Thirteen restaurant sites worth studying

Beautiful is not the same as effective

Ask ten restaurant owners what makes a good website and nine will start with looks. The photos, the typeface, the colour of the button. It is the most enjoyable part. It is not the part that decides whether an order comes in tonight.

The sites that perform are rarely the prettiest. They are the clearest. Within a few seconds they answer three questions: what do you sell, where are you, and how do I order. Everything after that is decoration, and decoration that gets in the way costs you money.

This is not a matter of taste. Someone landing on your site is usually standing outside with a phone in one hand and an appetite. They do not read, they scan. If they cannot find the menu or the order button straight away, they go back to Google or to the app they know works. That is exactly the app you pay commission to.

What follows is what the sites held up as examples have in common, and what you can borrow from them without rebuilding anything.

What the best restaurant sites have in common

The roundups of the best restaurant websites are almost all American. The businesses on them are wildly different: a vegan burger bar, a taqueria with three locations, a diner in a small city, a Thai restaurant that lives on reservations. Their styles have nothing to do with each other. The decisions underneath them do.

That the examples are American changes little about those decisions. The names of the delivery apps differ, the commission does not. Here too your guest is searching on a phone, looks at the menu first and leaves the moment they hesitate. What does need translating is the detail: iDEAL rather than an American card payment, prices including VAT rather than tax added at the till, and reviews on Google, TripAdvisor and Iens rather than on an American review site.

Taken one at a time these patterns are dull. None of them is a trick. Together they are the difference between a site that is finished and a site that produces something. And every one of them is reproducible in a twenty-seat restaurant.

The first second decides

The first screen is the only part of your site you know everyone sees. Yet on many restaurant sites it holds a large atmospheric photo with the name across it and nothing else. A stranger still does not know whether you do takeaway, whether you deliver, or whether you only open in the evening.

On the sites that work, that first screen carries everything a stranger needs in order to keep going.

After that you keep the button within reach. A bar that stays put while the page scrolls, carrying the menu, the order button and your phone number, costs nothing and pulls guests towards the till from any page. On a phone that is often the difference between scrolling and ordering.

Run the test on your own handset, on mobile data rather than the restaurant wifi. Can you tell within three seconds what you sell and where to tap? If not, you have found your first job.

Character is fine, as long as it blocks nothing

The most enjoyable sites in those roundups carry their character down to the details. One builds its entire site around horror films, with matching copy and a dark palette. Another names every burger after a metal band, which turns the menu itself into something people talk about. A third has a line about eating with your hands and drinking with your friends, and you know at once what an evening there looks like.

What they share is that the character never blocks the path. The joke lives in the words and the photography, not in the navigation. The order button is still the clearest button on the page.

That is the line. A menu that only appears when you hover over something, an intro animation, a typeface you have to decipher on a phone: charming in a presentation, expensive in practice. Character lives in what you say and show, not in how hard your site is to operate.

This is also the one part you cannot copy from anyone else, and you do not need to. The wood oven, your mother's sauce, the band that plays every Saturday. Put that on the page in plain sentences and you have something no template can give you.

The menu is the page where it happens

Most visitors come for one thing: the menu. On many sites it is the weakest page. A PDF from some earlier year, a photograph of a printed card, or a list with no prices on it.

A menu as a PDF costs you money in three ways. Google reads it poorly, so your dishes never surface in search results. Your guest has to pinch and drag on a small screen. And you have to export and upload a new file for every price change, so you do it less often than you should.

• Put the menu on the page as ordinary text, one line and one price per dish.
• Use photos of your own dishes. A sharp phone photo in daylight beats any stock library.
• Write descriptions that tell you something: the sourcing, the method, the one ingredient you are proud of.
• Build categories a stranger understands, not the layout of your kitchen.
• Put a most-ordered block at the top. What other people chose is easier for an undecided guest to choose.
• Show prices including VAT, 9 per cent on food and 21 per cent on most drinks, so nothing is added at checkout.

Give your signature dishes their own pages

The places that stand out in search results give their signature dishes a page each. Not a line on the menu, but a page with photos, the story behind it and an order button. American research into click behaviour in Google shows the top result takes roughly ten times as many clicks as the tenth. For a search along the lines of order pizza in your own neighbourhood, that is the difference between existing and not existing.

Start with two dishes: the two you sell most often and the ones people travel for.

Checkout without friction

Plenty of owners assume the sale is won the moment something lands in the basket. That is exactly where it breaks. Every extra field, every forced account and every surprise in the final total is a reason to give up. And someone who abandons at checkout does not come back five minutes later to order from you. They order from somebody else.

That last one matters more than it looks. A guest told forty minutes who eats after thirty-five comes back. A guest told twenty who waits forty orders through the app that was honest with them next time.

A second rule of thumb: do not ask a guest to fill anything in until they know what they are paying. A delivery charge, a minimum spend or a surcharge that only appears at the last step is the most expensive surprise you can build. Show them next to the basket, not behind it.

Walk your own checkout end to end once, on your own phone and your own bank account. Count the fields. Anything that is not needed to get food on a table can go.

Proof: stars, awards and honesty about commission

People believe other guests sooner than they believe you. That is not annoying, it is useful. You almost certainly have reviews already, on Google, TripAdvisor or Iens. They are simply in the wrong place: where someone is still comparing, rather than where they can order.

• Put your star rating and two or three short quotes on the homepage, the menu and the ordering page.
• Ask for a review at the moment things went well: at checkout, or just after the delivery.
• Route unhappy guests to you first and to the internet second. A problem you solve rarely becomes a review.
• Show press mentions and awards if you have them. One line is enough.

Just tell people that ordering direct is cheaper

The pattern that recurs most often on the best sites is also the simplest. They tell the guest outright that ordering through the restaurant's own site costs less than ordering through a delivery app, and why. No complaints about Thuisbezorgd, Uber Eats or Deliveroo, just an explanation in two sentences.

The sums work on both sides. An order through a platform costs you a percentage of the bill. An order on your own site does not. You can hand that difference back to the guest as a keener price, or use it to repair your margin. Explain it and you get both: trust, and a cheaper order.

Do read your platform contract before you start moving prices around. Saying that you have your own ordering channel and that an order there costs you less is always allowed.

One main goal: ordering or booking

Most restaurant sites line everything up at the same size: ordering, booking, gift cards, catering, the newsletter. The result is that visitors pick whatever is easiest, and that is usually leaving.

The sites that work do choose. A takeaway pushes everything towards ordering. A restaurant whose dining room is the product pushes everything towards booking and deliberately makes the order button smaller. That is not offering less. That is deciding what you want to see happening on a Monday evening.

More than one location

If you run several sites, give each location its own page with the address, opening hours, directions and its own order button. On the homepage put only the choice between them, not every address mixed together. The homepage stays calm, and each location gets a page that can compete in the search results for its own neighbourhood.

What to do on Monday

You do not need a rebuild to make your site sell better. Work through these seven points on your own phone and write down what is wrong.

1. Open your site on mobile data and count the seconds before anything appears.
2. Check whether you can tell within three seconds what you sell and where you are.
3. Count the taps from the homepage to a dish in your basket.
4. Confirm the menu is text rather than a PDF, and that every price is right.
5. Look for the reviews on your site. If they are missing, put three of them up.
6. Pay for one real order with your own phone and your own bank account.
7. Add one sentence to the homepage explaining why ordering direct is better.

What you find is usually not a design problem. It is a PDF, a button sitting too far down the page and a checkout asking for four fields too many. Three evenings of work, and your site does something it was not doing last week.