No, Don't Quit Deliveroo Tomorrow

We build direct ordering systems. We'd quite like you to buy one. So it's worth saying plainly, up front: switching the delivery apps off is usually a bad idea, and anyone telling you otherwise is selling harder than they're thinking.

The right move is almost always to run both. Here's why, and how to work out the split.

What the apps are actually good at

One thing, and they're very good at it: putting your name in front of people who have never heard of you.

Someone moves into the area. Someone's had a bad night and can't be bothered to think. Someone fancies Thai and doesn't know which Thai. They open an app, and there you are, next to your competitors, with your rating and your photos.

You cannot buy that reach cheaply any other way. Not with leaflets, not with Facebook ads, not with a website nobody's searching for by name. The marketplaces have spent enormous amounts of money to own the moment when a hungry person doesn't know where to order from, and when you're on them you rent a slice of it.

That is a real service and it's worth paying a real fee for.

What they're bad value for

Everybody who already knows you.

The eleventh order from the same person is not discovery. Nothing is being discovered. You're paying a 25–35% finder's fee to be found by someone who found you two years ago and has your menu memorised.

That's the distinction that matters, and it's not "apps bad". It's: an acquisition cost makes sense once per customer, not once per order.

Why switching off entirely usually backfires

Four things tend to happen to people who delete the apps in a fit of righteousness.

The volume doesn't follow you. People are lazy — pleasantly, ordinarily lazy. A meaningful chunk of your app orders will simply order from someone else on the app rather than go and find your website. Plan for keeping maybe half of them in the first few months, and be pleased if it's more.

Your new-customer tap turns off. However good your regulars are, some of them move away, some get bored, some stop eating fried food in January. If nothing is replacing them you're on a slow decline that takes six months to become visible.

The listing itself was advertising. Even people who order direct often found you on an app, or checked your rating there before ringing. Being absent isn't neutral.

You lose the quiet-Tuesday cover. Marketplace orders are often marginal on paper but they fill dead time. An order at 30% commission on a Tuesday that would otherwise have had no orders at all is still better than staff standing about.

The version that works

Run both. Treat them as two different jobs.

The apps are your shop window. Keep the listing good — decent photos, accurate hours, sensible menu. That's your acquisition channel and you're paying for it.

Your own site is where the relationship lives. Every marketplace order is a lead. It arrives with a bag, and you get to put something in that bag.

Then move people across, deliberately:

• A card in every bag — including every app order — with your ordering address and a first-order discount.
• A QR code on the counter and the shopfront for collection customers.
• Tell people on the phone. "There's a website now, it's cheaper for you too" — and mean it, because with no service fee it genuinely is.
• Make the direct price better — but size it against the right number. On a £25 delivery you're about £3.70 up ordering direct once the driver is paid, so a 10% discount leaves you £1.40 in front. On collection you're £5.30 up, and the same 10% still leaves you £2.99. Put your best offer on collection.

How to tell if it's working

Watch the ratio, not the totals. If direct orders are climbing as a share of the whole while your total order count holds steady, it's working. If total orders are falling, you've moved too fast and you should push the apps harder for a while.

The number to aim for depends on how established you are. These are reasoned from the economics, not measured across clients — we're new at this — so treat them as a shape to expect rather than a promise:

• New shop, first year: most of your volume will be marketplace, and that is fine. You're buying customers.
• Established, good local reputation: a roughly even split is a sensible first-year target once a direct channel exists.
• Long-standing, strong regular trade: this is where direct should dominate, because most of your orders are from people who stopped needing to discover you years ago.

Nobody sensible gets to 100% direct, and you shouldn't want to. Some proportion of marketplace volume is customer acquisition you'd otherwise have to pay for another way.

When leaving is the right call

There are cases. Be honest about whether you're in one:

• Your margins genuinely can't take it. If marketplace orders lose money after packaging and delivery, they're not filling dead time, they're subsidising a loss. Do that sum properly before deciding.
• You're at capacity. If the kitchen is full every Friday regardless, marketplace orders aren't incremental — they're displacing better-margin direct orders. Cut the channel or raise the app prices.
• You've already got the customers. A shop with a large, active, direct customer list is paying for discovery it no longer needs.

Even then, most people are better off going quiet on the apps — pausing during peak, dropping the promotions, coming off premium placement — than deleting the account.

The honest summary

The apps are an expensive acquisition channel that has been sold to you as a sales channel. Used as an acquisition channel, they're fine. Used as your only sales channel, they're taking a third of your business for work they did once.

Put a direct channel in, keep the apps, and move the regulars across. That's the whole strategy. It's less dramatic than deleting Deliveroo, and it makes a lot more money.

---

Work out your own split with the savings calculator, or try the ordering system — all four roles, no signup.