My sponsor on the Akron duplex said something offhand about STR analysis that I cannot get out of my head

Your sponsor named the right thing. The gap between gross nights times rate and actual cash in hand is where a lot of STR pro formas fall apart, and you have already found it yourself on this property, which puts you ahead of most people at this stage.

Your math looks directionally sound. Platform fees of around 15 percent on Airbnb are consistent with what hosts commonly report, though the exact percentage shifts depending on your pricing structure and whether you pass any fee to the guest. Cleaning reimbursements are frequently misread: guests pay a cleaning fee, but that money covers the cleaner, so it does not drop to your pocket. It reduces a cost rather than adding revenue, which is a meaningful distinction when you are building a pro forma. And the 8 to 10 percent reserve is reasonable for an STR specifically because the appliances, furnishings, and surfaces take far more wear than in a long-term rental.

So the $37k to $26k compression you are seeing is plausible, and the $26k is the number you would want to size debt service against, not the $37k.

On your actual question: I do not know what occupancy is running in Cleveland right now, and I would not want to guess. AirDNA and Rabbu both publish market-level occupancy data by city and property type, and they are the standard sources operators use to check assumptions against real booking data rather than estimates. A local STR property manager who is actively pricing units in that market is also worth a direct conversation, because they see live numbers.

One thing worth flagging from the strategy guide on this: Cleveland's regulatory environment for STR is worth checking before you model anything further. A single ordinance change can alter the whole picture.

What loan size were you working backward from on the Cleveland condo?

Your sponsor named the right thing. The occupied-versus-paid gap is where most STR underwriting quietly falls apart, and you've done the harder part by actually running it through.

The assumption doing the most work in your model is the 62 percent. Everything else you've adjusted is a function of gross revenue, so if that number is optimistic, the $26k shrinks again. Mid-tier Ohio markets are not tourism anchors, which means occupancy there is more sensitive to seasonality and local event calendars than a beach or mountain destination would be. A realistic STR in that category often runs 50 to 58 percent when you strip out the first few months of owner-optimism and account for shoulder-season softness. The difference between 62 and 54 percent at your ADR is roughly another $3k off the top before fees touch it.

The risk you did not model is regulatory. Akron and Cleveland are both municipalities that have been watching the housing-availability conversation closely. A single ordinance requiring a primary-residence permit, or restricting STR in non-owner-occupied units, eliminates the entire STR premium on a condo you'd be running as an investor rather than a resident. That scenario doesn't show up in a cash flow model until the ordinance passes. Confirm current city code and any pending zoning amendments in writing before you build a conversion case on that property. A local real estate attorney is the right person for that specific question.

Your reserve line at 8 to 10 percent is reasonable, though some operators in high-turnover urban markets run it higher because the appliance and furniture replacement cycle on an STR is roughly half that of a long-term rental unit.

What data source are you using for the 62 percent estimate? AirDNA, Rabbu, and direct comp pulls from active listings all produce different numbers, and the spread between them in a mid-tier market can be wide enough to swing your underwriting by several thousand dollars.

Cleveland two-bedroom STR is running closer to 48 to 54 percent right now, not 62.

The 62 percent figure is doing the most work in your model, and AirDNA's market data for Cleveland shows stabilized two-bedrooms tracking closer to 48 to 54 percent depending on neighborhood and season. Confirm that range against your specific zip code with AirDNA or Rabbu before you touch the loan sizing.