the App Layer | Why Startups Should be TokenMaxxing | Why VCs Should Reduce Weighting on Price & Ownership in an
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch · July 6, 2026
5 Selected Lessons:
heuristic · 4 reasoning_pattern · 1
Infrastructure-build-out precedes application-layer value creation
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why Now is the Time for the Application Layer | Why OpenAI & Anthropic Won't Win the App Layer | Why Startups Should be TokenMaxxing | Why VCs Should Reduce Weighting on Price & Ownership in an (July 6, 2026) · at 9:33 ▸
When this applies
when a new technology's foundational infrastructure has just been built and investors must decide where value creation shifts next
The takeaway
After a scarce, capital-intensive layer matures into a reliable utility, value creation migrates to the proliferation of lower-capital applications built on top of it.
Why it works
Once capital-intensive infrastructure (compute, models, chips) reaches sufficient maturity, the marginal advantage shifts to the many teams who can now cheaply build products on top of it, as happened when fiber/broadband enabled the internet application layer.
One level up
Track where capital intensity is decreasing — that is where the next wave of founders and returns will concentrate.
Where it stops holding
This stops working when the infrastructure layer itself keeps absorbing most of the value because it never truly commoditizes (e.g., if one lab reaches runaway recursive self-improvement and application layers can't differentiate).
Detail
approach: Compare the current AI infrastructure buildout (OpenAI, Anthropic, xAI, SpaceX capital intensity) to the historical fiber/broadband buildout before the internet application layer emerged.
signals:
- massive prior capital expenditure on infrastructure
- infrastructure reaching functional maturity
- proliferation of new application ideas emerging
rationale: History (the internet) shows that once infrastructure is built, value shifts to the many applications built cheaply on top of it.
In their words
It reminds me a lot of the early days of the internet when we built out fiber and we built out broadband, and then we had these new technologies to play with. And the Internet came along, and an application layer came along and took advantage of that new technology. To me, it feels like we are back in that mode right now where we've infrastructure, and now it's time for the applications to be built.
Market winners rarely take the whole pie
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why Now is the Time for the Application Layer | Why OpenAI & Anthropic Won't Win the App Layer | Why Startups Should be TokenMaxxing | Why VCs Should Reduce Weighting on Price & Ownership in an (July 6, 2026) · at 35:20 ▸
When this applies
when deciding whether to enter a category where a dominant leader has already emerged
The takeaway
In a sufficiently large market, even a clear leader leaves most of the market open — don't treat apparent winner-take-all dynamics as literal.
Why it works
Even strong category leaders typically only capture roughly a third of a market, because customer needs, trust, and product fit fragment demand, leaving a large remaining pool.
One level up
Perceived market concentration is usually overestimated relative to the actual addressable remainder.
Where it stops holding
This stops working in markets with strong network effects or winner-take-all dynamics (e.g., certain platform/marketplace businesses) where concentration really can approach 80-100%.
Detail
rule: Expect a market winner to capture roughly 30% of a market, not 80-100%.
threshold: ~30% share for the leader; ~70% remaining open to others.
rationale: Observed in the AI coding tools space where Claude/Anthropic Codex has 'run away' yet Lovable, Cognition, and Replit have independently crushed it too.
In their words
There usually isn't a market winner that takes 100% of the market or 80% of the market. Usually, you see is that the market winner takes something like 30% of the market, right? Which means the 70% remaining is totally up for grabs.
Mike's interpretation: context accumulation may act as a retention moat against bundling incumbents
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why Now is the Time for the Application Layer | Why OpenAI & Anthropic Won't Win the App Layer | Why Startups Should be TokenMaxxing | Why VCs Should Reduce Weighting on Price & Ownership in an (July 6, 2026) · at 39:43 ▸
When this applies
when a focused product competes against a bundling incumbent that can offer a 'good enough' version for free
The takeaway
In Mike's view, accumulated user-specific context can serve as a retention asset that a bundling competitor's feature parity alone may not fully neutralize.
Why it works
Mike suggests that once users accumulate rich, valuable historical context/data inside a product, switching away could mean losing that accumulated asset, which he believes creates switching costs independent of raw feature parity.
Where it stops holding
This stops working when the context itself is easily exportable/transferable to a competitor, or when the bundling incumbent also captures and migrates that context (e.g., via data portability tools).
Detail
rule: Mike believes that being first and accumulating organization-wide context may matter more than raw product quality when fighting a bundling incumbent.
threshold: N/A — qualitative moat, illustrated via Slack's continued $27B value despite Teams bundling.
rationale: In Mike's experience, if an org has built out rich, accumulated context inside a product, giving that up has a real cost, which he suggests protects the incumbent startup even against a 'meh' but bundled competitor.
In their words
What I've also learned about AI products and granola is that context is extremely valuable, right? If granola gets inside of your organization and everyone in the organization starts using it and producing and accumulating all of this amazingly rich and valuable context, that's not something you as an enterprise want to give up.
Fund size must match round size to preserve meaningful ownership
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why Now is the Time for the Application Layer | Why OpenAI & Anthropic Won't Win the App Layer | Why Startups Should be TokenMaxxing | Why VCs Should Reduce Weighting on Price & Ownership in an (July 6, 2026) · at 42:00 ▸
When this applies
when deciding whether your available capital is large enough to secure a meaningful stake at the price/round-size the market demands
The takeaway
Your capital base must scale with the market's prevailing deal size if ownership percentage is a input to your return model.
Why it works
As round sizes and valuations inflate, a fixed check size buys a shrinking percentage of the company, so funds must grow proportionally to maintain target ownership.
One level up
Strategy constraints (ownership targets) impose minimum viable scale, not just maximum scale.
Where it stops holding
This stops working once a fund stops optimizing for ownership percentage and instead optimizes for cash-on-cash multiple, which Mike says happens naturally at later stages (Series B/C and beyond).
Detail
rule: You can't do a credible Series A strategy without a fund large enough (Harry suggests ~$400M) to buy decent ownership at today's $80-150M post-money rounds.
threshold: Series A posts now commonly $80M, $100M, occasionally $150M.
rationale: In later rounds it's very hard to buy ownership, especially as a small fund, so ownership-sensitive strategies must concentrate capital at seed/Series A.
In their words
Series As are expensive now. They're 100,000,000 posts. They're 80,000,000 posts. They're occasionally 150,000,000 posts. And so, if you wanna write a check into around that size and get decent ownership for your fund, you have to have a slightly bigger fund than you had three years ago.
Don't project your own ideas onto founders
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Why Now is the Time for the Application Layer | Why OpenAI & Anthropic Won't Win the App Layer | Why Startups Should be TokenMaxxing | Why VCs Should Reduce Weighting on Price & Ownership in an (July 6, 2026) · at 49:08 ▸
When this applies
when an experienced operator-investor is tempted to impose their own playbook on someone else's venture
The takeaway
Evaluate and back a team's own judgment and plan rather than betting on your external hypothesis of what they should do.
Why it works
Even a correct external idea about what a company should do fails as an investment thesis if the team doesn't actually execute that idea, since the bet's success depends on the team's own judgment and plan, not the observer's.
One level up
The value of an outside opinion is capped by whether the people actually executing will act on it.
Where it stops holding
This stops working when the founder explicitly solicits the operator's input and would act on it — Mike notes you can still offer ideas if the founder wants them.
Detail
rule: Don't project your own operating ideas onto a founder's company when evaluating or supporting them.
rationale: If you make your investment bet based on your own evaluation of what a company should do, and the team executes a different plan, you made a bad bet — because it's the founder's company and their judgment that matters.
In their words
But if you make your bet based on that evaluation and that judgment, and that team doesn't do those four things, you made a bad bet. So don't project. Don't project.