Why Scrolling 50 Google Results To Pick A Restaurant Is Broken

Last week I spent 22 minutes picking a ramen spot in Dubai Marina. I opened Google Maps, scrolled past 14 places, switched to TripAdvisor, cross-checked three of them on Reddit, then opened Instagram to see if the photos matched reality. By the time I picked one, I wasn't hungry anymore. I was annoyed.

This is not a me problem. This is the product.

The platforms don't want you to decide

Google, Yelp, and TripAdvisor all run on the same business model: more listings, more sessions, more ad slots. If you found the right restaurant in 30 seconds, they'd lose money. So the interface is built around browsing, not deciding. You get 50 results. You get filters that don't actually narrow anything. You get sponsored slots that look like organic ones. You get a 4.3 next to a 4.4 next to a 4.5 and no way to tell which one is actually good.

The review counts make it worse. A place with 8,000 reviews at 4.2 stars and a place with 200 reviews at 4.6 stars — which is better? Nobody knows. The platforms won't tell you because the answer is "it depends," and "it depends" keeps you scrolling.

Reviews stopped meaning what they used to mean

The 5-star scale is broken at both ends. Restaurants buy reviews. Competitors leave fake 1-stars. Tourists rate based on whether the staff spoke English. Locals rate based on whether the bathroom was clean. A 4.5 in one city is a 3.8 in another. Half the reviews are people venting about delivery apps the restaurant doesn't even run.

And the review you actually want — "is this place good for a Tuesday dinner with someone I don't know well" — is buried under 400 reviews about parking.

"Best of" lists are the same problem with extra steps

So people give up on platforms and Google "best ramen Dubai 2024." Now you get 30-item listicles written by SEO farms that have never been to any of the restaurants. Or you get a Reddit thread from 2019 with 4 contradicting opinions. Or a TikTok where the criteria for "best" was that the noodles were yellow.

The list format is the bug. Any list longer than three items just moves the decision back to you. You wanted help. You got homework.

What I want instead

I want one answer. Not ten. Not three. One.

I want someone to have done the comparing already and to tell me, with a straight face, "go here." If they're wrong sometimes, fine. They'll be wrong less often than I am after 22 minutes of scrolling.

That's the bet behind Zerch. It picks one winner per category and shows you that. No infinite scroll. No 50 results sorted by some opaque algorithm. No sponsored slot dressed up as a recommendation. One place, one pick, decision done.

It won't be the right call for everyone. If you love research, if picking the restaurant is half the fun, if you want to read 30 reviews before committing — Zerch isn't for you. Keep your Google tabs open.

But if you've ever closed your laptop hungrier and grumpier than when you opened it, the problem isn't you. The problem is that every platform you tried makes more money the longer you take. A tool that just answers the question is a different kind of product. We'll see if people actually want one.