What this report is for
MyFoodFast suggests extra items to customers while they build an order on the kiosk, the mobile app and your website, for example a side or a drink at the basket. The Upsell Performance report shows how those suggestions are doing, and whether they are actually raising the value of an order rather than just being clicked.
Go to Reports > Online Reports > Upsell Performance. You can view it with a store selected in the switcher (one store) or at Company level (all your stores, with an extra By store table).
Use the date picker at the top to change the period. The default is the last 30 days.
The headline: does upselling raise order value?
The card at the top compares two groups of paid orders:
- Orders shown suggestions — carts where the engine offered items.
- The no-upsell holdout — a small, random share of carts (about one in ten) that are deliberately shown no suggestions.
Because the holdout is chosen at random, the difference in average order value between the two groups is a fair measure of what the suggestions add. It is the only number on the page that measures the effect of upselling; everything else describes activity.
Note: The card says not enough data yet until both groups have 200 paid orders. Until then it shows how many orders each group has so far. Small numbers swing a lot, so the report will not show a difference it cannot back up.
Once there is enough data you will see the two average order values, the estimated difference, and a ± figure. The ± is the margin of error. If the margin is larger than the difference, the report says the interval includes zero, which means no effect has been detected at that sample size.
The summary tiles
- Carts shown suggestions — how many baskets the engine offered items to. This is recorded by MyFoodFast itself, so it does not depend on the kiosk or app reporting back.
- Suggestions accepted — how many suggested items customers added, and the share of suggestions shown. The kiosk, app and website report this slightly differently, so compare like with like.
- Suggestions on a paid order — suggested items that ended up on the paid order, checked against the order itself rather than what the customer tapped.
- Retained value — the total paid for those items. This is not extra revenue on its own: some customers would have ordered the item anyway. The holdout comparison above is what tells you how much is genuinely extra.
Funnel by client
One row each for the kiosk, mobile app and website:
- Carts shown suggestions, Suggestions shown, Accepted and Accept rate.
- On paid order and Retained rate — how often a shown item ended up on the paid order.
- Retained value — what those items were worth.
- Holdout carts and Bought unprompted — how many carts were held out, and how often customers in the holdout bought the same items without being asked. If "bought unprompted" is close to the retained rate, the suggestions are not changing much.
Order value by client and by store
These tables repeat the headline comparison for each client and, at Company level, for each store. A difference is only shown once both groups for that row reach 200 paid orders; smaller rows show how far along they are.
Suggested products
The most-shown items with their accept rate, how often they landed on a paid order, the value retained, and how often the same item was bought unprompted in the holdout. A product that customers accept often but also buy unprompted just as often is not adding much; one that is accepted and rarely bought otherwise is earning its place.
Data quality
This section is collapsed by default. It shows how much of your paid order volume the report can see, and any gaps in what the kiosk, app or website send back, for example accepted items with no basket attached or rows with no price. If coverage is low for one client, the numbers for that client are less reliable, and the MyFoodFast team may already be working on it.