Stopping a Two-Week Rental Sample From Setting a Full Year

The situation

The letter gave Bogdan thirty days to respond, and by the time it reached him, worked its way through a stack of unopened mail, and actually got read, eleven of those days were already gone. What was left was not enough time to do what the letter was effectively demanding, which was to disprove a full year's worth of projected income built from a two-week snapshot, before the CRA moved from a proposal letter to a formal reassessment that would be far harder to unwind afterward, and a formal number, once it issued, carries a presumption of correctness that shifts the burden onto the taxpayer to disprove it.

Bogdan worked full time as a physiotherapist and, alongside that, owned several short-term rental properties in and around Brockville, a business he ran with his brother Carlos, an accountant, who handled much of the bookkeeping, and his sister Ines, who managed bookings and guest turnover on the ground. The three of them had built the rental side of things up gradually over several years, starting with a single property and adding more as the first one proved itself, and by the time of the audit the portfolio had grown into a genuine second income for the family. The properties did well in peak season and considerably less well the rest of the year, which is ordinary for short-term rentals in a seasonal market but had apparently caught the auditor's attention in a different way.

The CRA had audited a two-week stretch in the middle of the busiest part of the season, pulling booking platform records, cross-referencing them against reported income, and finding a gap between what the platform showed as collected and what had been reported for that window. Rather than treating the gap as isolated to those two weeks, the audit proposal applied that same rate of underreporting across the full fifty-two weeks of the year, producing a projected reassessment in the range of two hundred and fifty to three hundred thousand dollars once penalties and interest were factored in, a figure that bore almost no resemblance to what the properties actually earned across a full twelve months.

The trouble was timing on every front at once. The response deadline was closing fast, the family's usual division of labour on the file was about to be disrupted by circumstances no one had planned for, and the number itself, if it hardened into a formal reassessment rather than remaining a proposal that could still be argued down, would attach interest and penalties that grew the longer it sat unresolved. Every day the family spent unable to focus on the letter was a day working against them, and the clock did not care what else was happening in their lives.

The problem

Four days after the letter finally got Bogdan's full attention, their father, who had lived nearby and helped out informally with property maintenance, passed away suddenly. Carlos and Ines, who between them held most of the operational knowledge and the underlying booking records, were consumed for the following two weeks with funeral arrangements, out-of-town relatives, and the ordinary weight of grief, none of which left room for pulling a year of occupancy data together on a government deadline. Bogdan himself was trying to hold his physiotherapy practice together while also supporting his siblings, and the audit letter, understandably, kept sliding to the bottom of everyone's list of urgent things.

The substance of the problem, once anyone could turn to it, was the sampling method itself. Auditing a small window of activity and projecting the rate found in that window across an entire year is a recognized audit technique, and one that can be entirely reasonable when applied carefully, but it is only reliable when the sampled period is actually representative of the broader pattern it is meant to stand in for. A two-week stretch during the single busiest fortnight of a seasonal short-term rental business is close to the opposite of representative. Occupancy during that window regularly ran near capacity, with premium weekend rates and back-to-back bookings that left little room between guests. The rest of the year looked nothing like it, with stretches of low winter occupancy, discounted midweek rates, and entire weeks with no bookings at all, particularly through the colder months when the properties sometimes sat empty for days at a stretch.

Applying the peak-season underreporting rate to every week of the year assumed a level of activity, and a level of income, that simply did not exist for most of the calendar. If the projection stood, Bogdan would be taxed as though every week had produced peak-season revenue, a result that had no relationship to how the properties actually performed and that no amount of after-the-fact explanation would fix once a formal reassessment locked the number in and shifted the burden onto the family to unwind it.

Compounding the difficulty, the underlying booking and payment records needed to demonstrate the true seasonal pattern were split across Carlos's accounting files and a booking platform dashboard that only Ines knew how to navigate properly, and both of them were, understandably, unavailable to walk anyone through it during the weeks immediately after their father's death. The deadline did not move, but the people who held the knowledge needed to meet it were, for a time, simply not able to engage, which left a narrowing window to do work that genuinely required their input.

What we did

1. Contacted the auditor within days to request a short extension, explaining the bereavement plainly. Auditors generally have discretion to grant reasonable extensions for genuine personal circumstances, and being direct about what had happened, rather than offering a vague or generic excuse, secured an additional three weeks without argument, which took the immediate pressure off the family during the worst of it and bought time to do the work properly.
2. Began building the sampling challenge using only the records Carlos could access remotely in the meantime. Rather than waiting for the family to be fully available before starting anything, we started pulling twelve months of bank deposit records tied to the rental accounts, which gave a rough seasonal shape to work from even before the more detailed platform data was in hand, so no time was lost while everyone was otherwise occupied.
3. Once Ines was able to engage again, extracted a full year of platform booking and rate data covering every property. This meant exporting occupancy calendars, nightly rates, and booking volumes for the entire twelve-month period, not just the audited fortnight, to build a complete and defensible picture of how dramatically the business's activity actually varied from one season to the next.
4. Built a week-by-week occupancy and revenue chart comparing the audited two weeks against the rest of the year. Laying the data out visually made the seasonal swing immediately obvious to anyone looking at it, showing occupancy in the audited window running roughly three to four times higher than the year's average, which is the kind of comparison an auditor can absorb quickly without wading through pages of raw spreadsheets first.
5. Prepared a formal written challenge to the sampling methodology itself. Rather than disputing the specific dollar figures found in the audited fortnight, which were largely accurate on their own terms, we argued that projecting that period's rate across the full year was methodologically unsound given the documented seasonal pattern, and proposed that any adjustment instead be calculated using the actual full-year data now assembled.
6. Reconciled the small genuine reporting gap found in the audited window against the full-year records. The audit had correctly identified a modest discrepancy in how a handful of bookings were recorded for that fortnight, and rather than contest that finding on principle, we isolated it, corrected it, and showed through the surrounding records that it was an isolated data entry issue rather than evidence of a broader pattern running through the whole year.
7. Requested a meeting with the auditor to walk through the full-year data before any formal reassessment issued. Presenting the complete picture directly, rather than only in writing, let the auditor ask questions in real time, test the seasonal comparison against their own records, and see the swing for themselves, which moved the file toward resolution considerably faster than a written exchange alone would have.

The outcome

After reviewing the full-year data and the corrected figures for the audited fortnight, the auditor agreed the original projection could not stand and withdrew the proposed reassessment before it was ever formally issued. In its place, the small reporting discrepancy actually found in the audited two weeks was corrected on Bogdan's return, resulting in a modest adjustment in the low thousands of dollars, a fraction of what the original two-hundred-and-fifty-to-three-hundred-thousand-dollar projection had proposed only weeks earlier.

Because the matter never advanced past the proposal stage, no formal reassessment, and no wave of accruing interest and penalties on a six-figure number, ever attached to the file. The family avoided what would otherwise have been a lengthy and expensive objection and possible appeal process layered on top of an already difficult few months, and Bogdan avoided the disruption a formal dispute of that size would have caused to both his rental income and his physiotherapy practice.

Carlos rebuilt the bookkeeping system afterward so that full-year occupancy and rate summaries are generated automatically each season rather than needing to be reconstructed from scratch under pressure, and the small recording discrepancy that had triggered the audit in the first place has not recurred since the correction. Bogdan, Carlos, and Ines still run the rental business together, and they now keep a standing folder of seasonal comparison data on hand at all times, not because they expect another audit, but because rebuilding that picture once, under the worst possible circumstances, was enough for all three of them.

What you can learn from this