How our rent data works.

Sources

What counts as a rent report

The map carries three kinds of pins. A “what I pay” pin is a first-hand claim: this is the rent I actually pay, for this home, right now. A seeker pin says “I'm looking around here, and my budget is X”. A next-home pin says “a place opened up near here, asking around X”.

Only “what I pay” pins enter our rent aggregates — medians, ranges, heat overlays. Seeker budgets and asking prices are marketplace signals, not rents, so they are excluded from every rent statistic we publish. Rent reports are anonymous by default; a signed-in member can choose to attribute their pin, but never has to.

Statistics

How medians are calculated

For each locality, we compute percentile statistics directly in the database over the rent reports in that area: the 25th percentile, the median (the middle value — half of renters pay below it, half above), and the 75th percentile. Together they describe the typical band, not just a single point.

Locality stats use a trailing 180-day window, so a figure always reflects the last six months of reports rather than stale history. Statistics are recomputed as new pins arrive and cached briefly (a few minutes) to keep the map fast under load.

We show the median rather than the average because a handful of outlier rents — a typo, a luxury penthouse, a landlord's wish — can drag an average far from what anyone actually pays. The median shrugs those off. Wherever a figure is shown, the number of reports behind it is shown beside it, so you can judge how firm it is.

Trust

How outliers and bad data are handled

Statistical checks. Every new rent report is scored against recent reports in the same city (and the same configuration, when given). A rent that lands more than three standard deviations from the peer mean is flagged as an outlier. We never flag on thin evidence — with fewer than five peer reports to compare against, no rent is called an outlier.

Flagged reports are excluded from the public map and from aggregates by default, and queued for staff review — they are not deleted. Staff can suppress (hide) or restore any pin, and every moderation action is recorded in an audit log.

Community reports. Every pin has a Report button. A pin is auto-hidden from the public map only after it accumulates five distinct reports — and reports never influence outlier scoring, so a small group can't weaponize either mechanism against a pin they dislike. Suppressed pins stay reviewable by staff and can be restored.

Buildings

How buildings are grouped

Rent reports can name a society or building. Before grouping reports by building, we normalize those names: case and punctuation are ignored, suffix words such as Apartments, Society, or Towers are dropped, and the wings, towers, and blocks of one society are merged into a single building. “Prestige Towers Wing B” and “prestige tower, blk-b” are the same building.

A building gets its own median only when enough reports name it — otherwise no building figure is shown at all, rather than a guess built on one or two pins. And as everywhere else, the report count is displayed next to every building figure.

Freshness

Freshness labels

Every figure carries a freshness label based on when its underlying reports were submitted:

• Fresh — includes reports from the last 180 days.
• Recent — the newest reports are within 18 months.
• Older — the newest reports are more than 18 months old.

We label rather than hide older data: a three-year-old rent is still a useful anchor, as long as you can see that it is three years old. Locality medians on city pages always use the 180-day window; older reports mainly matter where a locality is thin, and the label tells you so.

Deposits

Deposits

The deposit is an optional field on a rent report — not everyone knows or wants to share it. We show deposit figures (typically as months of rent) only where enough reports in an area actually include one; below that, we show nothing rather than extrapolate. Deposit amounts are stored in paise, exactly like rents, so no rounding ever creeps into the math.

Privacy

Privacy and safety

Pin coordinates are rounded to three decimal places — roughly 110 metres — before they are stored. That is precise enough to say “this part of Indiranagar”, and never precise enough to point at someone's window. Pins are anonymous by default: no name, no account, no phone number required to drop one.

Browsing rents needs no login at all. When a member opts into direct contact, their details are stripped from every public map payload — contact happens only through our connect flow, never by exposing personal information on the map.

Corrections

Spot a wrong figure?

Use the Report button on the pin itself and tell us what looks wrong. Distinct reports stack up — five is enough to auto-hide a pin pending review, and staff can suppress or restore any pin at any time. If you dropped the pin yourself, you can edit your own report directly (a corrected rent is re-scored against outliers automatically).

For anything else — a figure that looks systematically off, or a moderation decision you disagree with — reach us through Help & FAQ or the grievance officer.