Magic Patterns CEO Alex Danilowicz Says AI-Built Startup Sites Are Losing Users at the Signup Form – BigGo Finance

The most expensive mistake on a startup website right now is not an ugly hero image. It’s a signup form that requires an extra click.
That was the recurring finding when Alex Danilowicz, co-founder and CEO of AI design tool Magic Patterns, sat down to tear apart four startup landing pages on the Y Combinator Startup Podcast. His framing is blunt: the only thing that matters about a landing page is whether it converts, and from the outside nobody can know that. What he can see are the patterns that get in the way — and in 2026, those patterns have a recognizable signature. They are the fingerprints of sites generated by AI coding agents and shipped without a human deciding what the page was actually for.
“There’s a lot of stuff that looks really cool but then fails at the core goal,” Danilowicz said. The gap between the two is where the money leaks.
The sharpest critique landed on Ego, a product whose tagline reads “Agents are machines. We make them people.” The call-to-action button in the top right says only “Ronin,” and Danilowicz couldn’t tell whether Ronin was the product, a model, or a person. The actual explanation — a voice agent with personality, memory, and a mind of its own — sits halfway down the page. The waitlist requires an extra interaction to reach.
He has tested this on his own company. Magic Patterns briefly stripped the prompt input off its landing page in the name of creativity, and “our conversion rate totally tanked.” Host Aaron Epstein, who ran conversion experiments at Creative Market, put a number-shaped lesson on it: “When you had the input form visible and expanded directly on the page, conversion rate was much higher than requiring somebody to click through to then put in an email address or pop up a window or get to the form.”
Ego had a second problem that is becoming a tell: a fake input box that looks interactive and accepts no typing. “It matters because you can notice how distracted we are by this input trying to type and we’re distracted from the core thing which is whatever the CTA is supposed to be,” Danilowicz said. The page also nudges the user to scroll — another habit he attributes to agents, which constantly prod visitors to act.

Sentient OS, which sells “your Mac, sentient,” drew the most praise of the four for its hero. A laptop lid closes as the copy says “while you sleep,” then reopens to show completed work. Danilowicz called it the best version of animation — one that reinforces the message rather than decorating it.
Then the page scrolljacks, forcing the visitor to fight the scroll position to reach each stage. “One thing I never see a decacorn company ever do is any kind of scroll jacking,” he said, “because it does not convert and it’s frustrating.” Other issues: an oversized glow on the download button, a “free” label rendered so small and at such an angle it is unreadable, and a hero that bolds random words — a known marker of AI writing. On the unreadable label, his principle was simple: “If you have something important to communicate, it shouldn’t matter the angle that the user is looking at it. It should be painfully obvious.”
The fonts are the giveaway. Space Grotesk and Instrument Serif are the defaults Claude Code reaches for, Danilowicz said, and he predicts mid-2020s site design will look as dated in ten years as an 1980s kitchen does now. “I would encourage people just resist the urge to do the default that the agents are spitting out when they’re designing,” he said. “Use something that is more unique to you. And that’s the way to create more timeless design.” His practical fix is to build a brand book early — he regrets Magic Patterns not doing so sooner.
Saranovas, which describes itself as “institutional intelligence and decision infrastructure,” is visually clean and fails anyway. Danilowicz couldn’t tell what it does. The page segments the audience into “enterprise, middle market, private equity” and asks the reader to self-identify before explaining the problem — it reads like a pitch deck aimed at investors.
Epstein’s diagnosis is that vibe-coded sites often pitch the company rather than telling users what problem is solved, for whom, and with what proof. One element both flagged as forward-looking: a trust-and-security link sitting in the main navigation, which Danilowicz expects to become a standard pattern for AI products handling sensitive data.
Slashy, “your intelligent inbox,” was the strongest of the four. Epstein immediately understood it as a Gmail replacement in the Superhuman category. The page is long, well-spaced, and professionally designed — which itself signals an established company. But the logo reads as “Sashy” without the URL for context, the hero video is too dense and fast to follow, and the tabbed feature sections auto-advance on a timer. Epstein noticed the timer appeared to start before the section was in view, meaning a visitor could land on the fourth-most-important feature. The one animation that lands is a side-by-side showing a draft reply being edited to match the user’s writing style. Social proof is present but skewed to investors and founders, which made Epstein wonder whether the product is only for VCs.

Sentient OS runs locally and would presumably have access to everything on a user’s machine, which raises a question the episode treats as newly urgent: what data does it see, how is it used, what is the retention policy, and is it secure.
Danilowicz’s evidence comes from his own Y Combinator product, an open-source text-message analyzer that read iMessage locally. The hero of that page was “open source, completely on device” — the internet never touches the data — and that is what users cared about most. “Products that really lean into trust and privacy and go above and beyond to make users feel super comfortable around what data we’re going to have access to,” he said, “are the ones that are going to have more success of getting users to download and try it and spread the word.”
Danilowicz rejects the Silicon Valley habit of equating design with “pixel-perfect border radius.” Design, in his framing, is thinking through the user experience; the polished UI is one part of it. To build taste he looks at beautiful things — he mentions a Matisse exhibit in San Francisco — and recommends The Design of Everyday Things.
Operationally, Magic Patterns makes every employee do customer support, so engineers see firsthand when users cannot find the share button because two primary buttons are the same color. The team spends time on the extra 20% rather than shipping raw vibe-coded output. On speed versus quality, Danilowicz argues the trade-off has weakened: teams that skip the polish lose users who cannot understand the site or whose workflow is blocked, and AI-assisted coding still lets you ship fast while doing the work.
Epstein’s closing advice to founders is to show the site to users without revealing it is theirs, because otherwise users are too polite to tell the truth.
None of this is abstract for Danilowicz. Magic Patterns spent about nine months pivoting before it found the product — a component-library editor, a “Storybook on steroids,” a Chrome extension that published a localhost app to the web — none of which produced the ten users who both paid and actively used the thing. Once Magic Patterns launched in October 2023, those ten arrived quickly.
The company’s structural bet is being model- and feature-agnostic. Danilowicz contrasts this with Anthropic’s Claude Design, which he says runs only Anthropic models and exports only to Claude Code with no Figma integration. Magic Patterns exports to Figma for precision or to whichever coding agent a customer prefers, and its MCP server — an internal hackathon project originally expected to go nowhere — became one of the main ways people use the product.
That logic connects directly to the landing-page advice. Code is becoming the design medium, the frontier labs are all investing in coding agents, and model access is commoditizing. What stays scarce is customer context — design systems, PRDs, accumulated skills — and the human judgment to know when the page is getting in the user’s way. Danilowicz concedes he does not know what the future of design holds — “anyone who thinks that they know for sure is probably just making it up” — and says a canvas is not for everyone; the company is hiring a designer specifically to work out when to show a canvas versus a single chat thread.
For investors, the read-across is less about one startup’s tool and more about where value sits in the AI design stack. If every competitor can call the same models, differentiation moves to workflow, distribution, and trust — the same three things that decide whether a landing page converts. The companies that win the next cycle may not be the ones with the flashiest generated hero, but the ones whose signup form is already open on the page.
The design industry’s own Accenture problem looms in the background: as AI reshapes how services are delivered, the firms that own the customer relationship and the context survive the repricing, and the ones selling undifferentiated output do not. A landing page is just the smallest, most measurable version of that fight.
References:
The news and data on this website are for reference only and do not constitute investment advice or an offer to buy or sell. Information is sourced from exchanges and public sources, and may be delayed, interrupted, or updated. While we strive for accuracy, we do not guarantee timeliness, correctness, or completeness. Content may include external links for which we are not responsible. By using this site, you agree that we and our partners are not liable for any losses. Investment carries full responsibility; please carefully assess risks and consult professionals. If there are errors in the content, please contact us for correction.

source