Skip to content
Design HackerDesign Hacker
On this page

The AI slop era: why great websites get crickets now

Everyone can look good now. Nobody can fake proof. The five moves that make a brand believable when polish stopped being a signal.

People are making websites in ten minutes that look like they cost ten thousand dollars. And they’re getting crickets.

That sentence explains more failed launches right now than any funnel critique ever will. Great became normal. Great is the floor. Great is enough to get ignored politely.

Barrier to entry vs barrier to trust

Here’s the pattern underneath it: those two barriers always move opposite. The easier it gets to MAKE something, the harder it gets to be believed. The second AI made pro-level design free, trust got expensive.

So the question changed. It used to be “does my brand look legit?” Now it’s “can my brand prove it’s real?” Anyone can look good. Nobody can fake proof.

The fastest win: a real face

A real face tells the brain this is real before a single word gets read. That’s human wiring, and it doesn’t update with the model releases.

Put your face on the hero. On the about page. On the sales page, the checkout, every automated email. And go one step further… put the ACCOUNTABLE person on the line. People want to know who takes the hit if it goes wrong. On the Bex+Co salon site, every page and follow-up email shows the actual salesperson’s face and name, loaded dynamically. That one move does more than any redesign.

Mark Rober runs face thumbnails across 79 million subscribers because they beat every other style he’s tested. Same wiring.

Build a proof library

Proof scattered across screenshots, DMs, and old emails is proof you don’t have. Build one Notion database that holds every testimonial, number, and story, with columns for type, who said it, photo, source, and consent.

Three kinds of proof go in:

  • Testimonials. What people said, with a face attached.
  • Numbers. Results, before-and-afters, revenue, timeframes. Specifics beat adjectives every time.
  • Stories. What happened, who it happened to, what it meant.

Then point your AI at it. Every sales page, email, and post it writes can pull from real proof instead of inventing filler. Build it once, pull from it forever.

Collect proof on purpose

Mining gets you started. Asking keeps it full.

Ask at peak happiness, right when the win lands. At the end of a happy call: “can we knock out a quick testimonial right now?” Three questions, 60 seconds, record the screen. A small incentive multiplies volume… one past client farmed video testimonials with prize sweepstakes and did millions in course sales on the back of them.

And place proof at the right moment. Numbers on the homepage. Testimonials right before the price. The exact objection-crushing story at the exact objection.

Turn every win into a story

Before, what changed, what they did, the after, and the one lesson. That’s the template. Tell the story free, keep the system paid.

That’s the whole play for the AI slop era. Everyone else is shipping prettier and prettier pages into a feed that stopped caring. You show up with a face, a number, and a story. Trust wins.