Where food businesses lose customers online.
The long version: five things to check on your own phone, what the delivery apps really cost you, and how a stranger becomes a regular.
Your hours are wrong on Google.
Your menu is a photo from three years ago.
The phone rings while your hands are full.
So I build the part you don't have time for.
Five things I check first.
Every one of these takes about two minutes to check on your own phone. Do it standing behind your own counter before you talk to anybody, me included.
- 01
Your menu is a picture.
A PDF, or a photo of the printed menu taped onto a page. Google can't read the words inside either one, so the dishes in it don't come up when somebody searches for them by name. And a customer holding a phone at a red light has to pinch and drag around a page that was laid out for paper. Most of them give up and order somewhere else.
Check it: open your menu on your phone and try to read a price without zooming.
- 02
Google doesn't know when you're open.
Blank hours, or hours that were right two years ago. Someone drives fifteen minutes across town on a Sunday and finds the lights off. You lost that customer and you'll probably read about it in a review. There's a quieter cost too. With no hours on file you fall out of Google's "Open now" filter, so you don't come up at all for the person searching at 9pm for somewhere still serving.
Check it: search your own name on your phone. The hours sit right under it. Look at the holidays as well.
- 03
Nobody can tap to call you.
Your number sits on the page as plain text. On a phone that means copying it by hand and pasting it into the dialer. Sounds like nothing. It's the easiest thing a hungry person can do and your site makes them work for it.
Check it: pull your site up on your phone and tap the number. If nothing happens, there's your answer.
- 04
Every online order goes through an app.
DoorDash, Uber Eats, Grubhub, Slice. If the only order button on your site sends people to one of those, a cut of every ticket leaves the building before you've paid for the chicken. The bigger loss is quieter. The app knows who ordered and how to reach them. You don't, so you can't text that person about Tuesday.
Check it: tap your own order button and see whose logo you land on.
- 05
The site loads fine at your desk and nowhere else.
On your office wifi it opens instantly. On a phone with two bars in a parking lot it's a white screen for six seconds, and people don't wait out white screens. When I check a site I count anything past three seconds as a problem, and if it hasn't loaded in eight I treat it as down. It's almost always one enormous background photo nobody ever resized.
Check it: turn wifi off on your phone and count out loud while your own site loads.
None of those five breaks. That's the whole problem with them.
A fryer that quits, you fix the same day. A menu that's a photo doesn't quit. Wrong hours don't quit. A phone that rang while your hands were full never tells you it rang. Nothing on that list sets off an alarm, so it sits there, working exactly the same way tomorrow as it did today.
Which is why I can't hand you one number for what it costs. It isn't a bill. It's a rate. Every week the hours stay wrong is another week of people driving over to a locked door. Every week the menu is a picture is another week your own dishes don't come up when somebody searches them by name. Every order that goes through the app is one more regular whose number the app has and you don't.
I'm not telling you the sky is falling. Plenty of places run like this for years and do fine. I'm telling you it's the one cost that never shows up on a statement, and it keeps running whether anybody looks at it or not.
About the delivery apps
You already know about the commission. So I'll skip that part and go to the ones that cost you more.
Your app price is the only price a lot of people have ever seen
When the apps take their cut, most owners do the sensible thing and raise the price on the app to cover it. Nothing wrong with that. But think about who is actually reading that number.
Somebody who found you on DoorDash has never stood at your counter. They have never seen the menu on your wall. They have no second number to compare against. That is just what your food costs, as far as they know.
There is a Jamaican place near me here in Atlanta. Oxtail runs $31.99 on DoorDash. The same plate is around $18 at the counter. I checked both in August 2026. Nobody in that kitchen is overcharging anyone, and the owner is the one printing the cheaper number. But plenty of people scrolling that app have quietly filed him under expensive, and he will never hear about it. That is the loss. It does not show up on any statement.
Marking it up gets you less back than it looks like
This is the part I would want to know if it were my place. The commission comes off the subtotal on the app, and the subtotal is your marked up price, not your counter price. So the markup raises the fee along with it.
- Your counter price
- $18
- Listed on the app
- $24
- You keep on the 30% plan
- $16.80
- You keep on the 15% plan
- $20.40
Say a plate is $18 at your counter and you list it at $24 on the app to cover the cut. On the 30% plan you keep $16.80 of that $24. The customer paid six dollars more than your regular price, and you finished a dollar twenty under it.
Run the same plate on the 15% plan and you keep $20.40. Now you are $2.40 ahead and the markup did its job.
Same menu, same markup, opposite outcome. The plans run from about 15% up to 30% depending on which one you signed, so that one number decides whether any of this works.
You never get the phone number
On a delivery order you get a first name and a masked number that stops working once the driver is gone. No email you can write to. No number you can text on a slow Tuesday.
The app has all of it.
So a regular you earned through the app still has to be bought back from the app on the next order, and the one after that, for as long as they keep eating your food. You did the work of turning a stranger into a regular. You are renting them back by the plate.
Two things worth checking on your own phone tonight
Both are free and neither one involves me.
Pull up your own listing on the app and set it next to your printed menu. Read down the two columns. I would bet you set those app prices once when you signed up and have not looked at them side by side since.
Then open your merchant portal and find your actual commission rate. Not what you think it is. The number on the plan you are on right now. Everything above runs on that one figure, so find it before you believe anybody's math about your place. Mine included.
I'm not going to tell you to leave the apps.
They put you in front of people who have never heard of you, which nothing else does as cheaply. Somebody scrolls past your photo at nine at night and orders. Keep that.
What I build is the other lane. Ordering that lives on your own site, at your counter prices. The kitchen ticket prints the same way it does now. The customer's name and number land somewhere you own.
Then the people who already love you have a door that does not take a cut of every plate. And the app goes back to the job it is genuinely good at, which is introducing you to strangers.
How a stranger becomes a regular.
There are four steps, and every one of them is a place you can lose somebody. Each one is simple on its own. It's just usually nobody's job.
- 01
Somebody two miles from you is hungry right now and has never heard of you.
They're not comparing you to anybody. They don't know you exist. Before you spend a dollar telling them, the free version has to be working: whether your Google listing is right, whether the hours are right, whether the photos are yours and recent, whether you're even filed under the right category. That's what somebody two miles out sees first. It costs nothing but the work, so it's where I start.
- 02
They found you, and now they're picking between you and the place next door.
This happens on a phone and it's over in seconds. They look at your photos. They check how many stars you've got. They tap the menu and find out whether it opens or whether it's a PDF that loads sideways. Nothing about your cooking is involved yet, which is the frustrating part, because the cooking is the thing you're actually good at. So I rebuild the menu as a real page, get current photos on it, and set the review requests to go out on their own so your rating isn't just whoever was annoyed enough to type something.
- 03
They want it. Ordering is the last thing in the way.
Ten seconds or three minutes. That's the whole difference, and it gets decided by things that sound too small to matter. A checkout that demands an account. A phone number that doesn't dial when you tap it. An app taking a cut of a ticket your own site could have taken for free. I put ordering and payment on your site so the order comes to you and the money stays yours. If they'd rather talk to somebody, the AI picks up, takes the order, and nobody sits on hold.
- 04
They came once. Can you reach them again?
You already paid to get them in the door. Getting them back costs almost nothing by comparison, and it's still the step that gets put off, because nothing breaks when you skip it. So here's the plain version of the question. Do you have their number? If somebody ordered from you in March and hasn't been back since, is there anything at all that reminds them you exist? I set up the review ask and the win-back text, and the list that makes either of them possible.
Those four words an ad guy will say at you.
Awareness, consideration, conversion, retention. That's all they mean. You already run most of it in your head every service. You just don't have anything set up to keep doing it while you're on the line.
Where I actually am with ads
I don't run ads. I got certified by Meta in marketing this month, and a certificate is not a track record — I still haven't managed a dollar of anybody else's ad spend. When that changes I'll put it on this page.
Here's why I'm telling you instead of quietly leaving it off. An ad is a bill you pay to send people somewhere. If the place you're sending them doesn't open the menu or take an order, the ad just buys you a faster way to lose them. Steps two, three and four are the thing an ad lands on, and that's the part I build. Get those right and step one turns into a decision you make later with real numbers, instead of a guess you pay for every month.
The free report on the homepage grades you on the first two steps. Takes about ten seconds.
The rest of it
- Every systemEight systems and the website under them: what each one does, what it needs from you, and its price.
- PricingThe packages for trucks and restaurants, the discount for building more than one at once, and the terms.
- Online orderingWhy marking a plate up on the app raises the fee with it, and what your own order page has to do to be worth having.
- WebsitesThe ten things I measure on a food business site, the thresholds I measure them against, and what each one costs you.
- AI phoneWhat an AI phone line does on a real call, what it gets wrong, and how to catch it getting things wrong in week one.
- Google listingHours, categories, the attributes people filter on, photos and reviews — and the one ranking factor nobody can sell you.
- Food trucksWhat it costs to run a text list, why a location page has to admit when it is out of date, and when a truck is better off not buying one.
- Start hereThe homepage, and the free report — type your name, it pulls your Google listing and your site and tells you what it finds. About ten seconds, no email.
Want these five checked on your place? Text me the name.