Your Delivery Listing Is a Shop Window. Most Are Filthy.

Here is a pattern you see in almost every café that delivers. The owner spent weeks on the commission conversation — pushed back, compared offers, got a point or two somewhere. Then the listing itself went live and has not been touched since. One flat photo taken on a phone in bad light, the menu uploaded in whatever order the spreadsheet happened to be in, no modifiers on anything, and four items that have been showing as unavailable every afternoon since spring. A point of commission is worth real money. So is a listing that converts twice as well, and that one is free.

This post is about the shop window rather than the economics. The economics — what the commission actually does to a cup of coffee, and how to price for the channel — are covered in aggregator commission and pricing your delivery menu. And a caveat worth stating plainly: nobody outside the apps knows how their ranking works. The algorithms are undisclosed and they change. Everything below is either what the apps themselves tell merchants, or what operators consistently report — not a formula, and anyone selling you one is guessing.

The photo is doing most of the work

In an app, the picture is the product. A customer scrolling at nine in the evening is comparing your flat white against a grid of thumbnails, at the size of a postage stamp, and the decision happens before they read anything. One poor photo does not just lose that item — it drags the perceived quality of the whole listing down with it.

The practical version: shoot on a clear day near a window, on a plain surface, from slightly above for food and closer to eye level for drinks, with nothing in the frame that isn't the product. Shoot every item that sells, not just the hero ones, and shoot them again when the recipe or the cup changes. Crop square, because that is what the grid shows. If you have one afternoon, spend it on the top ten sellers — they are most of your revenue and most of your impressions. Our food photography guide has the detail, and it applies here more than anywhere, because this is the one channel where the customer never sees the real thing first.

Category order is a decision, not a default

Most listings are in upload order, which means the categories a customer sees first are an accident of how your menu file was typed. Fix that. Put the category that actually sells first, and the thing you most want to sell second. For most cafés the first screen should be hot coffee, then cold coffee in summer, then food — and the order should change when the season does, which nobody ever remembers to do.

Inside a category, the same logic. Bestsellers near the top, the item with the best margin near them, and anything you are indifferent about further down. Keep categories small enough to scan: a list of thirty drinks under one heading gets abandoned, where four headings of eight get read. And cut what doesn't sell. A long listing is not a strength in an app — it is a longer scroll before the customer finds the thing they were going to order anyway, and it is more items to keep accurate.

Item names get searched, so write them to be found

People search inside these apps, and they type plainly: "iced latte", "croissant", "karak", "matcha". If your item is called "The Morning Ritual", it does not exist to those customers. Lead with the plain name of the thing, and put the personality after it: "Iced Spanish Latte" rather than "Spanish Sunrise". Keep house names for the in-café menu board, where someone is standing in front of you and can ask.

Then use the description field for the three or four things that actually settle a decision: size, what is in it, how sweet, how hot, whether it travels well. Not adjectives. "Double shot, 350ml, oat milk available, lightly sweetened" sells more than a sentence about passion for coffee. If you serve an Arabic-reading customer base, write the name in both languages where the app allows it — the same reasoning as a properly translated menu in the café itself.

Modifiers are where the margin hides

This is the most commonly skipped section of a listing and the most profitable one. If a customer cannot choose oat milk, an extra shot, a size, or a syrup, two things happen: you lose the upsell entirely, and some of them don't order at all because the thing they drink isn't available. In the café those choices happen in conversation. In the app, if it isn't a modifier, it isn't on offer.

• Size — the single biggest one. A large has barely any more cost in it than a regular, and a meaningful share of customers will take it if asked.
• Milk — oat, skimmed, lactose-free. Price the premium ones at what they cost you, as covered in milk alternatives and pricing, and let the customer decide.
• Strength and sweetness — extra shot, decaf, less sugar. Cheap to offer, and the fastest way to stop a customer leaving over a detail.
• Add-ons that travel — a side of dates, an extra pastry. Keep these to items that survive a thirty-minute drive.
• Notes the kitchen can actually honour — only offer a modifier you can deliver at peak. A choice you ignore when it's busy produces a complaint, not a sale.

The operational half: availability and prep time

The parts of a listing you cannot photograph matter just as much, because they are what the customer experiences after they have already committed.

Keep availability honest and current. An item that is out all afternoon should be switched off at lunchtime and back on in the morning, not left showing. Cancelling an order because something has been sold out since two o'clock is the single most expensive thirty seconds of the day: the customer gets a refund, a bad impression, and no reason to come back. The apps also report acceptance and cancellation behaviour back to merchants, and the operators who do best on these platforms are the ones who treat that figure as an operational metric rather than a nuisance.

Be honest about prep time too. An optimistic prep time does not get the order out faster — it just hands the customer a promise your kitchen then breaks, and a rider waiting at the counter blocking your walk-in queue. If you genuinely need longer at peak, say so, and look at whether the real problem is throughput at the rush rather than the number in the settings.

Reviews are part of the listing whether you manage them or not

In-app ratings sit next to your name on every impression, and delivery reviews are different from café reviews: most complaints are about temperature, spillage, missing items and packaging rather than the drink itself. That is useful, because all four are fixable without changing anything a barista does. A lid that seals, a bag that stands up, a bit of tape, and a check that the order is complete before it leaves — see takeaway packaging for the detail.

Reply where the platform lets you, briefly and without arguing, and treat a recurring complaint as a product problem rather than a customer problem. Three people saying the iced drinks arrive watery is not bad luck; it is an ice specification. The same approach to online reviews works here, with the difference that on a delivery app the rating is attached directly to the thing you are trying to sell.

A twenty-minute audit

1. Open your own listing on a phone, as a customer, at the busiest hour. Not the merchant portal — the customer app.
2. Look at the first screen. Is the first category the one that actually sells?
3. Find your three bestsellers. Would you order them from these photos?
4. Search your own café for "latte" and your top food item. Do they come up?
5. Try to order your most popular drink the way a regular takes it. Can you?
6. Check which items are showing unavailable, and why, and whether that is still true.
7. Read your last ten reviews and sort them into product, packaging and timing.
8. Fix the three worst things this week, and put the rest in next month's diary.

Where MidaOne fits

The listing itself lives in each app's own merchant portal, and no till edits it for you — that work is yours. What a POS can tell you is what to prioritise, which is the part most owners are guessing at.

MidaOne reports sales by item and by category, so the "top ten sellers" you reshoot are the real ten rather than the ones you assume, and sales by hour shows you when the delivery peak actually lands against your counter rush. The same item-level view is what tells you whether a delivery-only item is worth its place on the menu at all. It also gives you a channel you own outright: your menu is a public page you can link from anywhere, and car ordering covers customers who would rather collect from the car park than pay for a rider. Neither replaces the apps. Both are margin you keep.

Frequently asked questions

Spend an hour on this before you spend another week on the commission rate. The commission is a negotiation you will have once a year with someone whose job is to hold the line. The listing is the part nobody is defending, and it is sitting there, open in the app, exactly as you left it.