Unit 5 · Maximizing Performance Through Iteration · Lesson 5.8
Diagnosing a listing with no bookings
A four bedroom listing in Big Bear Lake took one booking in two weeks and had nothing on the books beyond it. The data showed why: one flat rate across all three pricing steps, sitting near the top of the comp set while midweek market occupancy ran under 20%. Here is the full diagnosis and the rates that replaced it.
The full lesson text below is an edited transcript of the video, published 2026-08-24. The complete course is free at the playbook.
The listing and the symptom
The property is a four bedroom listing in Big Bear Lake, running the three step structure: a hedge step covering the next zero to four nights, a control step from five to 42 nights out, and a test step from 43 to 365 nights out. The cost inputs behind the numbers are cleaning at $290, tax at 13.5%, and a channel deduction of 15.5%, which is what turns a listed rate into what the owner actually keeps.
One check comes before any of the pricing work. The listing’s conversion metrics were healthy: the search to listing click rate was good, and the listing to booking conversion was good. That matters, because a rate change only fixes a rate problem. If guests are seeing the listing and not clicking, or clicking and not booking, the constraint sits with the listing itself and repricing will not move it.
The symptom is simple to state. The strategy had been completed and implemented, and across the two week iteration that followed, the listing took exactly one booking. Past that booking the calendar was empty. Nothing in the control window, nothing in the test window, nothing at all.
Where the bookings landed
The iteration under review ran two weeks, from the 11th of November to the 25th. Inside that window the listing took one booking, for three nights, and it fell inside the hedge step. The control step took nothing. The test step took nothing. That is the area of concern, and it is not subtle: everything beyond four days out was dead.
The second number explains the first. The average booked rate came out at $404 against a listed midweek rate of $575, roughly 30% below the asking rate. So the single booking was not the strategy working. It was a last minute discount, taken inside four days of arrival, to avoid an empty night.
The third observation names the fault. Across the hedge, the control and the test there was no variability in the rates whatsoever. Every step carried the same number. The three steps had been written down but never actually implemented, so the listing was asking one rate for a night four days away and the identical rate for a night eleven months away. That flat rate, sitting where it was sitting, is why nothing past the hedge was selling.
Reading the hedge window
The hedge step covered the 27th of November through the 1st of December, five nights with no Friday or Saturday in them. That is why the weekend figures for the hedge came back blank rather than zero. It happens regularly with a short hedge step, and it is not a data error.
Forward occupancy on the listing was 0%. Market occupancy across the same nights was 18.23%, very low in absolute terms, and the 90 day occupancy trend was down 3%, a slight softening rather than a collapse. The market was quiet and the listing was quieter.
Price positioning is where it falls apart. The listed midweek rate of $575 sat between the 50th and the 75th percentile of the comp set, in a band running from roughly $395 to $593. Asking in the upper half of the set is a defensible position into a strong market. Into a market running under 20% occupancy it is not a position at all, because those nights are barely selling at any rate, let alone an ambitious one.
The reset follows from the occupancy number. With midweek market occupancy under 20%, the midweek hedge rate drops below the 25th percentile, and deliberately so: $225. The weekend hedge rate was left open at this point, because there was no weekend inside the hedge window to read it from.
Reading the control window
The control step ran from the 2nd of December to the 8th of January. Listing occupancy across it was 0%. Market expected occupancy was about 46% midweek and about 68% at weekends, and weekends in this market fill considerably better than midweek nights do.
Taken at face value, 68% weekend occupancy would justify leaving the weekend rate exactly where it was, and 46% midweek would justify a conservative midweek number. Neither conclusion is safe, because that window swallows the holiday peak. A peak lifts the average for the whole window, so the control numbers were describing the holidays and the quiet weeks on either side as though they were the same market.
So the control rates were left unset at this stage, and the decision moved to the recent record read month by month, where the peak weeks can be seen separately and taken out.
Reading the test window
The test step covers 43 to 365 nights out, and current market occupancy that far ahead always reads close to nothing, because those nights have not been sold yet. The number worth reading is the same period a year earlier, and it was low as well.
That gave a midweek test rate just under the 25th percentile at $325, which is $100 above the $225 hedge rate set a moment earlier. That is the shape the three steps are supposed to have: nights far out carry more room than nights about to expire.
Weekend occupancy that far ahead ran around 50%, giving a weekend test rate of $650, slightly above the $629 the comp set pointed to. Going slightly over is the point of a test step. It is the part of the calendar where you can afford to ask for more and find out whether the market pays it.
Stripping the peak out
Two questions were still open: the weekend hedge rate, and both control rates. Both needed the peak weeks taken out of the picture, which means reading the recent record month by month instead of trusting a window average.
November first, for the weekend hedge rate. The weekend rates at the 25th percentile through that month ran $289, $299, $319, $320, $320, $336 and $350, with market occupancy on those weekends between 38% and 50%. Occupancy in the 40% to 50% band puts you between the 25th and the 50th percentile: not far above the bottom quartile, but not sitting on it either. That gave a weekend hedge rate of $325.
December next, for the control. Midweek rates at the 25th percentile through the early part of the month came in at $299, $350, $320 and $317, and from around the 16th they start climbing steeply into the holidays. Averaging only the midweek nights before that climb gave about $335 at the 25th percentile, against average market occupancy of about 13% across the same nights.
13% is low enough that the 25th percentile is still an ambitious ask, so the midweek control rate went under it, to $295.
December weekends outside the peak told a different story: average occupancy around 50%, which supports pricing near the 50th percentile rather than the 25th. The weekend average across those nights was about $700, though one of them, the 16th and the 17th, sat about $200 above the other two and was dragging the average up. The weekend control rate was set just under that average, at $695.
That created a problem worth catching. $695 for a control weekend sat above the $650 already pencilled in for a test weekend, which inverts the structure: nights further out should carry the higher number, not the lower one. The test weekend rate was raised by about 10% over the control, to $765.
The new rate set
The rates that came out of the diagnosis: hedge at $225 midweek and $325 at weekends, control at $295 and $695, test at $325 and $765.
Run the shape check before changing anything. Midweek climbs $225, $295, $325 as the nights move further out. Weekends climb $325, $695, $765. Every step sits above the one closer in, and every number is anchored to a percentile position that matches the occupancy actually expected in that window. A single rate of $575 across all three steps did neither of those things.
Then record the next iteration before anything moves: a fresh two week window starting on the day the new rates go live, the step lengths unchanged at 4 and 42, and the new price points written against them. The rates themselves then go into whichever dynamic pricing tool the listing runs on, such as PriceLabs.
Note what actually happened here. The rates came down hard, and that was the deliberate response to a market running at 13% to 18% occupancy midweek while the listing asked upper quartile money. The next read is what tests the diagnosis. If the reset is right, bookings start appearing in the control and test windows instead of arriving late and cheap inside the hedge. If they do not, the next iteration says so, and the rates move again.