Start with the item-level report when you want to know whether an online menu change affected a particular dish. Then check the order-level report to see whether the change coincided with larger tickets, different order counts, or a shift in what guests bought together.
Neither report settles the question alone. More units sold can reflect more orders rather than a more appealing menu. A higher average ticket can reflect a different discount or a busier weekend rather than the item you changed. The useful answer comes from comparing the right denominators and recording what else changed.
Define the menu change and comparison period
Write down one decision before opening a report. For example: should the chicken bowl stay near the top of the online menu after being moved from a lower position? That question calls for a different analysis than deciding whether to remove the bowl or raise its price.
Record the old and new menu positions, the date and time the change went live, and the online ordering channels where guests could see it. If the item name, photo, description, price, or portion also changed, list those changes. They may affect the result even if your main interest is placement.
Choose a before period and an after period with the same number of comparable service days. A Tuesday-through-Monday week compared with another Tuesday-through-Monday week is usually easier to interpret than a slow weekday compared with a holiday weekend. Check opening hours and unusual events as well. If one period includes a closure or a large catering order, note it before calculating averages.
Decide which orders belong in the comparison. Direct online orders are a sensible starting point for a change made only to the direct menu. Keep pickup and delivery identifiable: they are dining options, not necessarily separate order sources. If the same menu change appeared on another ordering channel, you can analyze that channel separately rather than mixing its traffic and fees into the first result.
Assign someone to collect each input. The person who manages the menu can record the change and availability. A manager can check service hours, promotions, and unusual orders. The owner or bookkeeper can agree on the sales definitions used in both periods. Ask the software provider how to obtain the needed order and item detail if your current reports do not make it clear.
Use completed orders consistently, or choose another status that fits your operation and document it. Avoid counting canceled orders in one period but excluding them in the other. Decide whether refunds are reflected in the figures and apply that rule to both periods. The point is not to find a perfect report; it is to compare like with like.
Use each denominator to answer a different question
An order-level report treats each ticket as one observation. It can show order count, sales, and average sales per order. An item-level report treats each dish or unit as the subject. It can show units sold and item sales. Depending on the available detail, you may also be able to count how many distinct orders contained the item.
Those measures answer different questions. Units per order show the item's volume relative to overall orders. The share of orders containing the item shows how often at least one was chosen. Units per order containing the item shows whether those buyers tended to buy more than one. A report that only lists units cannot establish the last two measures without order-level item detail.
Hypothetical example: An owner moves a $12 chicken bowl higher on a direct online menu. The table compares one week before with one week after. Sales are gross menu sales before discounts, refunds, fees, and tax; the bowl price is unchanged. Fixed, variable, and exception costs are excluded, so these figures do not show profit.
| Measure | Before | After | What it shows |
|---|---|---|---|
| Completed direct orders | 100 | 120 | Overall order volume rose. |
| Gross sales across those orders | $2,000 | $2,520 | Sales per order rose from $20 to $21. |
| Chicken bowl units | 30 | 42 | Units per 100 orders rose from 30 to 35. |
| Orders containing a bowl | 25 | 30 | The share containing it stayed at 25%. |
| Gross bowl sales | $360 | $504 | Units multiplied by the unchanged $12 price. |
The bowl sold 12 more units, but the restaurant also had 20 more orders. Looking only at units would overstate what the change tells you. One quarter of orders contained a bowl in both periods. Among orders containing it, however, units rose from 30 ÷ 25, or 1.2, to 42 ÷ 30, or 1.4. Guests who chose the bowl bought more units per bowl-containing order in the second week.
That is a useful observation, not proof that the new placement caused it. Larger groups, a promotion on another item, or a few unusually large orders could explain the difference. The $1 increase in average gross sales per order is also a restaurant-wide result, not an extra dollar attributable to the bowl. A practical next move is to keep the placement for another comparable period and inspect whether the pattern holds.
Check availability, discounts, and substitutions
Item-level comparisons become misleading when guests did not have the same opportunity to order. Record whether the item was available throughout each service period and whether it was hidden, sold out, or temporarily unavailable. If it was unavailable for a busy Friday dinner in one week, a simple seven-day total does not give both periods equal exposure.
Where your data permits, compare orders placed while the item was available in both periods. Use that eligible order count as the denominator for the share of orders containing the item. If you cannot establish availability at the order level, compare similar service windows and flag the missing detail. Do not divide item units by all orders and describe the result as a true selection rate when many guests could not select it.
Discounts need their own column in your notes. A coupon may apply to the changed item, another item, or the whole order. Gross item sales at menu price can rise even when the amount collected after discounts rises less. Ask how your reports allocate order-wide discounts to items, and whether item sales include modifiers, refunds, and voids. Use the same definition in both periods.
Look for substitutions too. If the bowl gains sales while a similar salad loses them, the menu may have shifted demand rather than created new demand. Compare units per 100 eligible orders for the closest alternative, not just its raw unit total. Then check the combined sales of the two items. A shift can still be worthwhile if it helps the kitchen or supports a menu strategy, but sales alone will not establish that it improved contribution.
Also separate the sales result from operating cost. Food, packaging, labor, payment processing, delivery or dispatch charges, and other costs may change with the mix. An item report may not contain those costs. If cost matters to the decision, bring in reliable estimates from the people who manage purchasing and operations, and label them as estimates. Do not treat gross sales or average ticket as profit.
Use an order-versus-item worksheet
A short worksheet keeps the decision focused and exposes gaps before they become confident conclusions. It can live in a spreadsheet or a manager's daily notes. Complete these fields for each comparison:
- Decision and change: Name the item, the specific change, its launch time, and the decision you will make.
- Comparable periods: Record dates, days, service hours, channel, dining options, and included order statuses.
- Item inputs: Record available service windows, units, gross or net item sales as defined, and the count of orders containing the item if available.
- Order inputs: Record eligible order count, sales using a consistent definition, average sales per order, and any unusually large orders.
- Context: Note price changes, discounts, refunds, outages, closures, and changes to similar items.
- Decision rule: State what result would justify keeping, changing, or reversing the menu change, and when you will check again.
Make the decision rule specific to the goal. If the goal is more guests choosing the bowl, watch the share of eligible orders containing it. If the goal is more bowl volume without hurting similar items, review units per 100 eligible orders alongside the alternatives. If the goal is a higher ticket, use average sales per order as a supporting measure, while recognizing that other changes can affect it.
Before using the worksheet, ask your provider whether you can identify distinct orders containing an item, filter by order source and status, and compare the same date and time windows. Ask whether an export is available and how discounts, modifiers, and refunds appear. If those details cannot be produced, use the measures you do have and narrow the decision accordingly.
Fleksa's restaurant software includes dashboard, order, and menu views, with sales and order summaries for a selected period; available views and configuration can vary. Ask what detail is available for your setup rather than assuming every worksheet field appears in one report. The online ordering buying guide can help with the wider system decision. For the menu itself, consider how your restaurant website presents items alongside the ordering experience.
Set a follow-up before calling the change a success
A small sample can move sharply because of a few tickets. Check the number of orders behind each percentage, and inspect unusually large orders separately. You can show results with and without those orders, but do not quietly remove them just because they weaken the result.
Seasonality, local events, ads, weather, and changes in operating hours can all affect a before-and-after comparison. If several menu changes went live together, the report may show their combined result but cannot reliably assign the outcome to one change. A longer comparison or a later reversal may help you understand the pattern, provided you record other changes and avoid disrupting service unnecessarily.
Reconcile the totals before sharing a decision. Confirm that the order count uses the intended source and status, that item sales belong to those same orders, and that both periods use the same treatment of discounts and refunds. If the totals differ, ask whether timing, canceled orders, or report definitions explain the gap. Document the answer rather than forcing the numbers to match.
Use the worksheet on one current menu decision. Then ask Fleksa about its restaurant online ordering setup, the report fields you need, and verified US availability for your location.
Frequently asked questions
Which report should I check first after changing one menu item?
Start with item units and, if available, the share of eligible orders containing the item. Then use the order report to check overall volume and average sales per order.
Does a higher average order mean the menu change worked?
Not by itself. A different mix of guests, larger group orders, discounts, or other menu changes can move the average. Compare item measures and context before attributing the increase.
What if my report only shows item units?
Compare units per 100 eligible orders and record availability. Ask whether order-level item detail is available; without it, you cannot calculate the share of orders containing the item or units per buying order.
