We’ve audited a lot of UK restaurant websites over the last year. The independents in Soho, the small chains across Manchester and Bristol, the family-run places in Bath, the gastropubs outside London. And the same problems come up almost every time, in roughly the same order.
None of these problems are about the food. The food is usually great. The website is just losing the booking before the food gets a chance.
Problem 1: The menu PDF
A PDF menu is the single biggest revenue leak on most restaurant websites. The PDF takes 4 to 11 seconds to load on a phone, opens in a separate viewer that hides your branding, can’t be searched, can’t be styled, and is almost always out of date because the chef changed the specials at lunch and nobody re-exported the file.
The fix: render the menu as actual HTML on the page. Sections like Starters / Mains / Desserts. Real prices in pounds. Dietary tags (V, VG, GF) inline. A hidden field on each item for whether it’s currently 86’d so front-of-house can flip a switch.
Suddenly the menu loads instantly, looks like part of the brand, and Google can read every item. That last part is huge: search for "wood-fired pizza Hackney" and the restaurants whose menu items are crawlable text rank above the ones whose menus are locked inside PDFs.
Problem 2: Booking lock-in
OpenTable, Resy, SevenRooms, DesignMyNight. Useful tools, but each one of them takes a cut of every cover and, more importantly, owns the customer relationship. The diner books on OpenTable, OpenTable owns their email, OpenTable can promote the place down the road to that diner next month.
A custom website lets you take direct bookings. Two routes work: own-the-stack (a custom booking flow that writes to your reservation system, e.g. ResDiary or Tock, via API) or own-the-customer (a deposit-protected enquiry form for groups, with the smaller individual bookings staying on the platform).
A small group of London restaurants we work with shifted roughly 35% of their bookings off OpenTable in three months by adding a faster, branded direct-booking path on their own site. They still list on OpenTable for new-customer discovery, but the email list grows, the cover fees go down, and the regulars come straight to the source.
Problem 3: Photography that doesn’t do the food justice
Restaurant decisions are visual, made on a phone screen, in low ambient light, with a hungry brain. The site that looks appetising in those conditions wins disproportionately.
Most restaurant sites we see use either: stock food photos that nobody believes, phone photos lit by the kitchen pass, or a single nice shoot from 2019 that’s now seasonally wrong. None of these convert.
A real food shoot, twice a year, with a half-day photographer in your actual restaurant, is one of the highest-ROI marketing spends a hospitality business can make. The photos earn back the invoice on the first weekend. They power the website, the Instagram grid, the printed menu, the third-party listings, and the press kit.
Problem 4: The page is a desktop site shoved onto a phone
Hospitality has the most mobile-skewed traffic of any sector we work in. Roughly 80% of restaurant-website visitors are on a phone, often standing on the street looking at the place. The website still has a tiny logo, hover-only menus, and a "view our gallery" carousel that requires three taps.
A custom-built mobile-first restaurant site looks more like an app than a brochure: tap-to-call in the header, tap-to-direct straight into Apple/Google Maps, menu as the first thing you see scrolling, book a table as a sticky button on the bottom edge of the screen. All visible at the moment of decision.
Problem 5: Local SEO that doesn’t exist
"Sunday roast Hackney", "best pizza Manchester city centre", "tapas Bath". These are the searches that move covers. Most restaurant sites do nothing to win them.
The list, in order:
- Google Business Profile claimed, with at least 30 photos, fully populated hours, menu uploaded, and a steady drip of fresh reviews from regulars.
- Restaurant schema on the site (servesCuisine, priceRange, acceptsReservations, openingHoursSpecification). Without this, Google has to guess. With it, you get rich-result placements.
- Town and dish in page titles: "Sunday Roast in Hackney · Lamb · [Restaurant]", not "Welcome · [Restaurant]".
- One page per service: lunch, dinner, private dining, Sunday roast, set menu. Each ranks separately for a different search.
- Tripadvisor / SquareMeal / Yelp listings linked back to your real domain, with consistent NAP details.
Done together this is roughly two days of work for a competent studio. The compounding lift on covers is enormous because none of your competitors are doing it.
Problem 6: No way to capture the customer who almost booked
Most diners look at a restaurant website at least twice before they book. The first visit is research; they leave to compare. Most sites do nothing on that first visit to bring the diner back.
Two small additions move this hard. A well-designed newsletter signup with a real reason to subscribe (Sunday roast menu changes, supper-club drops, set-menu first looks). And an Instagram embed that’s the actual current grid, not a static placeholder. The diner who almost booked, scrolls Instagram on the train home, sees your post, and books on Tuesday.
Problem 7: Speed
A 6-second mobile load loses you 50% of the visitors before they see the menu. Restaurant sites are particularly bad for this because they’re full of unoptimised hero videos and 4MB photos straight from a DSLR.
A custom-coded restaurant site, with properly compressed imagery, lazy-loaded video, and no template bloat, lands in the 1 to 2 second range. That alone lifts conversion by double-digit percentages, in our own client data.
What a custom-built UK restaurant website looks like
Pulling all of this together, a 2026 hospitality site looks like: a homepage that loads in under two seconds with the food and the booking button visible immediately. A real HTML menu with dietary tags, refreshed weekly. Direct booking on your own site, with deposit protection for larger groups. Restaurant schema and proper local SEO so the site ranks for what people actually search. Real photography. A short, useful newsletter. And an Instagram feed that proves the food is still being made tonight.
None of that is exotic engineering. It’s just outside the lane of any off-the-shelf restaurant theme.
What to do next
If you run a restaurant in the UK and want a real assessment of your current site, we run a free 30-minute audit. We pull your site up, run the speed test, walk through the booking journey on a phone, and tell you the three things that would move bookings most. No pitch, no follow-up if you don’t want one.
Book the audit here. Or read the rest of the Blog.