Equity Podcast

Matt Murphy on why Anthropic's harness beat the model

Matt Murphy· Partner at Menlo Ventures at Menlo Ventures
·~32 min·English·TechCrunch
AI CompanyBusiness StrategyAgentsLLM
TL;DR

Menlo's Matt Murphy explains Anthropic's stunning revenue leap as less about model quality than the harness around it — Claude Code, MCP, and Skills — plus a VC willing to pay outcome-level prices to back the inevitable strong number two, in a wave that spreads product-led with almost no friction.

01Core Mental Model

The Harness, Not the Model

Anthropic's breakout was not the model alone — it was the harness around it, Claude Code, MCP, and Skills, that let people actually use the model without re-platforming their whole stack first.

obviously they took a strong model to power all that, but if they didn't do the rest of this stuff, the market wouldn't be growing nearly as quickly as it is.

— Matt Murphy, Equity Podcast
Key Insight
The harness is a distribution strategy wearing product clothing. A raw API forces every customer to re-architect before they get value, which throttles adoption; Claude Code, MCP, and Skills remove that re-platforming tax, so the model's quality finally converts into revenue at the speed the numbers show.
02The Bet

Your Outcome, Not Your Entry Point

Menlo backed Anthropic at a pre-revenue, four-billion-dollar-plus Series C valuation and then led its Series D — paying what used to be a home-run outcome just to get in — on the thesis that these markets always mint a strong number two.

These are numbers where you would be grateful and fortunate if that was your outcome, not your entry point.

— Matt Murphy, Equity Podcast
Key Insight
The courage here was not conviction that Anthropic would win — it was paying an outcome-level price to enter. What made that rational was the number-two thesis: these markets are never dominated by number one, so a durable second player is close to guaranteed, and the ex-ChatGPT founder was the lowest-risk bet on who that second player would be.
03The Moat

The Three-Layer Moat

Anthropic stacked three barriers at once — an elite team with real compute multipliers, Google and Amazon as partners and backers, and a war chest of capital — and together they priced out nearly every would-be rival.

Like, how do you keep up with team, technical advantage, massive amount of capital?

— Matt Murphy, Equity Podcast
Key Insight
Each layer alone is beatable — a rival can hire, or raise, or partner. Stacked, they compound: you now need all three simultaneously to compete, which is exactly why the early foundation-model hopefuls Murphy references quietly faded rather than catching up.
04Trust & Safety

Go Slow to Go Fast

Murphy reframes Anthropic's caution — the Mythos rollout, the deliberate releases — not as PR drag but as a necessary pause that lowers the risk of hacks or a catastrophe slowing the whole industry, so everyone can keep going fast.

Like it would be even a worse black eye for the industry if you just put something out there and then you'd had all these hacks and things going down

— Matt Murphy, Equity Podcast
Key Insight
Caution looks like cost only if you ignore the tail risk. Murphy's argument is industry-level: one high-profile failure early would invite exactly the heavy-handed, ad-hoc oversight that stalls everyone — which is why he leans toward an industry-run board of experts to govern the downside without killing the speed.
05Cost Optimization

Build First, Optimize Later

Enterprises do not optimize on day one — they adopt something that works, watch it become a runaway hit burning tokens, and only then tune cost, which is why the market is drifting toward a multi-model mix rather than a single provider.

And I think that's the way we're going through now is kind of diving in, getting going, building something that has value to now I need to optimize.

— Matt Murphy, Equity Podcast
Key Insight
The order matters: value first, cost second. A runaway hit generates the token bill that makes routing across Claude, open-source, and custom models worth the engineering. Murphy is careful not to predict the split — he floats a rough 50/50 only as one way the most sophisticated teams might go, while the majority without ML researchers likely stay almost entirely on Claude.
06Why This Wave Is Different

Product-Led, Not Sales-Led

Unlike past tech waves, AI needed no new device and no behavior change, so it spreads product-led instead of sales-led — bypassing the physics of hiring, ramping, and quota-ing a sales force.

this is more product led versus sales led. You know, a lot of companies are constrained by the physics of scaling a sales force

— Matt Murphy, Equity Podcast
Key Insight
Prior waves shipped, then waited — for devices, connections, comfort with payments. AI landed on hardware people already owned, in a chat box they already understood from search, so it skipped the behavior-change lag entirely. That is why the growth curves look nothing like SaaS: the ceiling that sales headcount used to impose largely falls away.
07The Long Tail

The Other 99%

Murphy's favorite pattern is arming the non-technical majority — Lovable lets the other 99% build software the way Uber and Rover unlocked drivers and dog-walkers, turning a shadow market into a platform.

it's basically putting the tools in the hands of people so that the other 99% can really just innovate, create.

— Matt Murphy, Equity Podcast
Key Insight
Arming the long tail is only half the pattern; the other half is keeping those makers on the platform. By stacking hosting and payments on top of build tools, Lovable lets non-technical creators graduate from one-off experiments to real, revenue-generating businesses without ever leaving — which is how a shadow market becomes a durable one.
08The New Normal

These Numbers Only Make Sense Now

Growth once unheard of — zero to a hundred million in a year — now recurs across more than a handful of AI companies, and after 25 years in venture Murphy calls this scale genuinely new, if still rarefied.

All these numbers don't make any sense except they make sense now

— Matt Murphy, Equity Podcast
Key Insight
The reset is double-edged. When zero-to-ten-million in a year used to mark the top 1% and now barely registers, the reset itself pressures founders even as it signals a genuinely larger prize — and coming from an investor with 25 years across internet, cloud, and mobile, the claim that this scale is unprecedented carries real weight.