How we rank schools
Almost every published ranking of UK schools tells you where a school sits but not why. The Houseroom Index 2026 publishes the formula in full so parents — and schools — can argue with it.
Nine pillars
Each school is scored 0–100 on nine pillars. The pillars are then weighted to produce a single composite score. The weights sum to 100 and we publish them.
| Pillar | Weight | What it captures |
|---|---|---|
| Academic outcomes | 22% | Raw A-level A* %, GCSE 9-7 %, Oxbridge offer rate, Russell Group offer rate. Composed from raw metrics — not from any external rank. |
| Destinations & outcomes | 17% | Top-tier university placement (Oxbridge, Russell Group, Ivy+), notable alumni density, prep-school leaver offers. |
| Future-readiness | 12% | AI / CS / coding / design tech / entrepreneurship in the curriculum — scored as a 5-tier rubric from the school's published curriculum. |
| Pastoral & wellbeing | 12% | ISI inspection wellbeing rating + parent reviews on pastoral care. |
| Pupil flourishing | 8% | Validated wellbeing items from parent reviews — peer connectedness, agency, identity, prosocial behaviour. Rolling out as the review corpus grows. |
| Value-add | 5% | Outcomes adjusted for selectivity. Real entry-vs-exit data is being collected; weight kept low at 5% until then. |
| Access & breadth | 10% | Bursary depth (means-tested), fully-funded place programmes, breadth of co-curricular signals. |
| Parent sentiment | 9% | Verified Houseroom parent reviews — overall satisfaction, with sample-size confidence. |
| External consensus | 5% | Sunday Times Parent Power + Tatler + Good Schools Guide consensus (capped at 5%) — sanity check, not anchor. |
| Total | 100% | Composite Houseroom Index score |
Why these weights
Academic outcomes (22%) carry the largest single weight because exam results remain the most comparable, audited signal of teaching strength. We compose this pillar from raw metrics — A-level A*%, GCSE 9-7%, Oxbridge offer rate, Russell Group offer rate — never from an external rank. Single-signal schools are blended toward neutral so a thin number cannot inflate a rank.
Destinations & outcomes (17%) rewards real placement evidence — top-100 universities, Oxbridge, Russell Group, US Ivy+, vocational/arts conservatoires — and where prep schools place leavers.
Future-readiness (12%) is the pillar nobody else publishes. A 5-tier rubric scores how seriously a school treats AI literacy, computer science, robotics, design tech and entrepreneurship. As of 2026, this is the new water-line for parents planning for a 2030+ workforce.
Pastoral (12%) today reads the ISI inspection wellbeing signal and pastoral language across school policies and the editorial review. As the verified parent-review corpus matures it feeds into this pillar with a low cap to prevent thin samples skewing the score. We give pastoral care a meaningful share of the composite so it isn’t lost next to exam outcomes.
Pupil flourishing (8%) is scaffolded for validated wellbeing items (peer connectedness, agency, identity, prosocial behaviour). Currently a placeholder until those items are added to parent reviews; the weight is reserved so flourishing is part of the score from day one.
Value-add (5%) attempts to credit non-selective schools that produce strong outcomes despite a less filtered intake. Today this uses a selectivity proxy; weight is deliberately low until real entry-vs-exit data is collected.
Access & breadth (10%) scores bursary depth, fully-funded place programmes, and breadth of co-curricular provision. Wealth-only access lowers this score.
Parent sentiment (9%) scaffolded for verified Houseroom parent reviews once volume passes the per-school threshold (currently 5 reviews). Schools below that threshold fall back to the neutral mean rather than risk one or two reviews swinging the rank. Weight is capped low even at scale, to prevent gaming.
External consensus (5%) deliberately small. Sunday Times Parent Power, Tatler Schools Guide and the Good Schools Guide are valuable complementary references and useful as cold-start signals. Houseroom is intended to add a distinct, multi-pillar view alongside them, so their consensus signal is blended in at 5% — it informs, but doesn’t dominate.
Data sources
- Public exam results & destinations — school websites, prospectuses, ISC bulletins, news releases.
- ISI inspection reports — published judgements on academic and pastoral quality.
- Curriculum & ethos — school websites (KS3/4/5 pages, computing, AI policy, design tech, careers).
- Verified parent reviews — Houseroom members only, moderated, signed. Per-school threshold of 5 reviews before the signal counts toward a rank.
- External rankings — Sunday Times Parent Power 2026, Tatler Schools Guide, Good Schools Guide (capped at 5% combined).
- School-supplied data — schools can claim their listing to fill gaps. Verified submissions earn a badge but don't inflate the score.
Confidence
Every school carries a data-confidence percentage — the share of the pillars that apply to that school (weighted) where we have real evidence rather than an imputed mean. Pillars we can't yet measure for anyone are left out of the denominator, and prep & junior schools aren't marked down for senior-only pillars like A-level results or university destinations — they're scored only on what applies to them. We show the confidence on every ranking row so you can weigh it.
Schools wanting to improve their data confidence (and accuracy) can submit verified data through the school portal.
What this index will not do
- · Take payment from schools to alter rankings.
- · Hide its formula. The weights are listed above and in code.
- · Reduce a school to one number without context. Every score is broken down into its nine pillars on the ranking row.
- · Replace a visit. The Index helps shortlist; an open day, lesson observation and pastoral conversation does the rest.
Changelog
Every change to the formula bumps the version. Each ranking row carries the version it was scored under, so older comparisons stay traceable.
- Confidence floor + sparse-pillar penalty tightened so a school cannot rank top-20 with confidence below ~55%. Every ranking row continues to show its confidence score.
- Parent voice pillar gated on review volume — schools with fewer than 5 verified reviews fall back to the neutral mean, preventing a single review from swinging the rank.
- Three sub-indices added — Academic, Whole-Child, Future-Ready. Each is the same data re-weighted to a 0-100 scale on a coherent dimension, sortable from the page header.
- Thin-signal penalty on academic and destinations pillars — schools with fewer than two independent sub-signals are blended toward neutral so a single fuzzy stat can no longer inflate a rank.
- Destinations now uses top-2 max-pooling rather than a flat max, so a single very strong signal isn't diluted by a single weak one.
- Godolphin and Latymer deduplicated — two near-identical rows merged to one canonical entry.
- Academic pillar de-tautologised — now uses raw A-level A*%, GCSE 9-7%, Oxbridge offer rate, Russell Group offer rate. No longer derived from any external rank.
- Future-readiness becomes a 5-tier rubric (0–4) classified from each school's curriculum text — replaces the v1 keyword scan.
- New pillar: Pupil flourishing (8%) — scaffolding for validated wellbeing items in parent reviews.
- Value-add weight cut 10 → 5 until real entry-vs-exit data is collected.
- Weights rebalanced: Destinations 15→17, Academic 25→22, Future-readiness 15→12, Parent 8→9.
- Initial release. Nine pillars; academic pillar approximated from external Sunday Times Parent Power rank as cold-start.
See the index
Every school's pillar breakdown is shown on the ranking page.
Browse the Houseroom Index 2026