Methodology
How we rank care homes.
The Best Match algorithm is published in full so families and providers can audit it. No provider can pay to rank higher. There are no promoted slots. Two homes with identical signals receive identical positions.
The short version
Best Match combines two signals: how well a home fits your stated needs (65%) and the home's Quality Score (35%). Family fit weighs more because a perfectly-fitting average home is more useful to most families than an outstanding home that doesn't suit them.
Family fit (65% of the score)
The fit score combines five inputs you provide:
- Care type (35%) — residential, nursing, dementia, respite, palliative, specialist. Multi-select.
- Weekly budget (25%) — homes within your stated budget score the full weight; homes above receive a proportional reduction.
- Location (30%) — region match (20%) and town match (10%). Distance-based scoring is on the roadmap; corridor scope keeps this simple for now.
- Specialisms (15%) — e.g. dementia-specialist, palliative, cultural or religious provision. Partial credit for partial overlap.
- Availability (10%) — only counts if you tick "must have a confirmed vacancy".
Quality Score (35% of the score)
The Quality Score is a 0–100 composite. It's computed at read time from auditable inputs:
- CQC rating (40%) — Outstanding=1.0, Good=0.85, Requires Improvement=0.5, Inadequate=0.2.
- Verified reviews (30%) — average rating, weighted by verified review count up to 20.
- Provider transparency (15%) — fee disclosure, real-time availability, profile completeness.
- Operational responsiveness (15%) — average response time to qualified enquiries.
Read the dedicated Quality Score page for worked examples.
What is not an input
The following are never used in Best Match ranking:
- Subscription tier paid by the provider (Foundation, Connect, Performance).
- Number of lead credits the provider has purchased.
- Whether the home is owned or operated by an entity connected to Akinn.
- How recently the provider last spoke to us.
- Whether the provider has paid for any other Akinn service.
Ties and determinism
When two homes produce the same combined score, the Quality Score breaks the tie. If those are also equal, homes are ordered alphabetically by name. Given identical inputs, the ranking is identical — there is no randomisation, no rotation, and no personalisation based on data you didn't provide.
Audit and changes
The weights and algorithm above are codified in our codebase at lib/ranking.ts and lib/quality-score.ts. Material changes (weight adjustments, new inputs, removed inputs) will be announced on this page with the date the change took effect.
See also our governance and disclosures page for the LWT Healthcare relationship and how it is handled.