← Journal6 min read
Design10 May 2026

Your restaurant website has 3 seconds. Here's what we'd fix first.

After building restaurant sites in Helsinki, we've watched the same five mistakes kill conversion rates over and over. This is the priority list.

A potential diner finds your restaurant on Google. They tap your website link. They get 3 seconds to decide whether to book a table, place an order, or close the tab and try the next restaurant. Three seconds.

We've built restaurant websites for clients in Helsinki and watched analytics tell the same story every time. The same handful of mistakes kill conversion. The fixes are simple. The order matters.

Here's the priority list. Fix from the top. Skip the bottom until you've nailed the top.

1. Load time under 2 seconds (non-negotiable)

Most restaurant websites we audit load in 6 to 12 seconds on a phone over 4G. The hungry diner has already left. They didn't bounce because of bad design or weak menu photography. They bounced because the spinning loader exceeded their patience.

Speed is the foundation. Everything else is decoration on top of speed. If your site loads in 8 seconds on a phone, every other improvement in this article is wasted effort until you fix this one.

Quick check: open your site on your phone (not WiFi, use mobile data) and time it. If it takes more than 3 seconds to be readable, you have a speed problem that's quietly killing bookings.

2. Show food. Hide everything else.

The single biggest mistake we see: restaurant homepages that lead with chef stories, brand philosophy, or aesthetic photography of empty tables. Diners don't care. They want to see food.

The order people care about, in our experience: (1) photos of actual dishes, (2) menu with prices, (3) opening hours and location, (4) ability to book or order, (5) maybe the chef's story if you scroll.

Most restaurant sites have this reversed. They lead with mood photography and bury the menu. Move food up.

3. Make the menu actually readable

Half the restaurants we audit either have no online menu, a PDF menu that takes 15 seconds to download on a phone, or a menu image so small you have to pinch-zoom to read it. Each of these costs you bookings.

Your menu should be: HTML text (not a PDF, not an image), searchable by Google, readable on a phone without zooming, with photos for at least your signature dishes. Plus allergen indicators and dietary tags (vegetarian, vegan, gluten-free) because diners check for these.

4. One clear primary action

Every restaurant site should have one primary action that's impossible to miss. For most: "Book a table." For takeaway-focused places: "Order online." For lunch-buffet places: "See today's buffet."

When we audit restaurant sites, we count how many "call to action" buttons sit in the visible header area. Often it's three or four (Book, Order, Contact, Reserve, Menu, About). That's three or four too many. One primary, one secondary, the rest hidden in a menu.

5. Practical info above the fold

Address. Hours today. Phone number. These three pieces of information should be visible on the homepage without scrolling. Don't hide them under a "Contact" page. Don't put them in a tiny footer. Diners check these in 5 seconds and decide whether to visit.

Bonus: if you're closing in 30 minutes, show that fact. "Open until 22:00 (last orders 21:30)" is information that converts visits.

What to ignore (for now)

Things that don't matter at the start:

  • Newsletter signup forms (nobody is signing up for restaurant newsletters)
  • Social media feeds embedded on the homepage (they slow the site down)
  • Animated entrance effects, cursor trails, parallax scrolling (they delay loading)
  • Photo of the empty dining room at night (atmospheric but converts nothing)
  • Chef philosophy paragraph (only matters once people are already coming)

The simple test

After we fix a restaurant site, we run a 5-second test on a friend who's never seen it. We show them the homepage for 5 seconds on a phone, then close it and ask: "What did this restaurant serve? Where are they? Could you have booked a table?"

If they can answer all three, the site is doing its job. If they can't, we go back and fix it. Try it on your own site right now.


If your restaurant website fails the 5-second test, we can probably help. We rebuild small restaurant sites in 1 to 3 weeks. Fixed price, fast turnaround, no agency overhead. Get in touch if it sounds useful.

★ Work with us

Like this thinking?
Let’s build something together.

Book a free call →

★ More from the Journal

All articles →