His math is sound and he clearly states his underlying assumptions. Using the recently published infographic, AirBnB revealed it did 5,000,000 nights in a 143

His math is sound and he clearly states his underlying assumptions. Using the recently published infographic, AirBnB revealed it did 5,000,000 nights in a 143 day period. Grossing this up by 365 days equals 12,750,000 nights projected for 2012 (5M / (143/365) = 12.8M). This calculation implicitly assumes that the rate of bookings does not increase for the rest of 2012, which is probably not true looking at their impressive growth rate. The author backs out a 19% month-over-month historical growth rate using other statistics published in the same infographic. Using this rate of growth, he derives a projected 17.6M nights for 2012 versus the 12.75M above. > with the length of each reservation increasing substantially from the year before The 17.6M projected nights makes no explicit assumption about the "length of each reservation." The 5 million nights published by AirBnB could have been 1 user booking a 5 million night stay (it wasn't) or 5 million users booking 1 night stays. It doesn't matter, the length of reservation is embedded in the statistics and not separately dealt with, so the author cannot be growing that assumption. > the post is assuming maximum commission on each reservation He does not use the maximum commission on each reservation, which we established would be 15%. Instead, he uses 13%. > Payments collected on behalf of other parties (i.e., the payment fee) are not treated as revenue under GAAP. I'm not going to argue GAAP accounting, but from a cash perspective, AirBnB is collecting 9 to 15% on each reservation. If a listing is for $100, the Guest pays somewhere between $106 and $112 per night (i.e., 6 to 12% markup) and the Host gets $97 ($100 less $3 or 3%). Therefore, AirBnB is generating revenue of $9 to $15 on every $100 in reservations.