What every café and restaurant site needs in 2026 — the eight essentials that fill tables.

Why most restaurant websites are quietly losing tables

Here's how it actually goes. It's a Thursday. Someone's in town, hungry, looking at their phone. They tap "good lunch near me." Google shows them five places. They open three in tabs, scan each one for about 8 seconds, and pick the one that looks least like a chore. The other two get nothing.

That 8 seconds is the whole game. In that window the visitor wants to know three things: what's the food like, are you open right now, and can I get a seat. If your site makes any of those three a treasure hunt — a PDF menu that won't open on iOS, hours that contradict your Google listing, a "Reservations" link in the footer next to the cookie policy — they tap back and pick the next result. You never knew they were there.

The maths is brutal. The average UK restaurant cover is somewhere between £24 and £42. A small café might do £9. Lose six tables a week to website friction and you've quietly given a competitor £40,000+ a year. Lose two Friday-night bookings of four because your booking widget is buried, and that alone is £15k gone in twelve months. Nobody bills you for it. It just doesn't show up.

The good news: nearly every restaurant site I audit fails in the same eight predictable places. The fixes aren't expensive. Most of them are a weekend.

1. A real menu — not a PDF

The single most common mistake. I'd guess 6 in 10 independent restaurant sites I audit still link the menu as a PDF. Sometimes it's a scan of a printed menu, photographed at an angle. Sometimes it's a beautifully designed PDF from the brand agency. Either way, it's a disaster.

Three reasons. First, PDFs are awful on phones — they download instead of opening, they zoom badly, and on a flaky 4G signal they sit at 40% loading while your customer goes to the next tab. Second, you can't update them quickly. Specials changed yesterday? Tough — the printed PDF still shows last week's. Third, and this is the big one: Google cannot meaningfully index a PDF menu. When someone searches "ramen near me", Google's looking at the text on your page. If your dishes are inside a PDF, they don't exist to the search engine. You're invisible to exactly the people most likely to walk in.

Put the menu in real HTML. Section headings, dish names as proper text, descriptions, prices, allergen tags. If your menu changes weekly, build it so a staff member can update it from a phone in 90 seconds. Google will start surfacing your dishes in search results within a fortnight — and that's free traffic you currently don't get.

2. Accurate hours, synced everywhere

Wrong hours are a quiet killer. I've watched a café owner change Sunday opening from 10am to 9am, update his website, and forget about Google Business Profile, Facebook, TripAdvisor, OpenTable, and the third-party delivery apps. For six months he wondered why Sunday mornings were dead. Half the regulars who'd showed up at 9:30 and seen "Closed until 10" on Google had simply switched to the place across the road. They never came back to check.

Three places need to agree, always: your Google Business Profile, your website, and your structured data (the invisible JSON-LD that tells Google your hours in machine-readable form). If you've got special hours — bank holidays, summer extended trading, January reduced hours — they all need updating in all three. Set a recurring calendar reminder for the first Monday of each month to check.

And for the love of god, embed your hours into your homepage as real text. "Open today 8am–4pm" near the top of the page tells Google and the customer the same thing. The "23% of lost customers are wrong-hour visits" figure isn't a marketing slogan — that's what mystery-shopping studies of independent hospitality consistently show.

3. Booking that doesn't suck — and how to choose

About 68% of restaurant bookings happen outside trading hours. If your only "book a table" option is "call us" on a page hidden behind a hamburger menu, you're filtering out two-thirds of your demand. Worse, you're filtering it in favour of the customer who's prepared to leave a voicemail — and that customer is increasingly extinct.

The choice of platform depends on what you are:

• OpenTable — fine if you're a destination restaurant with high cover charges and you want their marketing reach. The flip side is the cover-fee — currently around £1 per seated diner — which adds up fast. Worth it for some, expensive for a 30-cover bistro.
• Resy — increasingly the favoured platform of independent restaurants under chefs and operators. Lower fees than OpenTable, cleaner UX, but smaller marketing reach in the UK.
• SevenRooms — best for groups with multiple sites or upscale operations that want real CRM data on guests. Overkill for a single café.
• In-built (Tock, Eveve, ResDiary, or a widget embedded directly on your site) — best for casual independents and cafés. You keep all guest data, no per-cover fees, and the booking happens on your site rather than a marketplace.

If you're a single-site independent doing under 200 covers a day, I almost always recommend in-built. You own the relationship and you don't pay a marketplace tax on guests who were going to find you anyway. If you're new, opening soon, or struggling for visibility — then OpenTable or Resy's marketing reach can be worth the fee for a year while you build word-of-mouth.

4. Photography that makes people hungry

Food photography is the single highest-leverage spend in restaurant marketing, and most independents under-invest in it by an order of magnitude. The data is consistent: sites with real, high-quality food photography convert visitors to bookings at roughly 2× the rate of sites with stock images, and 3–4× the rate of sites with no images of the food at all.

You don't need a £3,000 shoot. You need a half-day with a local food photographer — £450–£700 in most UK cities — shooting 20 hero plates in natural window light, plus the room, the bar, the front door, and one or two real staff portraits. That's it. That carries you for two years.

If you genuinely can't justify a shoot yet, the second-best option is to ban yourself from stock photography entirely and use your own phone photos consistently. iPhone Pro shots in natural light, served on a real table, edited lightly in the default Photos app. Honest beats slick. The thing that destroys conversion is bad generic food photography — the kind of stock library shot of "burger and chips" that screams "this isn't even my restaurant."

And one more thing: optimise the file size. A 4MB hero plate JPG is the single biggest reason restaurant sites are slow. Compress every image to under 200kb. Use modern formats (WebP, AVIF). Lazy-load anything below the first scroll.

5. The "hero plate above the fold" rule

Open your homepage on your own phone. Don't scroll. What does your customer see?

If the answer is "a giant logo, a beige overlay, and 'Welcome to [restaurant name], established 2014'", you've blown your eight seconds. Hungry people don't care about your founding year. They want to see the food, immediately.

The hero — the bit above the fold on a phone — should contain three things and nothing else: a beautiful photo of one hero dish (not a collage, not a slideshow that's still loading), your name and one-line positioning ("Wood-fired neighbourhood pizza, Cheltenham"), and two buttons: "Book a table" and "See menu". That's it. If you can fit "Open until 10pm" in small text, even better — it answers the open-right-now question without them scrolling.

I've A/B tested this against the "welcome message" homepage on three different restaurant clients. The hero-plate version wins on booking conversion every single time, by margins between 30% and 110%.

6. Instagram pull-through

Your Instagram is doing more marketing work than your website right now. That's not a slight — it's a structural reality of hospitality in 2026. The under-40 crowd discovers restaurants through Reels, saves them to a "places to try" collection, and then checks the website 48 hours later to actually book.

Two things to do. First, embed a live Instagram feed somewhere visible on your homepage — usually mid-page, after the menu teaser. This solves three problems at once: it gives your site fresh-looking content without you writing anything, it shows your most recent specials and events, and it gives Google evidence of an active, current business (which matters for local ranking).

Second — and this is the one most miss — make sure the link in your Instagram bio points to a page on your site that has the booking widget at the top. Not a link tree, not a Linktree, not your homepage with the booking buried six scrolls down. A dedicated "/book" page with the widget visible immediately. Every percentage point of bio-link conversion is direct money.

7. Order-ahead and delivery, front and centre

Even if delivery isn't your main business, you almost certainly have at least one third-party order link — UberEats, Deliveroo, Just Eat, DoorDash, Square Online, or Toast. Treat those links like a primary CTA, because for some of your customers they are the primary CTA.

Specifically: the top of your homepage should have a clear "Order takeaway" or "Order delivery" button alongside "Book a table". On mobile, both should be reachable with one thumb. Don't bury delivery links inside a "Services" dropdown — that's a 50%+ drop-off.

If you take direct online orders (via Square, Toast, or a custom integration) prefer that route — you keep 100% of the margin instead of giving 25–35% to the marketplaces. But have the marketplace links available too. Customers default to the platform they trust, not the one with the best margin for you. Meet them where they are.

8. Mobile speed — why a slow site costs you Friday nights

Around 78% of restaurant-website traffic in 2026 is on a phone, frequently outside on 4G or a struggling 5G signal. If your site takes more than 3 seconds to render the first meaningful content, roughly half your visitors leave before they see the menu. Google's own data on this has been consistent for nearly a decade — every additional second of load time costs you 7–20% of conversions, with the curve getting steeper above 4 seconds.

Open your own site on your phone. Time it from tap to "I can read the menu." If it's over 3 seconds, that's a problem. The usual culprits, in order: uncompressed hero images, a slow third-party booking widget loading synchronously, four tracking scripts fighting each other (Facebook Pixel + Google Tag Manager + a CRM + a chat widget), and a cheap shared host.

The audit-grade fix list: compress every image under 200kb, defer non-critical scripts, host with a modern CDN-backed provider (Vercel, Cloudflare Pages, Netlify, or a decent UK host), and load the booking widget only when it's actually scrolled into view. Done properly, a restaurant site should hit a Lighthouse mobile score above 85. Most independents I audit score in the 30s.

What I'd do this week if I owned a café or restaurant

You don't have a free month. So here's the realistic 80/20 list. Two weekends of effort, one half-day photo shoot, ~£600 in cash. Returns show up in about 60 days as Google re-indexes and your booking conversion lifts.

1. Saturday morning, hour one. Check your hours on Google Business Profile, your site, Facebook, OpenTable/Resy if applicable, and TripAdvisor. Make them all match. Add the next three months of bank holidays.
2. Hour two. Turn your PDF menu into a real HTML page. Section headings, dish names, prices, allergen badges. If your developer can't do this in 90 minutes, get a new developer.
3. Hour three. Add "Book a table" and "Order takeaway" buttons to the top of every page. If you only have one, just that one — but visible.
4. Week one. Book a half-day local food photographer. Twenty dishes, the room, the front door. Around £500–£700.
5. Week two. Rebuild the homepage hero around the best plate from the shoot. One dish, name, one-line positioning, two CTAs. Nothing else.
6. Week three. Compress every image, defer scripts, embed your Instagram feed mid-page. Run a mobile speed test. Aim for sub-3-second load.

Do those six things and your bookings will lift measurably inside a quarter. Not 5% — usually 25–60%, depending on how bad your starting point was. The hard part isn't the work, it's resisting the temptation to "do a full rebrand" as a six-month project that never ships. Ship the boring fixes first. The fancy redesign can wait until you've got the extra Friday-night revenue paying for it.

---

If you want a second opinion on your site, the free audit was built for exactly this. It scores against the eight essentials above, ranks fixes by booking-impact, and gives you a one-page action list. No signup, no sales call unless you ask for one.