Booking conversion: from page view to confirmed

Booking conversion is the share of guests who arrive at your booking page and leave with a confirmed table. It is one of the most useful numbers a restaurant website has, because it measures the moment where interest becomes revenue. A guest who reads your menu, looks at your photographs and opens the booking page has already done most of the work of choosing you. If they leave without booking, something in the last few steps stopped them.

Most venues know how many bookings they take online. Far fewer know how many guests started a booking and did not finish, or where in the flow they gave up. Without that, every change to the booking page is a guess. With it, you can see which step loses the most guests and fix that one first.

This guide breaks booking conversion into the five numbers that make up a booking funnel, from the page view to the confirmation. For each, it explains what the number tells you, what a healthy pattern looks like and what usually goes wrong. It then looks past the confirmation, to the guests who actually arrive, and ends with the one change that tends to lift every step at once.

Booking conversion is calculated simply: confirmed bookings divided by the guests who reached the booking page, over the same period. If 1,000 guests open your booking page in a month and 300 of them confirm a booking, your booking conversion is 30 per cent.

That single figure is a useful summary, but it hides where the losses happen. A guest can leave at any step: before choosing a date, after seeing the times, halfway through the form or at the final screen. The same 30 per cent can come from a flow that loses most guests at the time screen or one that loses them at the form, and the fixes are completely different.

That is why the funnel matters more than the total. The idea of a conversion funnel, a sequence of steps that each lose a share of the people who reach them, comes from marketing, and it fits a booking flow closely. Each step has its own number, and each number points to its own fix.

A typical booking funnel has five steps, and each produces a number worth watching:

Each number divided by the one before it gives the rate for that step. A drop between two steps is booking abandonment at that point. The rest of this guide takes each step in turn.

The first number is how many guests reach the booking at all. It is not a conversion rate in itself, but it sets the size of everything that follows, and it is often lower than it should be because the booking is hard to find.

Compare booking page views with total website visits. If many guests visit your site and few reach the booking page, the button may be hidden, unclear or missing on some pages. A healthy pattern is a steady share of visitors moving from the menu, the home page and the location page to the booking, especially on a phone.

To raise it, put a clear booking button on every page, in the same place, with the same words. Link to the booking directly from your social profiles, your newsletter and your Google listing, rather than to the home page, so guests who are ready to book arrive at the right page first time.

The second number is how many guests who reach the booking choose a date and party size. A low rate here usually means the first screen is getting in the way: it loads slowly, asks something unexpected first, or does not look like a booking at all.

A healthy pattern is that most guests who open the booking make a search. Some will leave immediately, having only wanted to check something, but if a large share leave without searching, look at what they see first. A welcome message, a pop-up, a request to sign in or a choice of venue for a single restaurant can all stop guests before they start.

The fix is usually to make the first screen do one thing: ask for the date and the party size, with sensible defaults such as tonight and two guests. Everything else can wait until the guest has found a time.

The third number is how many guests who search go on to choose a time. This is where booking abandonment is often highest on busy nights, because the time the guest wanted may simply not be available.

A healthy pattern varies with demand: on quiet nights nearly every search finds a time, while on busy nights some will not. What matters is what happens to the guests who do not find their time. If the screen shows only a message saying the time is full, those guests are lost. If it offers the nearest available times, other nights and a place on the waitlist, many of them will stay.

Read this number by day of the week and party size. If larger parties rarely find a time, the problem may be table combinations rather than the booking widget. Our guide to designing a restaurant floor plan that sells covers how to plan tables for larger groups.

The fourth number is how many guests who choose a time complete the form. Here the booking form fields are usually the cause of any drop. Every field takes time, every required field is a reason to stop, and every unexpected question is a moment of doubt.

A healthy pattern is that nearly every guest who chooses a time completes the form, because they have already decided to book. A significant drop here is almost always fixable. Count the fields, remove those you do not use, make the optional ones clearly optional, and never ask the guest to create an account. Show errors beside the field, as the guest types, in plain words.

Verification belongs in this step too. A short code sent to the guest's email makes every booking more reliable, and done well, it costs few guests. Keep the screen open while the guest checks their inbox, and make the code easy to paste. Our guide to the booking widget and the four taps that lose a guest covers the form in detail.

The fifth number is how many guests who complete the form end with a confirmed booking. In a good flow, this is close to all of them. Losses here come from errors, slow final steps, verification codes that never arrive, and screens that leave the guest unsure whether they have booked.

A healthy pattern is that the final step is quick and the outcome is unmistakable: the table is booked, for this many guests, on this date, at this time, with a confirmation on its way. If bookings arrive pending until somebody confirms them, the screen should say so clearly and say when the guest will hear back.

Watch for duplicates as well. A guest who is unsure whether a booking worked may try again, creating two bookings for one party. A rise in duplicate bookings is often a sign that the final screen is not clear enough.

A confirmed booking is not yet a seated guest. Some bookings are cancelled, some are changed, and some guests do not arrive at all. The most useful version of booking conversion follows the guest past the confirmation, to the table.

Look at how many confirmed bookings end up seated, and how many are lost to cancellations and no-shows. Read the figures by source, because guests who book through different channels can behave differently. Our guide to your no-show rate covers how to measure it fairly.

The flow affects this too. A booking made with a verified email address can be reminded and can be changed easily, so plans that change arrive as cancellations rather than empty tables. Our guide to reducing no-shows without annoying guests covers the rest of that toolkit.

The same funnel can behave very differently on a phone and on a laptop. If guests on phones abandon the booking at the form much more often than guests on laptops, the form is hard to use on a small screen. If they abandon at the time screen, the times may be hard to tap or hidden in a drop-down list.

Read the funnel by source as well. Guests who arrive from your newsletter already know you and usually convert well. Guests who arrive from a search engine may be comparing several restaurants and convert less often. A guest arriving from a social post may be browsing rather than ready to book. Comparing like with like stops a change in your traffic from looking like a change in your booking widget.

Finally, read it over time. Booking conversion varies by day of the week, by season and by how busy you are. Compare each week with the same week before, and each change with a period long enough to include busy and quiet nights.

If one change tends to lift every number in the funnel at once, it is booking page speed. Every step in a booking flow loads something: the widget, the available times, the form, the confirmation. Every second of waiting is a chance for the guest to change their mind, check something else or close the page. A faster flow loses fewer guests at every step, not just one.

Booking page speed matters most on phones, where most guests book, often on a mobile connection. A widget that opens instantly on office broadband may take several seconds on a phone in a busy street. Test it there, on a real device, and time each step. Google's guidance on Core Web Vitals is a useful way to think about how quickly a page appears and responds.

Speed comes from small things: loading the widget with the page rather than after the tap, keeping images and scripts in the flow light, and returning available times quickly when the guest changes the date or party size. None of them are visible to guests, which is exactly why they work.

Not every guest who leaves the booking flow is a guest you lost. Some open the booking to check whether a Saturday is free before asking their friends. Some are planning a trip months away. Some are comparing two restaurants and will book the one with the better time. A share of booking abandonment is simply guests doing their research, and no change to the flow will remove it.

That matters when you set expectations. Chasing a booking conversion of nearly 100 per cent leads to changes that make the flow worse, such as pop-ups that try to stop guests leaving or emails sent to guests who never gave permission. The aim is to remove the losses the flow causes, not the ones the guest chooses.

A useful test is to ask whether the guest had what they needed to book. If they found a time, saw a short form and could confirm quickly, and still left, that is their choice. If they left at a dead end, a long form or a slow screen, that is yours to fix.

A sudden fall in booking conversion is usually a fault rather than a change in demand. When the numbers drop sharply from one week to the next, check the obvious things first. Has the website been updated, and does the booking widget still load on every page? Have the booking rules changed, so fewer times are offered? Is the booking page slower than it was?

Check verification too. If guests must confirm their email with a code and the codes stop arriving, or land in spam folders, the confirmed number will fall while the earlier steps stay steady. Our guide to email deliverability covers why mail goes missing and how to fix it.

Finally, book a table yourself, on a phone, from start to finish. Most sudden falls become obvious within a minute of trying the flow as a guest would.

You need two sources of data. Your website analytics can show how many guests reached the booking page and, if your booking widget reports its steps, how many moved through each one. Your booking system shows how many bookings were confirmed, from which source, and what happened to them afterwards.

Set up a simple weekly view with the five numbers and the step rates between them. Keep it in the same format every week, so changes stand out. When a number moves, look at what changed that week: the website, the booking rules, the season, or your marketing.

Avoid over-reading small numbers. A small venue may have only a few hundred booking page views a week, so a single busy evening can move a rate noticeably. Look at four-week periods for decisions, and at single weeks only for warning signs.

Booking conversion improves fastest when changes are made one at a time and measured. Change the first screen, watch the search rate for a few weeks, and keep the change only if it helped. Then move on to the next step.

Changing several things at once makes it impossible to tell which helped and which hurt. A new booking widget, a redesigned website and a new marketing campaign launched in the same week will produce numbers that cannot be explained.

Write down every change and its date, alongside the weekly numbers. Over a year, that record becomes the most useful document you have about your booking flow, and it stops the same failed idea being tried twice.

Here is an illustrative month for a venue's booking funnel, before and after two changes. The figures show the method, not benchmarks.

The faster page lifted every step a little; the alternatives on the time screen lifted one step a lot. Together, they brought 370 more confirmed bookings from the same number of visitors.

In servearound, guests book on a page in your colours, on its own link or embedded in your website, with the time held while they finish, a six-digit email code before the booking is confirmed and a link to change or cancel in every confirmation. When a time is full, the booking page offers the waitlist rather than a dead end.

Analytics shows bookings by source, including the booking widget, the embedded widget, the phone and walk-ins, with covers, no-show and cancellation rates and how much notice each source gives. It also follows bookings from booked to confirmed, seated and completed, showing how many were lost to cancellations and no-shows at each step. Page views come from your own website analytics. See Analytics and the reservations module.

It depends on your venue, your traffic and how busy you are. The most useful comparison is your own booking page over time, after each change you make.

Some are checking availability, comparing restaurants or planning far ahead. The share that leaves because of the flow is the part worth fixing, usually at the time screen or the form.

Not in the online funnel, but track them alongside it. If online booking conversion improves, phone bookings during service often fall.

Usually four weeks, so the period includes busy and quiet nights. A single weekend can mislead.

Not if it is easy to skip. Guests who care will use it; guests who do not should move on in one tap.

Every second of waiting gives guests a chance to leave, at every step. On a phone, the effect is larger than it looks on a laptop.

Booking conversion is the share of guests who reach your booking page and leave with a confirmed table. Break it into five numbers, page views, searches, times chosen, details completed and confirmed, and read the drop between each pair to find where booking abandonment happens. Then follow confirmed bookings through to the table, by source.

Fix the steps one at a time: make the booking easy to find, keep the first screen simple, offer alternatives when a time is full, cut the booking form fields to the ones you use and make the final screen unmistakable. Above all, improve booking page speed, because a faster booking widget loses fewer guests at every step.