Could an inventory rise itself change the apparent condition mix? More choice would let buyers favour presentation without requiring any broad loss of demand.
I’d record both new listings and completed transactions by segment. Prices alone cannot show whether a gap widened during active trading or because very few deals completed.
Another caveat: percentages can conceal the cash difference across price bands. Compare like-priced properties before interpreting a common percentage as equal buyer resistance.
True, though adding too many subdivisions will leave tiny groups. Start with broad property type, condition and neighbourhood, then disclose when a cell is too sparse to interpret.
A rolling view may help once the raw monthly table exists. It can soften calendar noise, but it should not replace the underlying October 2025 observations.
Do not smooth away the very shift being investigated, though. Show raw periods beside any rolling measure so readers can see whether the result depends on smoothing.
What would count as evidence of selectivity here? Longer times only for properties needing work, larger reductions for them, or lower completed prices relative to their own initial asks?
Probably a combination. Longer marketing alone might reflect stubborn pricing; a consistent condition-related difference in reductions and outcomes would make the selectivity interpretation stronger.
There is a seller side too. Owners of renovated properties may price more realistically, while owners pricing future potential may hold out. Buyer behaviour is only half the negotiation.
Agreed. That is why “buyer selectivity” should remain a hypothesis. The observed split could arise from buyers, sellers, listing composition, or all three.
I would ask contributors not to post only successful examples. Unsold, withdrawn and reduced listings are essential if this is meant to describe inventory rather than completed deals alone.
A workable template now seems clear: neighbourhood, type, condition category, first listing date and ask, reductions, status date, completion date and price where available, plus relisting notes.
Add the date the information was captured. Listing status can change, so two people looking at the same property at different times may otherwise appear to contradict each other.
Good addition. I would also leave unknown fields blank. Filling them from memory or assumptions would create false precision around an already small-looking difference.
Is there any reason to treat serviced status as presentation? A serviced apartment can still need work. Those should be separate fields rather than one combined category.
Exactly. Property use or operating model is not the same variable as physical condition. Combining them makes the opening comparison harder to interpret.
That weakens the original wording somewhat. The fast group may be defined by two characteristics—serviced and well-presented—while the slow group is defined only by condition.
The clean comparison would be well-presented versus work-needed within serviced apartments, then the same comparison within other property types, provided each group has enough observations.