Changing a menu price is one of the quickest ways to move your margin, and one of the riskiest. Raise everything at once and some dishes stop selling; leave everything alone and rising costs eat the profit. This guide explains how AI smart pricing in Kappino works, from the data it reads to the moment you decide whether a price changes.
The data it uses
Smart pricing works on your own data: the last 90 days of sales and demand, your item costs and how each dish performs on the menu. Demand models are trained on that data. It does not use competitor prices or weather data.
Item costs come from recipes linked to stock in menu management, which show the cost and margin of each dish.
Step 1: refresh the recommendations
Kappino reviews recent sales and demand for every menu item and produces a recommendation per dish. There are six kinds:
- Raise
- Lower
- Hold
- Surge
- Discount
- Clearance
Each recommendation shows the expected change in revenue and profit per week, a confidence level and a risk level.
See where each dish stands
The menu engineering matrix places every item, based on the last 30 days, as a Star, Plow Horse, Puzzle or Dog, so you can see what to promote, re-price or rethink before you change anything.
Price elasticity shows which items are price-sensitive and which are not, so you know where a higher price is likely to cost you sales.
Step 2: simulate before you commit
Pick an item and a price change in the what-if simulator to see projected revenue, profit, volume and retention over 7, 14 or 30 days. The price impact forecast adds the expected effect of the change.
Step 3: decide and apply
Kappino never changes a price on its own. You apply one recommendation, or all of them from AI Insights, and only admins and managers with pricing permission can do it. An applied price updates the menu across the POS and online ordering, including the QR menu.
Step 4: track the result
After you apply a recommendation, Actual vs predicted shows how the dish really performed against the prediction. As your data grows, re-train the model.
How it fits with forecasting
AI sales forecasting predicts demand for the next 7, 30 or 90 days, with peak hours and top-demand items, and its recommendations include pricing suggestions next to staffing, reordering and promotions. You can also ask Pino AI about your sales and menu performance in plain language.
A seasonal routine
- Look at the menu engineering matrix before you reprint the menu.
- Refresh recommendations and read the confidence and risk of each.
- Simulate the changes you are considering.
- Apply the ones you agree with.
- Check actual vs predicted a few weeks later.
AI item smart pricing is part of the Professional plan. See pricing for what each plan includes.
