The Unit File
How The Unit File works
Look up a place. Search a building, address, or management company.
Read what renters said. Each report is one renter's own account: what they paid, what went wrong, and whether they'd sign again. Reports come from renters and share their own experiences. Your name is never shown.
See the score once enough people weigh in. A Unit Score from 1 to 5 shows after 3 reports (marked limited), along with how many would sign again. From 5 reports it shows in full.
Add yours. Tell us what you paid and how it went. It helps the next person before they sign.
What the score means
Unit Score is out of 5 (5 is best). It averages renters' ratings (or tags) in six categories, plus the share who would sign again (all yes = 5, all no = 1) once 3 renters have answered. 3.6 or higher is good, below 3.0 is poor. A number shows after 3 reports (marked limited) and in full from 5.
Landlords can't pay to remove or bury them.
Check this place before you sign.
For the curious
Unit Score is out of 5 (5 is best). It averages renters' ratings (or tags) in six categories, plus the share who would sign again (all yes = 5, all no = 1) once 3 renters have answered. 3.6 or higher is good, below 3.0 is poor. A number shows after 3 reports (marked limited) and in full from 5.
Each category that has answers is averaged on the 1–5 star scale, with equal weights. A category with no data is left out — it is not scored as zero.
Stars count as given. Older files without 1–5 category ratings use the experience tags already on the report (responsive = 5, mixed = 3, unresponsive = 1, and the same pattern for maintenance, noise, safety, fees, and amenities). The result is shown with one decimal. Cleanliness is N/A until someone rates it.
Would Sign Again is a signal in that same average once three people have answered (the share who said yes, from 1 = all no to 5 = all yes). One or two answers never enter the number. Report volume is also a signal: it decides whether we publish a headline and whether we stamp it Limited data. More reports do not add quality points.
Most Reported lists only structured tags, valenced 1–5 categories, and issue topics, with counts — never themes guessed from comments.
Sample size is always shown. One or two reports cannot speak for a complex: we hide the headline score and the percentage until at least three people have filed. Five or more is the normal display. Word bands such as Good are printed only next to a number the formula earned.
Formula
- Map each of Management, Maintenance, Noise, Safety, Value, and Cleanliness from a 1–5 rating when present, otherwise from the matching experience tag. Cleanliness has no tag fallback. Maintenance days are used only when a report has neither a maintenance rating nor a maintenance tag.
- Stars count as given (1–5). Tags count as good = 5, mixed = 3, poor = 1. Each report's own star stamp is the average of its stars; a report with no stars shows 4 if they would sign again and 2 if not. That stamp is not used in Unit Score. It is used, rounded to a whole star, for the Rating breakdown, the highest and lowest city averages on company pages, and the review stars we share with search engines.
- Would Sign Again = the share who said yes, on the same 1–5 scale (all yes = 5, all no = 1). It joins the average only when at least three people answered.
- Unit Score = equal-weighted average of those signals that exist. Missing categories are omitted. Report volume does not add points. It gates the headline (hidden at 1–2, Limited data at 3–4, normal at 5+). Shown on the 1–5 star scale with one decimal.
When we show a number
- 0 reports: no score. File is empty.
- 1–2 reports: no headline Unit Score or Would Sign Again %.
- 3–4 reports: numbers labeled Limited data. No rankings.
- 5 or more: Unit Score shown with the report count.
- Bands (Good, Fair, …) only next to a published number.