Intuned (YC S22) Got Into YC With a Completely Different Idea. How Much Does Your Idea Matter for F26?

Intuned hit the Hacker News front page today with a Launch HN for its browser automation platform. Buried in the middle of the post is the most useful sentence for anyone working on a YC F26 application: "we were accepted into YC for a completely different idea."

Founders Faisal Kilaiwi and Ahmad got into the Summer 2022 batch, then pivoted during the batch itself, after batchmates kept asking whether RPA-style automation could fill API gaps in their products. Four years later, that pivot is the company. So if YC funded a team whose original idea didn't survive its own batch, how much does the idea you submit in your F26 application actually matter?

What actually happened with Intuned

The backstory, in the founders' own words: they entered YC S22 with a different product. During the batch, several batchmates approached them because of Faisal's background at UiPath, asking whether browser automation could stand in for missing APIs. When it was time to pivot, they went back to those founders, dug into the problem, and discovered the real pain wasn't building browser automations, it was maintaining them when websites change and selectors break.

That became Intuned: browser automations written and healed by an AI agent, run as code on managed infrastructure. They launched it on HN today, in 2026, four years after their batch.

Notice what carried them through the pivot. Not the original idea. Domain expertise (UiPath), proximity to real customers (batchmates with a concrete problem), and the willingness to throw away a nearly-shipped version of their agent when better models made a different architecture possible.

Does YC fund the idea or the founders?

YC's history answers this clearly. Reddit's founders applied with a mobile food-ordering app; Paul Graham rejected the idea, then invited them to come back with something else. Justin.tv was a single lifecasting channel before it became Twitch. Brex's founders entered YC pitching a VR company before moving to corporate cards. Pivots inside or shortly after the batch are a normal part of how YC companies find their footing.

Paul Graham wrote in "What We Look for in Founders" that determination, not brilliance of the initial idea, is the number one predictor. YC partners repeat a version of this every cycle: ideas are cheap to change, founders are not.

So the honest answer to "does my idea matter?" is: less than most applicants think, but not zero. And the way it matters is specific.

How your idea actually gets evaluated

YC readers use your idea as evidence about you. Three things they can read directly off your application:

Clarity of thought

Can you explain the idea in two plain sentences? The application's short-answer format is a thinking test disguised as a form. A muddled description of a good idea reads worse than a crisp description of a mediocre one.

Earned insight

What do you understand about this problem that others don't, and how did you earn it? Intuned's pivot worked because Faisal had years inside UiPath and could see exactly where RPA broke down. If your application shows that kind of specific, lived knowledge, the reader trusts you to navigate a pivot if one comes.

Customer proximity

Intuned found its real business by talking to batchmates with a concrete problem, then iterating with a handful of customers through 2024. An application that shows you in constant contact with users signals that, whatever happens to idea number one, you'll find idea number two quickly.

What this means for your F26 application

First, stop polishing the idea and start strengthening the evidence around it. A week spent getting three more user conversations or shipping one more iteration moves your application more than a week spent rewording your one-liner.

Second, don't hedge with three ideas. Applicants sometimes try to de-risk by presenting multiple directions. That reads as lack of conviction. Pick one, commit, and let your reasoning show. YC knows you might pivot; they want to see how you commit anyway.

Third, answer the "why you" question like it's the whole application, because functionally it is. If a partner believes you are the right people, an imperfect idea is a detail. If they don't, a perfect idea won't save you.

Fourth, show motion. Intuned scrapped a nearly-finished agent when Claude-style coding harnesses made a better approach possible, and rebuilt. Applications that show you changing your product in response to evidence, even small evidence, demonstrate the exact muscle YC is betting on.

Will YC hold a 'small' idea against you?

Sometimes, at the margin. If two teams look equally strong, the one pointed at a bigger market wins the tiebreak. But the inverse failure is far more common: weak teams hiding behind enormous TAM slides. Partners have seen thousands of "AI agent for X" applications this cycle. They are not selecting the biggest claimed market; they are selecting the teams most likely to still be moving fast in four years, the way Intuned still is.

One practical way to pressure-test which signal your application actually sends is to have someone who has been through the batch read it cold. YC Roaster exists for exactly this: YC alumni review your application and tell you whether it reads as "strong team, evolving idea" or "idea in search of a team." That distinction, as Intuned's story shows, is the one that decides outcomes.

The takeaway from today's Launch HN is not that ideas are worthless. It's that YC funded the founders, and the founders eventually found the idea. Write your F26 application so the reader can see the founders who would survive the pivot.