OH SNAP! Spiegel Said the Quiet Part Out Loud: Distribution Is The Only Moat Left

2 — the number of consumer apps Evan Spiegel says broke through in the last fifteen years. Two, in a decade and a half of unprecedented venture funding, frontier-model handouts, three trillion dollars of M&A, and the largest concentration of engineering talent in human history. Snap was one. Spiegel told Lenny on Sunday that every meaningful feature his company invented — Stories, swipe nav, AR lenses, Specs — got cloned inside a year. Software isn't a moat. Distribution is.

Write down what about your business is distribution-shaped, not feature-shaped. Spiegel said every Snap feature got cloned inside twelve months — Stories, swipe nav, the camera-first UX, AR lenses, all of it. He's a public-company CEO saying out loud what most pitch decks won't. Run the same exercise on your product. Which parts can a competitor with bigger distribution copy in a year, and which parts are anchored to a pipe they can't replicate? If the second list is empty, you've got a feature, not a business.

Run the Spolsky question on your category before lunch. Tomasz Tunguz mapped Anthropic onto Google's 2003 Spolsky playbook on Friday — MCP, Claude Code Security, Cowork, Claude Design, Interactive Apps. Each one is a free or near-free complement shipped to drive usage of the actual product, the model. Same play that killed Garmin and Hotmail. The conversation to have with your team this morning: which of our products would still have demand if a frontier lab shipped a free version next week? The ones that survive that question are your real business. The ones that don't are the part of the company you should be pivoting now, while you still have the cash to do it gracefully.

Stop letting your VC, your bank, or your advisor sell you the Rolodex. Ask for the apparatus. A community manager and an annual founder summit is not distribution. A podcast with millions of subscribers, a placement team that lands founders inside enterprise buyers, a vertical-specialization stack that compounds — that is. The Rolodex was a real moat for thirty years. It still is, for one-time door-opening. But if you need I can get your product in front of 100,000 people on Tuesday and again next Tuesday, that's a different product. Make sure the firms you're choosing actually have it.

Today's actions all touched on distribution as moat — what it is, what it isn't, and what to do when yours doesn't hold up. That's the work Anthony and I have spend decades helping operators sort out. If you're wondering where to start, that's the conversation we're built for.

Outsider Labs Advisory →

Monday, April 27

1. Snap's CEO confirmed what every pitch deck has been pretending isn't true. Evan Spiegel told Lenny Rachitsky on Sunday that only two consumer apps have broken through in fifteen years. Every Snap feature got cloned. Pure software is no longer a defensible business — hardware and distribution are the only moats he can see. The framing should be the first slide in every consumer pitch deck this quarter. (Full analysis below.)

2. Tunguz drew the diagram. Anthropic is running Google's 2003 playbook. Tomasz Tunguz at Theory Ventures ran Joel Spolsky's "commoditize your complement" framework over Anthropic's product line. The tables look identical to Google's 2003-2008 lineup. Different decade. Same flywheel. Your margin is their opportunity. (Full analysis below.)

3. OpenAI shipped GPT-5.5 wide in ChatGPT. The API stayed locked. Marcus Schuler at Implicator caught the cleanest tell yet that the labs know the game. GPT-5.5 went broad where OpenAI controls the surface, narrow where third-party builders could wire it into their products. Pricing doubled. UK AISI found a universal jailbreak during testing. The model race just moved from the leaderboard to a hallway with a velvet rope. (Full analysis below.)

4. Adobe just bet the next decade on platform plus audience. Liam Lawson at The AI Report filed from Vegas this weekend. Jensen Huang's keynote: 100% of NVIDIA's software engineers are now paired with agents. Adobe's positioning underneath: become the connective tissue stitching NVIDIA, Anthropic, OpenAI, Google Cloud, AWS, IBM, and Accenture into a single agentic content supply chain. Output is no longer scarce. Judgment is. The taste layer becomes the product.

5. Anthropic moved Claude Code from $20 Pro to $100+ Max. OpenClaude v0.6.0 shipped the same day. A 5x pricing hike on the developer-facing surface, met inside the same news cycle by an open-source community release. This is the substrate moving under the application layer in 72 hours — exactly the pattern Tunguz mapped. The lab tries to monetize the lead. The community ships a free alternative the same afternoon. The compute-competitor paradox isn't theoretical.

Distribution Is The Only Moat Left

Spiegel sat with Lenny Rachitsky on Sunday and named the thing publicly. Only two consumer apps have broken through in fifteen years. Every meaningful feature Snap invented got copied by a bigger competitor inside a year — Stories went to Instagram in 2016, Pixy got cloned, swipe nav, camera-first UX, AR lenses with face tracking. The only thing the clones can't replicate, in his telling, is the distribution layer underneath. And the distribution layer in consumer is now hardware-shaped, which is why Snap is betting the Specs franchise has to work.

Sit with that number. Two consumer breakthroughs in fifteen years. That period produced the largest pool of engineering talent and venture capital in history. SaaS multiples were rewriting financial textbooks. Founders had unlimited cloud credits, free model APIs, and an industry telling them software was the highest-margin business ever invented. Two products broke through. Software was never the moat.

Tunguz drew the diagram on Friday. He took Joel Spolsky's twenty-three-year-old "commoditize your complement" framework and laid it side-by-side over Google's 2003-2008 lineup and Anthropic's 2024-2026 lineup. The tables look identical. Free Maps, free Gmail, free Chrome, free Android — none of those products were the business. Search advertising was the business. Each free product was a strategic move to keep the toll booth open between the user and Google's actual revenue stream. Anthropic is running the same play with MCP, Claude Code Security, Claude Cowork, Claude Design, Interactive Apps. Each is a free or near-free complement to the real product, the model. Figma's stock dropped 14% in three trading days when Claude Design shipped — and Figma was paying Anthropic for compute the entire time. Bezos used to walk into rooms and tell SaaS CEOs your margin is my opportunity. The labs are running the same line on application software in 2026, except instead of competing on price they're shipping a free version to drive usage of the model underneath.

OpenAI shipped GPT-5.5 on Thursday. The benchmarks read like a victory lap. The release calendar told a different story. The model went wide inside ChatGPT and Codex, where OpenAI controls the surface. The API stayed locked. Anthropic ran the inverse play with Mythos two weeks ago — frontier capability behind a forty-organization whitelist. Both labs are now selling frontier capability through trust, not documentation. The model race left the leaderboard. The product race is now about who controls the door. Pricing doubled on GPT-5.5 list rates. DeepSeek V4 shipped Friday on Huawei silicon at one-eighth the output price, runs on consumer hardware. None of this is about whether the models are good. It's about who owns the access layer between the model and the customer.

The pattern isn't new. From 2008 through 2016, three names dominated venture conversations: Fred Wilson at USV, Paul Graham at YC, Mark Suster at Upfront. Different funds, different theses, same actual product. They were the most prolific bloggers in Silicon Valley. AVC, pg's essays, Both Sides of the Table — one-person media franchises that made their firms famous and their partners into industry stars. If you wanted into YC in 2014 the prize wasn't just the SAFE check or the tutorial-style office hours. It was getting inside the gravitational pull of one of the most-read writers in the Valley. a16z took that one-person model and industrialized it. Marc didn't invent distribution-as-VC. He scaled it. The question for everyone else is what happens when every VC has a podcast, every senior partner has a Substack, every X account has a content team behind it, and the platforms that made AVC famous have commoditized attention to the point that one extraordinary writer in 2012 is competing with fifty well-organized media operations in 2026. The credential is still real. The distribution layer underneath it is now operating at a scale the one-person-blog model can't match.

Three data points in seventy-two hours. Spiegel said it. Tunguz framed it. OpenAI's locked door enacted it. The CEOs already know. The pitch decks haven't caught up yet.

At COAI today: Full Signal/Noise briefing — the Spiegel confession, the Spolsky play on application software, the Pallotta-era Rolodex versus the a16z agency, and the consumer-tech opportunity nobody's assembled yet — at getcoai.com.

— Harry and Anthony

Sources: