Methodology · version 1.0 · scores as of 8 Sep 2026
How the Pinly RERA Compliance grade is calculated
The Pinly RERA Compliance grade rates how closely a developer follows Karnataka RERA's rules, using nothing but the developer's own filings on the Karnataka RERA public register. It is a score out of 100 made of 5 checks, banded A to E. As of 8 Sep 2026, 295 builders are graded and 3,303 more have scorecards without a grade. Everything below — weights, bands, averages — is generated from the same code that computes the scores.
Principles
- Own filings only. Only what the developer itself filed with Karnataka RERA is used — no reviews, opinions, news, price data or anything from outside the register.
- Shares, not counts. Every check is a share of the developer's registrations, so a large portfolio is not penalised for being large.
- Small portfolios are blended. Each share is pulled toward the state average by 2 pseudo-registrations, so one or two projects cannot decide a grade.
- Not applicable is not zero. A check with nothing to measure (say, no project due yet) is left out and the remaining weights are re-scaled.
- Dated and counted. Every score carries its as-of date (8 Sep 2026) and the counts behind each check.
- People are not rated publicly. Individuals, proprietors, HUFs and land owners who register a project are scored internally but never named or ranked on public pages (1,602 promoters as of 8 Sep 2026).
The 5 checks
“State average” is the compliant share across every registration in Karnataka — the value small portfolios are blended toward.
| Check | Weight | Counts against the developer when… | Measured over | State average |
|---|---|---|---|---|
| Delivers on the promised date | 30% | A registration counts against the developer when its proposed completion date (as filed with RERA) has passed and the developer's own quarterly update after that date still shows it below 100% complete. | registrations due between 1 Oct 2025 (when quarterly updates start on the register) and the snapshot date, that filed an update after their due date | 75% |
| Withdrawals in line with collections | 20% | RERA requires 70% of money collected from buyers to go into the project account, withdrawn in step with construction. A registration counts against the developer when a quarterly filing shows more withdrawn from that account than was collected in the quarter; a borderline quarter counts half. This is a signal from the filing, not a finding that the law was broken. | registrations with at least one quarterly money filing | 54% |
| Files quarterly progress updates | 25% | Running projects must file a quarterly progress report. A registration counts against the developer when it was approved before October 2025, its promised completion date is still ahead or recently passed, and it shows no quarterly update on the register. | approved registrations still inside their promised timeline on 1 Oct 2025 | 55% |
| Reports progress consistently | 10% | A registration counts against the developer when its filed completion % went backwards between quarters, or stalled while the deadline approached. | registrations with two or more quarterly updates | 81% |
| Registration in good standing | 15% | A registration counts against the developer when it was revoked or withdrawn, or its own filed dates contradict each other (completion before approval). | all registrations | 92% |
The formula
For each check i, with n registrations measured and a counting against the developer:
blended_i = (n − a + 2 × state_average_i) / (n + 2)
score = 100 × Σ (weight_i / Σ live weights) × blended_i
(sum over checks with n > 0 only)The score is rounded to one decimal. A borderline escrow quarter counts as half a registration against the developer.
Grade bands
| Grade | Score | Meaning | Builders (8 Sep 2026) |
|---|---|---|---|
| A | 85–100 | Consistently compliant | 2 |
| B | 70–84.9 | Mostly compliant | 177 |
| C | 55–69.9 | Mixed record | 115 |
| D | 40–54.9 | Frequent lapses | 0 |
| E | 0–39.9 | Poor record | 1 |
Who is graded
A developer is graded once it has at least 3 registrations and at least 3 of the 5 checks can be measured. Below that, its scorecard shows the checks that can be measured and says “Not enough filings to grade yet” — never a letter.
Rank is among graded organisations only (1 = highest score; ties go to the larger portfolio). Percentile compares against every developer scored, including those not shown publicly.
Worked example: Sobha Limited
Computed live from the same data as its scorecard.
| Check | n | Against | Own share | Blended | Weight used | Points |
|---|---|---|---|---|---|---|
| Delivers on the promised date | 11 | 0 | 100% | 96% | 30.0% | 28.8 |
| Withdrawals in line with collections | 60 | 42.5 | 29% | 30% | 20.0% | 6.0 |
| Files quarterly progress updates | 23 | 1 | 96% | 92% | 25.0% | 23.1 |
| Reports progress consistently | 37 | 9 | 76% | 76% | 10.0% | 7.6 |
| Registration in good standing | 121 | 8 | 93% | 93% | 15.0% | 14.0 |
| Score | 79.5 → B | |||||
Sobha Limited's scorecard shows the counts behind each row in words.
Data source
The Karnataka RERA public register at https://rera.karnataka.gov.in — registrations, promoter details, approval and completion dates, and the quarterly progress and money filings developers submit. Pinly's copy is a snapshot taken 18 Sep 2026 covering 8,919 registrations and 5,200 promoters; scores are computed as of 8 Sep 2026.
Promoter names are grouped where the register spells the same organisation several ways. Public pages show registration identity, status, dates and counts only — rupee amounts, financials and document contents are part of Pinly's paid due-diligence report, not the rating.
Limitations
- Quarterly updates are on the register only from 1 Oct 2025. A project whose promised date fell earlier can't be told apart from a late one, so it isn't judged on delivery.
- A completion percentage is the developer's own figure. Pinly does not inspect sites; a grade reflects what was filed.
- The escrow check is a signal from the filing (withdrawn more than collected in a quarter), not a finding that the law was broken — there can be legitimate reasons.
- Revocations and withdrawals are counted as filed; the register doesn't always say why.
- Construction quality, customer service, title disputes and court cases are outside the register and outside the grade.
- Karnataka only, for now. A developer's projects in other states are not counted.
Correction policy
Anyone — buyers, developers, journalists — can ask for a correction from the form on any builder’s scorecard. We check the request against the register, fix data errors at the next refresh, and record any change to the method itself below. We do not change a grade on request when the filings support it, and we never accept payment to change one.
Citing these ratings
The grades and this method are free to quote, with attribution: “Pinly RERA Compliance rating, as of 8 Sep 2026” and a link to the developer's scorecard or to https://pinly9.khosha.tech/rera/methodology. A machine-readable summary is at /llms.txt.
Changelog
- v1.0 · 21 Oct 2026 — First public version: five components, A–E bands, blending toward the state average, quarterly-update era from 1 Oct 2025.