Two restaurant reports can both have a column called sales and still answer different questions. In the fictional example below, a seemingly better result disappears once the definitions match. Then one change in the activity covered makes the whole comparison unavailable.
Valsoft’s acquisition of Mirus offers a useful setting for that exercise: a Canadian buyer purchasing a U.S. restaurant-data business. The practical owner question is what must be understood about a number before using it to compare operations.
What’s In this Issue?
The business and the acquisition
The work behind a comparable report
Two historical customer examples
A fictional comparison: 20% becomes 0%
What changes when a channel is missing
A restaurant-data business joins Edelweiss
In its September 16, 2026 announcement, Montreal-based Valsoft said the acquisition of Houston’s Mirus Information Technology Services was completed through its wholly owned Edelweiss Software Group. That establishes the announcement date, not the precise legal closing day. Price and financing terms were not disclosed in the release.
The release says Mirus was founded in 1999 and brings restaurant operating data into reports, dashboards, calculations and alerts. Valsoft positioned the purchase as Edelweiss’s entry into restaurant analytics and said Mirus would remain autonomous with its team. Those are the announcement’s rationale and plan, not demonstrated results.
Mirus’s undated company account, inspected September 27, traces its origins to menu-data analysis and development of a warehouse and report-writing product. Its current biographies describe pipeline work and custom connections across point-of-sale and workforce systems. That explains the kind of work behind the interface; it does not establish the exact staffing or product configuration on the acquisition date.
Access is one step; meaning is another
Mirus’s undated implementation article, inspected September 27, describes receiving data through connectors, file transfers or APIs. It then describes cleaning and mapping data, applying client-defined business rules and comparing the result with source-system reports alongside the client. This is the provider’s account, not an independently tested performance claim.
The client’s role matters because access to records does not supply every definition needed to interpret them. Mirus describes clients as a bridge to other vendors, with knowledge of their own reports and business practices. Our reading: a usable comparison depends on both the connection and the meaning supplied around it. A matched report alone does not prove all activity was included.
The company’s warehouse page, also undated and inspected September 27, identifies inconsistent descriptions across departments as a problem and describes centralized reports updated with incoming transactions. The relevant distinction is between collecting records and deciding which records should contribute to a particular measure. Neither a tidy screen nor a familiar label resolves that decision.
Historical examples show different reporting jobs
In an undated Mirus interview about Dewey’s Pizza, Jason Busch was described as providing point-of-sale management through his consulting firm. He discussed year-prior comparisons used for scheduling and forecasting, and an exception report identifying stores that had not polled. The latter addresses whether expected data has arrived, a different question from what a sales field means.
A Portillo’s interview published by Mirus, expressly conducted before COVID, describes bringing in Aloha point-of-sale sales and later adding general-ledger account names, subaccounts and mappings. Here the reporting task expanded into a different classification structure. These vendor-published accounts are historical illustrations, not evidence of current customer systems, measured returns or acquisition benefits.
A broader historical reference is Oracle’s hotel-restaurant reporting datasheet, version 1.01, bearing a 2020 copyright. It describes a warehouse and reports covering measures such as sales, labour and discounts in a hotel/casino food-and-beverage setting. The shared idea is organizing operating data for decisions; the setting and product are different. This is context for the reporting task, not a current vendor ranking or evidence that either product connects to the other.
A fictional comparison: 20% becomes 0%
Imagine two fictional restaurants, A and B, reporting one trading day in Canadian dollars. Start by stipulating identical activity coverage. Exclude taxes, tips, returns, foreign exchange and every adjustment except the discounts specified below. The chosen operational measure is sales after those discounts. It is not accounting revenue, cash or profit.
Site A’s export has a field labelled sales showing C$10,000. Its definition already deducts C$1,000 of discounts. Site B’s sales field shows C$12,000, but its definition is before C$2,000 of discounts. These definitions and amounts are invented, not Mirus settings or actual customer records.
Read just the two labelled fields and B appears 20% higher: C$12,000 divided by C$10,000, minus one. The arithmetic is correct for those inputs. The comparison is unsuitable for the stipulated after-discount measure because the inputs represent different things.
To calculate that measure, subtract B’s C$2,000 discounts from its C$12,000 field. B becomes C$10,000. Leave A at C$10,000: its discount has already been deducted. Subtracting A’s C$1,000 again would create a new error while trying to fix the first one.
Now the difference is 0%: C$10,000 divided by C$10,000, minus one. No underlying activity changed. We changed which values entered the comparison so that they answered the same stated question. Equal totals still do not establish equal volumes, business performance or profitability.
The useful record is therefore more than the number. For this exercise it needs the field’s definition, whether the discount is already included, and the activity and period covered. Someone reviewing the calculation should be able to reproduce the treatment without guessing what the label was intended to mean.
Then discover that a channel is missing
Change one assumption. A still covers all intended activity. B’s same C$10,000 after-discount report now covers only its in-store channel; a digital channel is missing and its amount is unknown. Keep the field definitions and discount treatment unchanged.
B’s calculation can still agree exactly with its in-store source report. That agreement answers whether the available report was reproduced. It cannot answer whether the whole restaurant matched A. The previous 0% whole-site comparison must be withdrawn because the common-coverage premise no longer holds.
Do not enter zero for the missing channel or assume it must contribute a positive amount. Either would add an input the example does not provide. There is no supported full-site total for B and no revised whole-site percentage to report.
The next useful step is specific: identify the intended channels and obtain the missing channel’s source for the same day, with a compatible definition. Until that input is available, label B’s result as partial. If the intended task instead becomes an in-store comparison, A would also need an appropriate in-store figure; its full-site total cannot simply be reused.
This change in the example is consequential. The first problem could be resolved using definitions and numbers already supplied. The second cannot. Repeating the calculation more carefully will not create the missing evidence. The owner’s record should show which question remains unanswered, so a neat equality does not travel into the next decision as a complete result.
Keep the definition with the number
For the fictional exercise, a compact record could hold the measure, the source field, its discount treatment, the day and channels covered, the comparison report and who can resolve a mismatch. That is a proposed operating aid, not an accounting policy or a term of the Mirus acquisition.
The final test is whether another person can tell what the result includes and what would invalidate it. In our example, that makes the difference between a correct calculation, a meaningful comparison and a conclusion that should wait. A number becomes more useful when its limits can travel with it.
For general information and education, not legal, tax, investment or valuation advice. Examples are illustrative and do not predict your business’s value, financing terms or sale outcome. Consult qualified advisers about your situation.
Subscribe to Deal Flow Canada for practical owner guides and Canadian deal updates.