When a Sales Projection Assumed a Restaurant No One Was Running

The situation

Anh worked as a farm worker and Tuan as a factory technician, and between them they had spent several years saving enough to buy a small mixed-use building in Oakville — two residential units upstairs and a street-level commercial unit they leased out. It was meant to be a straightforward landlord arrangement: collect rent, cover the mortgage, keep the building in good repair. For most of the time they owned it, that is exactly what it was.

The commercial tenant, a small breakfast-and-lunch counter, closed partway through one year when its operator retired. Rather than leave the space empty while they searched for a new tenant, Anh and Tuan decided to run a scaled-down food counter themselves for a few months — coffee, pastries, a short lunch menu — mostly to keep some income flowing from the space and to make it easier to show as a going concern to prospective tenants. Neither of them had run a food business before. They registered for HST since the commercial unit was already generating rental income that pushed their combined activities over the threshold that requires registration, kept the till receipts and supplier invoices in a shoebox, and filed HST returns based on what the register actually took in. A few months later they found a new long-term tenant, handed the space back over, and went back to being landlords with a single commercial rent roll.

What the review found

Close to two years later, the Canada Revenue Agency selected the period covering the food counter for an HST audit. Food service businesses are audited more often than most sectors, largely because they tend to handle a high proportion of cash and because sales are easy to under-record if a business is inclined to. Anh and Tuan's operation was neither large nor deliberately under-reported, but it looked, on paper, exactly like the kind of file the audit selection criteria are built to catch.

The auditor did not simply compare the till tapes to the HST returns. Because the till records were thin — a few months of handwritten summary sheets rather than a proper point-of-sale system — the auditor used an indirect verification method commonly called a sales projection: take the food and supply purchases recorded on supplier invoices, apply a standard industry markup percentage that CRA's audit manuals use for the restaurant sector, and back into what sales should have been if the business ran at a typical margin. The projection came out well above what Anh and Tuan had actually reported, and the CRA issued an assessment for the shortfall in HST, plus interest and a penalty, totalling just under $14,000.

The trouble was that the standard markup assumption did not fit what the food counter actually was. It priced supplies as though every ingredient purchased became a sold item at full restaurant pricing, with none of it lost to spoilage, given away, or sold at a discount. In practice, a large share of what Anh and Tuan bought never turned into a paying sale at all: unsold pastries were donated to a local shelter at the end of each week rather than thrown out, staff and family ate meals on shift as part of the arrangement with no cash changing hands, and the short lunch menu was priced deliberately low to draw in a new customer base for whatever tenant eventually took over the space — a strategy, not a bookkeeping gap. None of that showed up in a projection built for a full-service restaurant operating at standard margins for a full year.

What we did

1. Reconstructed the actual purchase-to-sale relationship. We worked with the couple and the bookkeeper they brought on partway through the process, Genevieve, to go through every supplier invoice for the period and separate purchases into three categories: items sold for cash, items donated, and items consumed as staff or family meals. This turned a single lump purchase figure into a breakdown the CRA's markup formula had never accounted for.
2. Documented the donations with third-party confirmation. We obtained written confirmation from the shelter of the weekly food donations during the relevant months, giving the CRA something more concrete than the couple's own recollection to support the reduction in taxable sales the donated stock represented.
3. Challenged the applicability of the standard markup benchmark. A sales projection is a reasonable tool when a business genuinely resembles the sector average it is measured against. We put together a short, factual comparison — a deliberately short-term operation, a below-market lunch pricing strategy aimed at attracting a future tenant's customer base rather than maximizing the couple's own margin, and no drive-through or delivery volume — to argue the standard restaurant markup was the wrong benchmark for this particular few months.
4. Filed a formal objection within the required deadline. An assessment like this can be disputed through a notice of objection, which has to be filed within a strict window after the assessment is issued. We filed on time, attaching the reconstructed purchase breakdown and the shelter's donation records as supporting evidence.
5. Negotiated directly with the CRA's appeals division rather than pushing straight to the Tax Court. Once the objection was under review, we had several exchanges with the appeals officer assigned to the file, walking through the donation and staff-meal adjustments line by line and proposing a revised markup figure that split the difference between the standard restaurant benchmark and the couple's own reported numbers, given that both had some basis and neither was likely to be accepted outright.

The outcome

The appeals officer accepted the documented donations and staff meals as legitimate reductions to the projected sales base, but did not accept the couple's argument that the standard markup should be set aside entirely — the CRA's position was that some independent verification method was appropriate given the thin original records, even if the specific figure needed adjustment. The two sides settled on a revised markup that fell between the CRA's original benchmark and the couple's actual reported margin, reflecting the below-market pricing strategy without ignoring that a projection tool serves a legitimate purpose when records are incomplete.

The final HST reassessment came to roughly $6,900, down from the original assessment of just under $14,000 — close to a fifty percent reduction — with the associated penalty reduced proportionally as well. Anh and Tuan paid the revised amount in full shortly after the settlement was confirmed, closing the file without a trip to the Tax Court of Canada, which would have meant a longer process and costs that were unlikely to be worth it given the amount actually in dispute by that point.

Since the audit, the couple has kept the commercial unit strictly as a rental and has not run a business out of the space themselves again. When the current tenant's lease comes up for renewal, they plan to require proper point-of-sale reporting as a condition of any future short-term arrangement they might consider, rather than relying on a shoebox of receipts the way they had.

What you can learn from this