Occupancy at 61 percent but gross revenue up 11 percent year over year and I don't know how to feel about that

Your CPA is mostly right, and the number that actually settles it is net operating income, not gross revenue and not occupancy.

Here is the mechanism that explains what you are watching. At $109 average with lower occupancy, your cleaning costs, supply restocking, and platform fees per-night-booked are a smaller drag on total revenue than they were at $94 with more nights. Each additional occupied night carries real variable cost. Fewer nights at higher rates can improve margin even when gross revenue only moves modestly. So the 61 percent unit may actually be outperforming the 68 percent unit on a per-dollar-deployed basis, but you cannot know that without running the variable cost per booking side by side.

The $316 February question is worth taking seriously as a calibration exercise, not as regret. If your variable cost per stay (cleaning, consumables, fees) on a four-night booking at $79 is, say, $80 to $100, you were looking at $216 to $236 in net contribution, not $316. Whether passing on it was correct depends on whether accepting it would have reset your pricing signal in PriceLabs or attracted a guest profile inconsistent with the property. Those are real considerations, not rationalizations.

The risk your post does not name is demand mix. Emory-adjacent, between-lease demand is relatively inelastic on timing but price-sensitive. If your rate compression keeps the medical-visit and relocation renter, you are fine. If $109 starts pushing that renter toward extended-stay hotels or furnished apartments, occupancy drops further without a corresponding rate gain, and the math reverses. Monitoring your booking lead time and cancellation rate alongside occupancy would tell you whether demand is shifting or just pricing tighter.

What does your net per occupied night look like on each unit after variable costs?

Your CPA is mostly right, and the February gap is the clearest proof of it.

Four nights at $79 is $316 in gross revenue. But if PriceLabs held the floor because it calculated that accepting $79 would set a price anchor that dragged future rates down, then you may have traded $316 now for a higher average over the rest of the year. That is the actual logic behind dynamic pricing tools: they are not just reacting to today, they are protecting tomorrow's rate. Whether PriceLabs got that call right is something you can check by pulling your rate calendar and seeing what the unit booked for in the weeks after that gap.

Your CPA's point stands at the portfolio level too. Your 61-percent unit at $109 average produces more gross revenue per available night than your 68-percent unit at $97. Running a quick comparison: 61 percent of 365 nights is about 223 booked nights at $109, which is roughly $24,300 gross. The other unit at 68 percent gives you about 248 nights at $97, which is roughly $24,100. The "worse" unit is ahead on revenue.

The thing worth watching, and I'd want to confirm this with your actual numbers, is whether the 61-percent unit also has lower cleaning and turnover costs from fewer bookings. Fewer stays can mean meaningfully lower operating costs, which changes the net comparison more than the gross one does.

The February $316 is almost certainly a personality thing, as you said. Anchoring on a specific missed booking is a known behavioral pattern in hospitality. Worth noticing, worth setting aside.

Which unit has the higher net after cleaning and platform fees? That number probably settles how you should feel.

Your CPA is right but $316 at 83% of your floor isn't a personality problem, it's a $316 question worth answering once.