How to Use AI in a Doner Shop: Daily Work, Data and Limits

This guide shows how a doner shop can use AI to read the daily spit prep, the gram card, the delivery line and the end-of-day summary. (Doner is meat stacked and roasted on a vertical spit, then carved to order.)

When the kilos to load in the morning, the lunchtime delivery rush and the portions left in the evening sit side by side in one report, the owner talks about the spit with data instead of guesswork.

Prices, allergens and the portion standard stay with you. AI tidies up scattered notes and turns Ritapos records into the language of your business.

Taking an order at the doner counter, where the portion grams and the delivery note get settled.

Get the basics straight before you bring AI into the doner shop

At the doner grill, the day starts with the first cut from the spit. With lavash, sauces, delivery notes and the till all moving at once, the job you give AI isn't to decide for the boss. It's to make clear which question needs reading first today. The distinction set out in the AI guide hub applies here too: numbers stay in Ritapos Ai, and text drafts stay in general chat tools.

Which jobs can you hand to AI?

Good jobs for AI include reading a sales summary by day and hour, interpreting the gram card next to the report, gathering up delivery notes, drafting replies to platform reviews and simplifying menu descriptions.

A draft calculation of how many kilos of meat to load on the spit, a summary of the weekly shopping list and short sentences comparing dine-in with delivery sales also fit. All of them rely on the Ritapos report, gram card and shift notes you supply.

What data do you need, and who checks it last?

At minimum you need units sold per product, the split between dine-in and delivery, your gram target and the end-of-day waste note. If the meat price has changed, add the cost-per-kilo line from the kitchen as well. AI won't make that number up.

The owner or kitchen lead approves the portion standard, price labels, allergen wording and the final reply that goes to a guest. Ritapos Ai explains the report, and it doesn't change the price at the till on its own.

Jobs AI can take on across a doner shop day

A doner shop day splits into three peaks: morning prep, the lunchtime rush and the evening close. At each one, AI asks the same question: which record is still missing today? The answer should be a one-paragraph briefing, not a long report.

Morning: how many kilos of doner to prepare

The morning decision covers the kilos to load on the spit, the number of lavash breads and the sauce prep. Put the same weekday's sales from the last few weeks in Ritapos, the delivery share and the kitchen note side by side, and AI gives you a readable draft along the lines of "prepare within this range today".

If you are just opening, matching the prep list in the how to open a doner shop guide to the product records in your POS is the first step towards a morning estimate that works.

Role: You are an operations assistant writing the morning spit prep briefing for a doner shop owner.
Facts given: [today's weekday], [units sold on the same weekday over the last four weeks], [dine-in and delivery share], [kilos of meat currently in the kitchen], [gram card target], [special day or event note]
Do not: Don't guarantee an exact number of kilos, suggest prices, make up the weather or write guest names.
Output format: Morning briefing: suggested prep range, three lines to check, and a time to look again before lunch.

Lunchtime peak: speed and delivery orders

At lunch, mistakes usually come from the delivery note rather than the portion: no chilli, no onion, sauce on the side, the drink forgotten. AI groups the notes that keep coming up and produces a short checklist for the till and the delivery line.

When platform orders land on the Ritapos screen, you can watch the dine-in queue and the delivery queue at the same time. Even if courier tracking and the delivery flow run through separate channels, the sales channel label must stay clear in the report.

Evening: leftover meat and the day's summary

At closing, leftover meat, unused lavash and any repeated delivery complaint go into the same daily summary. AI links those three lines to the next morning's briefing, so the spit prep doesn't ignore yesterday's waste.

The end-of-day summary for the boss should be a three-point note on what to check tomorrow, not a long table. The reading your Ritapos report guide explains how to keep that closing note short.

Estimating daily doner volume

An estimate means combining past sales with what is actually happening today. AI doesn't produce a magic number. It reads your Ritapos sales history and your shift notes.

Sales history by day and hour

Monday lunch is not Friday lunch. Ask about the durum (wrap), the plated portion and the delivery meal lines separately, and AI sums up in a sentence which product piles up at which hour.

To go deeper on stock, the stock and kitchen cost guide helps you track meat, lavash and packaging lines alongside the same daily note.

The effect of weather, matches and special days

You write the special day note yourself: schools going back, a derby, the eve of a bayram (religious holiday), a festival nearby. AI combines that note with sales history, but it never invents events on its own.

An estimate should always stay a range. Before lunch starts, the kitchen looks at the real queue and sets the pace of the spit by eye. AI just gets the prep moving earlier.

Portion grams, cooking loss and cost

The gram card is the doner shop's unwritten contract. If the units sold in the report don't match what is actually being carved in the kitchen, takings look fine at the end of the day while costs start telling a different story.

Portion yield per kilo

Cooking loss, waste on the spit and the gram difference between a plate and a durum all go in the same log. You can ask Ritapos Ai "which product seems to have drifted from the gram card this week?" and the answer comes back with the product name and the lines to check.

The inventory waste cost calculator and the kitchen count back up the yield per kilo with numbers. AI brings those two sources together and writes a short interpretation.

Price decisions when meat costs go up

When the purchase price of meat changes, AI doesn't push a new selling price on you. What it gives you is the product that needs discussing once your current menu price, portion cost and units sold sit side by side.

The price decision belongs to the owner alone. The menu price increase scenario and the menu profitability guide make clear which line to look at before you raise prices, and AI gives you a draft summary of that reading.

Delivery and platform reviews

As the delivery channel grows, you need to read dine-in and platform sales separately. When you talk about commission and net margin, use the delivery commission calculator together with the channel label in the report.

Comparing dine-in and delivery sales

The same durum carries a different cost in the dining room and in delivery: packaging, the platform's cut, courier waiting time. AI puts units sold and average basket per channel side by side in sentences, so you can see which channel only brings volume.

For businesses with similar speed demands, the way delivery peaks are read in the AI guide for fast food businesses can be adapted to a doner delivery line too.

Replying to complaints about delays and portions

The recurring themes in platform reviews are usually delays, sauce on the side or how big the portion felt. Instead of reading every review one by one, AI pulls out the theme, and you get a calm reply draft on behalf of the business.

Before you send the final reply, check it against what really happened in the kitchen and at the till. The guest reviews and content guide explains which sentences should never go out automatically.

Role: You are a customer relations assistant drafting replies to platform reviews for a doner shop owner.
Facts given: [review text], [order channel], [complaint topic: delay or portion], [kitchen confirmation note], [no compensation offer]
Do not: Discount offers, made-up guest names, blaming language, legal threats, guaranteeing portion size.
Output format: A two-paragraph reply draft: thanks, showing you understand the issue, a concrete fix, and a space for a human signature.

Durum, portions and menu descriptions

Short, clear descriptions on the QR menu and the platform listing cut down on order notes. AI writes the durum, plated portion, iskender and rice options in the same tone, and keeps the lines on spice level, sauce and sides separate.

Don't put allergen or definite ingredient claims in menu copy, and don't publish any sentence the kitchen recipe hasn't confirmed. For image generation, use the checklist in the food images for ads guide.

Step by step: starting with AI in a doner shop

In the first week, start with the morning briefing only: yesterday's sales, today's prep range, the delivery rush. Give the same data to Ritapos Ai or a general chat tool and see which answer reads better.

In the second week, ask about gram card drift once a week. In the third week, add the themes from platform reviews. Pick one job at each step. Don't write prices, social media and a stock overhaul plan all on the same day.

When product and channel labels are right in the POS, your questions to AI get shorter. Check where Ritapos Ai sits inside the app on the AI assistant page.

Decisions a doner shop should not leave to AI

Selling prices, promotional discounts, cutting grams or shrinking portions are not decisions for AI. Allergens, halal or regulatory claims and official document wording need human approval.

A complaint reply never automatically offers compensation, a refund or a free item. Using a doner photo in place of the real product is risky for both platform rules and guest trust. Any image must always match a shot taken or approved in the kitchen.

Going digital: Ritapos Ai for doner shops

Ritapos Ai reads the sales, product and channel data in your doner shop account and answers questions you speak or type. There is no separate program to install, and it opens from the app's Tools menu.

Sample questions to ask Ritapos Ai

At what time of day are delivery orders peaking today?

Did the durum or the plated portion sell more this week?

Compare dine-in and platform sales for the last seven days.

Is any product drifting from the gram card, and which line should I check?

What Ritapos Ai won't do

It doesn't fill in sales figures that aren't in your account. It won't change prices, set the spit's kilos automatically or post reviews to a platform on your behalf.

General chat tools can write text in a doner shop's voice, but they don't know your till data. Keep the two apart: the numbers live in Ritapos, the draft text comes from outside, and the decision is yours.

Frequently Asked Questions

What can a doner shop use AI for most?

The most practical place to start is the morning spit prep and the end-of-day summary. Once Ritapos sales history and the shift note are combined, the conversation about how many kilos to prepare gets much shorter.

Can AI estimate how many kilos of doner to prepare each day?

Yes, but as a range rather than a single number. It reads past sales for the same weekday, the delivery share and the special day note you wrote, all together. The kitchen still makes the final call as service begins.

How do you track the cost of a doner portion?

Bring the gram card, the cooking loss note and Ritapos product sales into the same weekly routine. Ritapos Ai tells you which product's cost line you need to discuss, and you enter the figures from the kitchen and your purchase invoices.

When meat prices rise, can AI set the selling price?

No. AI summarizes which menu lines you should look at, and the new price is the owner's decision alone. You complete the price increase scenario yourself with the calculator and the menu profitability guide.

Can delivery orders be managed on one screen?

Ritapos brings delivery and dine-in orders into the same till flow. If you use a platform integration, the channel label stays separate in the report, and AI helps you read that split.

Should doner photos be generated with AI?

AI can help with drafts or background edits, but it should never stand in for the real product on display. Before publishing, match the image to a reference shot taken in the kitchen and don't promise an exaggerated portion.