Legal

01The rules

1. Minimum cohort. A benchmark compares a customer against a group of other schemes. The group must contain a minimum number of schemes besides the recipient’s own, run by a minimum number of operators unconnected to the recipient.
2. Concentration. Excluding the recipient’s own schemes, no single scheme and no single operator may account for more than a set share of the group.
3. Isolation. The reports issued to one recipient and its connected companies, taken together over the previous 12 months, must not allow a single scheme’s figures to be worked out by subtraction. A group changes only by adding or removing several schemes at once.
4. Small cells. A figure that represents fewer than a set number of individuals is suppressed or merged into “other”.
5. Output. A benchmark shows the recipient’s own figures beside a single city-level aggregate, split no further than broad room type and letting year. No row-level data, building name, operator name or tenant attribute is released.
6. Banding. Aggregate figures are rounded or shown as a band, not to the unit: percentages to a set number of points, counts as a range or a floor.
7. Unranked lists. Where a benchmark lists categories, any category below the small-cell threshold is listed without its rank or share.
8. Timing. Rent and occupancy for a live letting cycle are released no more often than a set interval and with a delay.
9. Failure. A group that stops meeting this standard is suppressed until it does.

We do not attempt to re-identify any individual, scheme or operator from Aggregated Data. If we find that any Aggregated Data could identify one, we treat it as personal data, and delete or re-anonymise it without undue delay.