Eduard Lataretu

Stripe just acquired OpenRouter in a deal reportedly worth more than $8 billion. OpenRouter is an AI infrastructure company that gives developers access to hundreds of AI models through a single platform. But the bigger lesson here isn’t the acquisition. It’s the reason I keep telling developers: BUILD. If you have an idea that you genuinely believe solves a problem, but nobody around you is building it, don’t wait for someone else. Build it. Put it in the market. Get users. Even if it isn’t making money yet, keep improving it. Make sure people are actually using it. Make sure the product works. Because you never know what happens next. A bigger company may eventually discover that your product solves a problem they were about to spend years and millions of dollars solving themselves. And instead of starting from scratch, they might decide: “Why don’t we just acquire them?” We’ve seen versions of this before. ByteDance acquired Musical.ly for roughly $800 million to $1 billion in 2017, then merged it with TikTok the following year. Facebook acquired Instagram and WhatsApp. Google acquired YouTube. The lesson isn’t: “Build something so somebody will buy you.” The lesson is: Build something valuable enough that somebody would want to buy it. So if you’re a developer sitting on an idea because you think it’s too simple, BUILD IT. If you’re waiting for investors, BUILD IT. If you’re not making money yet, BUILD IT. Get your first 10 users. Then 100. Then 1,000. Keep improving. You don’t need to know exactly where the journey ends. You just need to start building. The little product you’re building today could become the missing piece a much bigger company needs tomorrow. Don’t just have ideas. Build them.

1

YC's advice of "Focus on one thing" has killed more startups than bad products. I know because it almost killed ours. 3 months in, a close mentor sat us down and said: "You're doing too many things. Pick one and go all in." So we listened. We picked one approach and spent three months perfecting it. By the time we shipped, nobody cared. The market had moved. The irony? In trying to follow YC's advice, we forgot to follow their most famous one: talk to your customers. We were so heads down building the "one thing" that we stopped listening to the people we were building it for. Meanwhile, one of our engineers had been tinkering with something on the side. Turns out he'd been talking to users on his own and building out deep IDE integrations. His side project is now the fastest growing feature at Willow Voice. I realized that I had interpreted the advice as "focus on one solution" when the advice actually meant "focus on one problem." Dropbox focused on one problem but tested dozens of approaches. Airbnb focused on one problem but pivoted three times. They didn't focus on one solution. They focused on one problem. And they never stopped talking to customers while doing it. That's what we do now. Every engineer owns a bet end-to-end. Then every Sunday, we dogfood with Willow Voice and build out the feature in a few hours for fun. This way we can run a ton of experiments. Last month two survived! One of them doubled our activation rate. One metric. Multiple bets. Kill ideas fast and never stop talking to users 👉 — Written with Willow

54

4 Yorum

BREAKING: Google is offering buyouts to staff who aren't "all in" on AI 😳 Chief Business Officer Philipp Schindler sent a memo: everyone in the global business org needs to be "all in" on the mission and "embracing AI to have even greater impact." Eligible roles: solutions teams, sales, corporate development. Customer-facing roles and large sales teams are excluded. The message is clear: if you're not thriving in the fast-paced AI environment, here's the door. Google has offered similar voluntary exits over the past year for YouTube staff and employees affected by return-to-office mandates. This is the new reality. Companies are filtering for AI believers and letting everyone else self-select out. Do you think more tech companies will follow this playbook?

76

30 Yorum

A team getting bigger is rarely good for productivity. It is usually a scaling approach, with a productivity tax. I remember a PM years ago telling me his prior startup shipped much faster than “teams at Amazon.” Later, when he had his own org, the tune changed. The work did not get harder. But the system certainly did. At scale, speed drops for reasons that have nothing to do with motivation: 1. Live systems create constant operational pull: refunds, incidents, capacity work, bugs that only happen in production. 2. Customers make change expensive. A minor UI tweak can move revenue enough that you have to prove it’s safe. 3. Abstraction has a hidden bill. You split one service into two so that two teams can work, and then you spend months building and maintaining that glue. This is why small teams can outmaneuver big ones, even when the big company has money. It is not “startup magic.” It is fewer layers and fewer coordination paths. To learn more about abstraction costs and why efficiency drops as organizations grow, read on in my linked article.

25

4 Yorum

Anurag Goel was the 5th engineer at Stripe and part of the early team that built Stripe's API. One of the things he said about building product stood out to me:  Often you can have all kinds of arguments around how to build something, but really the way to win arguments is to show the code -- And to show the product as you think it should exist And then scaling it with good documentation, good design, and excellent customer support. Anurag is now the CEO of Render -- we had a great conversation on building MCP servers and the future of AI platforms. Tune in to the next episode of Request // Response -- dropping Thursday!

62

3 Yorum

This valuation tells founders who capital is for. Waymo raising at a roughly 110 billion valuation is not about revenue today. It is about category ownership tomorrow. Investors are paying for inevitability. The belief that autonomous mobility becomes baseline infrastructure. For founders this clarifies a harsh truth. Capital is still abundant, but it is highly selective. Markets are rewarding companies that look unavoidable. Not just impressive. That means clear leadership in one narrow problem. Clear proof that losing to you would be expensive. If you are building in a capital heavy space this changes the pitch. You are not selling traction. You are selling why no one else should win. The danger is copying the valuation logic without the substance. Big numbers only work when paired with dominant position. If an investor asked why your category should consolidate around you what would you say?

15

6 Yorum

Everyone's dunking on Stripe for paying $7 billion for OpenRouter (a metering tool). Stripe has done metering for ages, and routing is a commodity. They bought the markets primary LLM endpoint, and you cannot buy that later. I run 25+ software companies, and one of them is teaching me this lesson. Ayrshare is a social media API in our portfolio. It's growing like crazy right now, and the reason is kind of surprising. When you ask an LLM to build something that posts to social media, it picks Ayrshare, because LLMs are lazy. Ask a model to set up metering, or tooling for switching between models, and it reaches for the most complete, best-documented answer it already knows and moves on. Whoever is that answer wins the customer. OpenRouter became that answer for model routing. So the models propose it by default, and the company keeps growing without doing anything. Stripe acquired OpenRouter for $7.8 billion. Stripe already won this exact game once. They have the best API documentation you'll find anywhere, which is why every LLM proposes Stripe when someone builds a payment flow today. So my guess (and it's just speculation, they haven't said this) is they'll do the same thing with model selection, maybe even an agentic layer that auto-selects for you. Why not? Is it worth $7 billion? Honestly, I don't know. But looking at our own AI spend across the portfolio, which has grown in multiples over the last months (multiples, not percentages), this is becoming a fundamental piece of everyone's cost base. The only way to win an LLM's defaults is to already be the most complete, best-documented answer. If the models aren't picking your product, the fix is the product itself. The next few years will show whether being the default answer in an LLM is as durable an advantage as being the default answer in a search engine was.

14

5 Yorum

You can feel the anxiety in the tech hiring market, but it's a useful reminder that big offers are still being landed daily. A Hello Interview user just landed a $1M TC offer from Stripe 🤯 They were kind enough to share tons of great lessons and advice to other candidates here: https://lnkd.in/g4uMJKme

306

8 Yorum

The FAANG → startup gap is widening. When I was at Uber, we could still reliably hire great engineers out of FAANG. The velocity difference existed, but it was bridgeable. Now? For a seed or Series A team, the ramp is often too steep: ambiguity everywhere no guardrails shipping is the strategy “ownership” means end-to-end, not a slice Here’s what I think happens next as big tech layoffs continue: If you’re hiring: expect more noise. A larger pool of “available” talent won’t automatically translate to startup-ready talent. Your filter gets harder: slope > pedigree. If you’re a FAANG engineer: if you want risk, speed, and real ownership, jump earlier. After ~3+ years in a highly structured environment, the pivot gets meaningfully harder. Not because you’re not capable, but because the operating system is different. And I’m seeing a pattern: A lot of FAANG engineers want to move faster… But many early-stage startups aren’t prioritizing candidates who’ve spent 5+ years in big tech. Not a knock. Just a market reality. Curious if others are seeing the same thing: What’s your “threshold” where the FAANG → seed jump gets tough? 2 years? 4 years? More?

261

53 Yorum