Online ordering conversion: fixing the checkout

Every venue that takes orders online loses most of the guests who start. They open the menu, look around, perhaps add a dish or two, and then leave without ordering. Some were only browsing. Many were hungry, intended to order and gave up somewhere between the first tap and the confirmation, because something in the way made it too slow, too confusing or too expensive.

Online ordering conversion is the share of guests who start an order and finish it. Small improvements to it are worth a great deal, because every guest who gets to the end is an order the venue has already paid to attract. This guide walks through the five steps between a hungry guest and a confirmed order, where guests give up at each, and the fixes that keep them going. Every figure in it is an example.

Conversion is a simple ratio: of the guests who opened the menu, how many placed an order? It can be measured for the whole journey or step by step, and the step-by-step view is far more useful, because it shows where guests leave.

A venue might find, for example, that of every hundred guests who open its online menu, sixty open a category, forty open a dish, twenty-five add something to the basket, fifteen start the checkout and ten complete an order. The overall conversion is ten in a hundred. The steps show that the biggest losses come between opening a dish and adding it, and between starting the checkout and finishing it.

The same thinking applies to bookings, where the equivalent problem is booking abandonment: guests who start a booking and leave before confirming. Our guide to booking abandonment and where guests give up covers that journey; this guide covers ordering.

Most online ordering journeys pass through five steps:

Each step has its own reasons for guests to leave. Fixing conversion means finding the step that loses the most guests and removing what stands in their way. The discipline of doing this systematically is often called conversion rate optimisation, and its principles work as well for a menu as for any online shop.

The first step loses guests before they see a dish. The commonest cause is speed: a menu that takes several seconds to appear on a phone loses guests who were never very patient. The second is a poor first screen: a large banner image, a pop-up asking them to sign in or download an app, or a page that does not make clear what to do next.

Fix speed first. Keep images small enough to load quickly, avoid heavy scripts, and test the menu on a phone over an ordinary mobile connection, not the office wifi. Then make the first screen useful: the venue's name, whether it is open for orders, and the menu itself, or a clear way into it.

Avoid anything that stands between the guest and the food. An app download request on arrival loses a large share of guests who only wanted to order dinner.

Once the menu has loaded, guests need to find what they want quickly. Long menus, unclear category names and dishes buried at the bottom of long lists all slow them down, and slow browsing leads to giving up.

Organise the menu the way guests think: clear categories with familiar names, the most popular dishes near the top, and a short way to jump between categories. A search box helps guests who know what they want. Filters for dietary needs help guests who need to rule dishes out.

Watch what guests search for. Searches that find nothing are especially useful: a guest searching for "vegan" or "kids" or a dish you do not serve is telling you what the menu is missing, or what it is failing to show clearly.

Choosing a dish is where many guests hesitate. They open a dish, read the description, look at the photograph, consider the options, and then close it again without adding it. A large gap between dishes opened and dishes added usually points to something on the dish itself.

Look at the description, the photograph and the price. A description that does not say what the dish is, a missing or poor photograph, or a price that seems high compared with similar dishes can each stop the add. Options can stop it too: a dish with too many required choices, or choices that are confusing, makes adding it feel like work.

Keep required choices to those the kitchen genuinely needs, such as a cooking point, and make optional extras clearly optional. Show the price of each option as the guest chooses it, so the total is never a surprise.

The checkout is where many orders are lost, and often the most fixable part of the journey. Guests who have already chosen their food leave because the checkout asks too much of them: too many fields, a forced account, unclear delivery or collection times, or fees that appear only at the end.

Ask only for what the order needs. A name, a phone number and an email, an address for delivery, a time for collection: each extra field loses some guests. Let guests order without creating an account, and offer the option to save their details afterwards rather than before.

Show the full cost early. A guest who sees delivery fees, service fees or a minimum order only on the last screen often leaves, and does not come back. Fees shown clearly from the start, even if they are the same, lose far fewer guests.

The last step is the guest's confirmation that the order has gone through. If payment is taken online, this is where payment failures, slow payment screens and unfamiliar payment forms lose guests at the very end.

Offer the payment methods guests expect, keep the payment screen fast and familiar, and make errors clear and fixable: a card declined message that explains what to do next keeps the guest, while a blank error loses them. After payment, show a clear confirmation with the order details, the expected time and a way to contact the venue.

Send the confirmation by email as well, so the guest has a record. A guest who is unsure whether their order went through often orders again elsewhere, or calls the venue in a panic.

Direct online ordering, through the venue's own website rather than a delivery platform, has a particular reason to care about conversion. The venue has usually worked harder to bring the guest there, through its own marketing, its regulars and its social channels, and every guest it loses at the checkout is one it paid to attract and will not get back through the platform.

Direct ordering also gives the venue control over every step of the journey, which platforms do not. The venue can shorten the checkout, change the menu layout, test new descriptions and see exactly where guests leave. Our guide to direct online ordering and keeping the whole margin covers the wider case.

That control is only useful if the venue uses it. A direct ordering page that is slower and clumsier than the platforms will lose guests to them.

Most guests who order on a venue's website do so on a phone, often with one thumb, often while doing something else. Every step of the journey should be designed for that.

Make buttons large and easy to reach. Keep text readable without zooming. Avoid layouts that need horizontal scrolling or tiny tap targets. Make the basket visible at all times, so guests can see what they have chosen. Our guide to mobile booking: thumbs, not mice covers the same principles for booking, and nearly all of them apply to ordering.

Test it yourself, on your own phone and on an older one, on a mobile connection. Order a meal from start to finish as a guest would, and note every moment of hesitation.

Online upsells, suggestions of extras, sides or drinks during the order, can lift the average order value. They can also lose orders, if they interrupt the journey with pop-ups or make the guest work to decline them.

The best online upsells appear at the natural moment: options shown when a dish is chosen, a side suggested with a main, a drink or dessert suggested in the basket. They should be easy to accept and easy to ignore, with no extra screen to dismiss.

Watch whether upsells help or hurt. If adding a suggestion step lowers the share of guests who finish their order, the extra spend from those who accept may not make up for the orders lost. Our guide to upselling that does not sound like selling covers the principles in person; online, they matter even more.

Guests hesitate to order from a venue they are not sure about. At the checkout, a few signals reassure them: a clear name and address for the venue, a phone number, clear prices, a secure payment screen, and an honest estimate of when the food will arrive.

Allergen information is a trust signal too. Guests with dietary needs want to know that the venue takes them seriously, and a menu that shows allergens clearly and invites guests to mention allergies gives them confidence to order.

Remove anything that looks careless: spelling mistakes, broken images, dishes that are listed but unavailable. Each small sign of neglect makes guests wonder about the food.

To improve conversion, measure it step by step. Many ordering systems record how many guests open the menu, view dishes, add items, start the checkout and complete an order. If yours does, look at the numbers each week; if it does not, ask your provider what it can show.

Look for the step with the biggest fall. That is where to start. A large fall between opening a dish and adding it points to descriptions, photographs, prices or options. A large fall at the checkout points to fields, fees or accounts. A large fall at payment points to the payment screen.

Look at the time guests spend too, where you can. A step that takes guests a long time is a step that is confusing them.

Change one thing at a time and measure the result over a fair period. Remove a checkout field, show fees earlier, add photographs to the most popular dishes, shorten the menu: each change should be measured before the next.

Compare similar periods: weekday with weekday, a normal week with a normal week. A change that seems to work in a week of good weather may simply have coincided with more hungry guests.

Keep a record of what you changed and what happened. Over a few months, the record shows which changes moved conversion and which made no difference, which is far more useful than general advice.

Most checkout problems fall into a few patterns, each with a straightforward fix:

Fixing even one or two of these often lifts conversion noticeably, because each one loses guests at the point where they had already decided to order.

Not every guest who leaves is lost. Some leave to check with someone else, to look at another menu or simply because they were interrupted, and some of them return. Make it easy for them.

Keep the basket when a guest leaves and returns on the same device, so they do not have to start again. Keep the menu's address simple and memorable, so a guest who closes it can find it again. Guests who have given their details and agreed to hear from you can sometimes be reminded, with their consent and within the rules, but never chased.

A guest who returns and finds their basket waiting is far more likely to finish than one who has to rebuild their order.

Guests often open an online menu outside the hours it can take orders: before opening, after the kitchen closes, or on a day the venue is shut. A menu that lets them build a basket and then refuses the order at the checkout loses them at the worst possible moment.

Show the ordering hours clearly at the top of the menu, and say when the next orders can be taken. If guests can order ahead for a later time, make that obvious; if they cannot, say so before they start. Keep the hours in step with the kitchen's real hours, including holidays and early closures, so a guest is never told the kitchen is open when it is not.

Groups order differently. One person often collects everyone's choices, and the order grows long and complicated, with options and notes for each guest. A checkout designed for one person ordering one meal can struggle.

Make it easy to review a long basket: each item clearly listed with its options, easy to change or remove, with a running total. Let guests add a note for the kitchen where the system allows, and keep the basket safe if the person ordering leaves the page to check with someone. A group order that has to be started again is often abandoned.

An invented pizzeria takes orders on its own website. Of every hundred guests who open the menu, about eight complete an order. Step-by-step figures show the biggest falls at the dish page and at the checkout.

The pizzeria makes four changes over two months. It adds photographs to its ten best-selling pizzas and rewrites their descriptions. It lets guests order without an account. It shows the delivery fee and minimum order on the menu, not just at the end. And it reduces the checkout to name, phone, email and address.

After the changes, about thirteen in every hundred guests complete an order. The falls at the dish page and the checkout have both narrowed, and the average order value is unchanged. With the same number of visitors, the pizzeria takes over half as many orders again.

servearound does not take online orders for delivery or takeaway today, so ordering on your website for those runs through whatever ordering service you use. Where guests order at the table from their phones, by scanning the table's QR code, servearound records the same journey this guide describes: how far each visit gets, from opening the menu to opening a category, opening a dish, adding a dish and starting the checkout, recorded without identifying the guest.

The menu report also shows each dish's views against its adds and orders, how far down the list guests look, how far down the page they scroll, which categories are never opened and what guests searched for that matched no dish. At the table, guests confirm they have reviewed the allergen and dietary information before an order is sent, and orders reach the kitchen screen like any other. See the menus module and Analytics.

The share of guests who open an online menu and go on to complete an order, best measured step by step to show where guests leave.

Arriving at the menu, browsing, choosing and adding dishes, checking out, and confirming, with payment if it is taken online.

Often at the dish page, where descriptions, photos, prices or options put them off, and at the checkout, where fields, accounts and late fees do.

No. Let them order as guests and offer to save their details afterwards. Forced accounts lose many orders.

Ask only for what the order needs, show all fees and minimums from the start, show the expected time, and keep payment fast and familiar.

They can, if they appear at the natural moment and are easy to ignore. Pop-ups that interrupt the journey often cost more orders than they add.

It is the same problem in a different journey: guests who start a booking and leave before confirming. The same step-by-step approach applies.

Count guests at each step, from opening the menu to completing the order, and look for the step with the biggest fall.

Change one thing at a time and compare conversion over similar periods before and after.

Online ordering conversion is lost step by step: guests leave when the menu is slow, when dishes are hard to find, when descriptions, photos, prices or options put them off, and above all at a checkout that asks too much or reveals fees too late. Measure each of the five steps, arriving, browsing, choosing, checking out and confirming, and start with the one that loses the most guests.

Keep the menu fast and organised the way guests think, the dishes clear and photographed, the options few and priced as they are chosen, and the checkout short, with no forced account and every fee visible from the start. Place online upsells at natural moments, make the confirmation clear, test one change at a time, and treat the guests who almost ordered as the easiest guests you will ever win.