Start by making the dashboard and export answer the same question. Select one location, one date range, one time zone, and one sales measure. Then compare the dashboard total with the corresponding export total—not with a payment deposit or a tax-inclusive amount.
If the numbers differ, first check whether the two views include the same orders and adjustments. A mismatch can come from timing, filters, refunds, rounding, or a genuine reporting problem. The buying decision is whether the vendor can explain the difference and give your team a repeatable way to resolve it.
Set the boundaries before comparing totals
Write down the dashboard settings before requesting an export. A screenshot or written record should show the location, reporting period, time zone, selected sales figure, filters, and the time the dashboard was viewed. Save the export with its download time and the settings used to create it.
Use a completed reporting period for the first comparison. An active shift is harder to reconcile because new orders, adjustments, and refreshes can change the total while you work. Ask whether the system assigns orders to a calendar day, a business day that ends after midnight, or another period. The same displayed date can cover different transactions if those boundaries differ.
For a useful comparison, identify the source of each input. The dashboard display is the source for the dashboard number; the downloaded report is the source for the export number. Order-level records are the next place to investigate a difference. Record what each surface actually shows separately from what you assume its labels mean.
For example, do not assume that a filter labeled “completed” treats canceled orders, partial refunds, and edited checks the same way in both views. Ask the vendor to explain the rule, then confirm it in the demonstration. Also check whether the export contains individual order lines, order totals, or a summary. That determines how independently you can trace a mismatch.
For a broader look at how reporting fits into operations, see this guide to one platform versus separate restaurant tools. When reviewing a particular system, inspect its report workflow as well; Fleksa describes a sales report workflow. Confirm the fields and export options available for your setup rather than assuming a workflow supports every comparison.
Use one sales definition and compare the outcomes
Choose a measure that both surfaces can show. Gross sales, net sales, tax collected, payment deposits, and processing fees answer different questions. A deposit is money settled to an account; it is not automatically the same as sales. Fees may affect the deposit without changing the sales figure. Do not add either one to a sales reconciliation unless you are separately tracing settlement.
The following worksheet is hypothetical. Assume one location, a completed business day from 4:00 a.m. to 3:59 a.m. in the location’s time zone, and the same included orders in both views. Assume both reports assign discounts and refunds to that business day. The chosen measure is gross sales less discounts and refunds, excluding tax, tips, payment fees, and delivery charges.
| Worksheet item | Dashboard | Export or calculation | Decision |
|---|---|---|---|
| Gross sales | $1,240.00 | $1,240.00 | Same starting measure |
| Discounts | $60.00 | $60.00 | Subtract once |
| Refunds | $25.00 | $25.00 | Subtract once |
| Net sales | $1,155.00 | $1,240.00 − $60.00 − $25.00 = $1,155.00 | Pass the total comparison |
| Alternative export outcome | $1,155.00 | $1,154.99 | Investigate the $0.01 difference |
The first outcome is a match: $1,240.00 less $60.00 and $25.00 equals $1,155.00 on both surfaces. Assume the reports also show $92.40 in tax. That tax is excluded from net sales under this worksheet’s definition; adding it would produce $1,247.40, a different measure. The example does not specify a tax rate or determine how tax should be treated in your books.
The alternative outcome is a one-cent difference, not a match. It may be explainable, but the vendor should show why it occurs and which figure staff should use for the agreed purpose. A repeated, documented rounding difference may be manageable if the report provides a clear audit trail. An unexplained difference—or a report that omits adjustments without making that clear—is a stronger reason to reconsider the workflow.
Ask how the system defines gross sales and net sales. Does a refund reduce sales on the original order date or the date the refund was issued? Are discounts shown before or after tax? Does a partial refund appear as its own row? There is no useful comparison until both surfaces apply the same rules, even if their totals happen to match on one quiet day.
Trace a one-cent failure to its source
Continue the hypothetical example with the export at $1,154.99 and the dashboard at $1,155.00. The difference is $0.01. Assume all included orders and adjustments match, but the two surfaces round a small group of calculated line amounts differently.
In this illustration, three amounts after the relevant adjustments are $10.005 each. With half-up rounding, rounding each line to cents gives $10.01 + $10.01 + $10.01 = $30.03. Adding the unrounded amounts first gives $30.015, which rounds to $30.02. Assume the remaining net sales total $1,124.97 on both surfaces. The line-rounded dashboard therefore shows $1,155.00, while the aggregate-rounded export shows $1,154.99.
A shift manager can detect this issue during the daily review by comparing the selected dashboard total with the export. The manager should flag the difference rather than silently changing a spreadsheet cell to force a match. The reporting owner, such as the restaurant owner or bookkeeper, can then ask the vendor to identify the affected rows and document the rounding rule.
The practical fix is a consistent reporting procedure: record which view rounds at which stage, preserve the original export, and use the agreed figure for the specific internal report. If your accounting process needs a particular treatment, ask a qualified adviser. Do not assume rounding explains every cent difference; an omitted refund or a different cutoff could produce the same visible gap.
Work from totals toward details. Compare gross sales, discounts, refunds, and tax separately. If one component differs, narrow it to order IDs or line items. If every component agrees but the final total differs, ask about rounding and calculation precision. Keep the original files and note any filters changed during the investigation, so another person can repeat the comparison.
Run a vendor demonstration with clear pass criteria
Give each vendor the same scenario and ask them to use a simulated or demonstration environment. Practice checks must not create a real card payment, send a guest message, or dispatch a courier. If the vendor cannot safely demonstrate a transaction, use existing non-sensitive sample records and ask how the missing step would be verified.
- Set the scope. Select one location, a completed business day, and a named time zone. Have the demonstrator show the exact start and end boundaries.
- Define the measure. Agree on gross sales, discounts, refunds, and tax treatment before viewing the answer. Specify whether tips, delivery charges, and fees are excluded.
- Record the observations. Capture the dashboard total, its filters, and the time it was viewed. Download the corresponding report if the workflow supports it; record its filters, generation time, and refresh information.
- Recalculate. Use the export’s displayed fields to reproduce the agreed total. Check that the report contains enough detail to investigate at least one discount and one refund.
- Introduce or inspect a difference. In the simulation, compare a known exception or a prepared sample showing a mismatch. Ask the demonstrator to trace it without altering the original report.
- Assign the exception. Name the restaurant reporting owner and the vendor support contact or support channel. Agree on what evidence goes with an escalation and when the next update is expected.
Pass: for the agreed scope and measure, the dashboard and export match, or the vendor traces every difference to identifiable records and a documented calculation or timing rule. The team can repeat the steps, and the reporting owner knows which figure to use and how to escalate an unresolved exception.
Fail: the vendor cannot show equivalent filters, reproduce the calculation, explain a difference, or identify a route for investigation. A total that matches without a usable explanation of its inputs is not enough. Record a limitation if the needed export or detail is unavailable, then decide whether another report or a separate tool can meet the requirement.
Refresh timing matters. Ask when each surface last updated, whether an export is a snapshot, and how long a newly completed order or refund normally takes to appear. Set an agreed wait and recheck time for your team. A difference observed before both views have refreshed should remain open until that recheck, not be recorded as a permanent discrepancy.
After the simulation, a live service pilot is a separate decision. If you proceed, review naturally occurring orders and refunds after service rather than creating real guest transactions solely for the test. Have the reporting owner log exceptions, their resolution, and any manual work required. That record makes the buying decision more useful than a single successful demo.
Choose the workflow your team can repeat
A small restaurant may accept an export with manual checking if the report has clear fields and the daily review takes a manageable amount of staff time. A multi-location operator may need location-level filters and a consistent process for escalating differences across sites. Ask for a written demonstration of the exact workflow and any operational quote for the relevant setup; confirm what support is included.
Before deciding, compare your requirements with Fleksa’s restaurant management software and confirm current support for the report, filters, and export details you need.
Frequently asked questions
Should a dashboard always match an export to the cent?
For the same measure, period, filters, and refresh state, expect a match or a traceable explanation for every difference. Document any rounding rule and decide whether it works for your reporting process.
Why does a sales total differ from a bank deposit?
Sales and deposits measure different things. Settlement timing and payment fees can affect a deposit, while the sales report may use a different period or include amounts outside the deposit. Reconcile settlement separately.
Who should handle a reporting mismatch?
A shift manager can flag it during the daily check. Assign a reporting owner to preserve the evidence, review order-level details, and contact vendor support when the cause remains unresolved.
