Two reasonable denominators, two different percentages
Imagine 100 active listings at a cutoff. Twenty-five have ever been reduced, while ten were reduced during the preceding 30 days. The ever-reduced share is 25%; the recent-reduction share is 10%. These are different measures, not conflicting results. Neither reveals a closed sale price.
| Measure | Illustrative result |
|---|---|
| Any reduction in observed history | 25 ÷ 100 = 25% |
| Reduction within the last 30 days | 10 ÷ 100 = 10% |
| Final sold-price discount | Cannot be calculated from this active set |
Handle relists and incomplete history
Specify whether each property or each listing ID is counted once. Retain the original price, dated changes and current ask, and record where history begins. A relisted property may appear newly priced if earlier records are unavailable. Exclude or separately label incomplete histories instead of assigning them “never reduced.” Publish the cutoff time and exact status filters.
Preserve the observation window when publishing a reduction rate
Store the listing identifier, first observed price, subsequent prices, status and observation dates. Decide how relists and duplicate records are handled before calculating the rate. Label the denominator as active listings, listings observed during the window or another defined cohort. A percentage without those decisions cannot reliably be compared with another publisher’s figure.
Sources & further reading
- CFPB determine a comfortable home budget checked 2026-09-26
Your next chapter
Let’s make it
a good move.
Bring your questions. Start a conversation with Gary’s team.