Scores
One explained score for every place, inside your product.
Turn a neighbourhood's signals into a single 0 to 100 score for the purpose your product serves, with our weights or your own. Every score opens into the dimensions behind it, with confidence on each, and carries the methodology version that produced it.
See it run
A score, and the reasons behind it. Weighted your way, explained every time.
One place is scored. The number resolves from its seven dimensions, each with its weight and its confidence, and carries the version that produced it.
A score your product can show and explain.
Every number shows what it is built from, how sure we are of each part, and the version of the method that produced it.
One number, and what it is built from
Every score opens into its dimensions, how much each counted and the reasoning behind each, so a customer can see why a place landed where it did.
Confidence on each part
Every dimension carries a confidence level from the data behind it, so thin data never hides behind a confident-looking number.
The same inputs, the same score
On the same data and the same methodology version, the same place always scores the same. No model sets the number.
Versioned, so you can explain it later
Every score carries the methodology version that produced it, and every change to the method is published, so a score a customer saw months ago can still be explained.
0 to 100 · four presets · your weights · your cohort
Tuned to the decision your product supports.
Four presets, or your own weights
Moving home, choosing a business location, weighing up an investment or a general read. Each preset weights the seven dimensions for that job. Send your own weights with a request instead, or save a mix for your organisation and every call uses it.
Ranked against your own cohort
By default a place is ranked against neighbourhoods of its kind across the nation. Give us a cohort of your own areas, at least 30, and transport, schools and amenities rank against them instead, so a score reads as 'among your places'. The response says which reference was used.
One explained number for every place, weighted for your product's purpose, and the same on the same data and version.
Get a sandbox key
The Scores workspace in the console
Where it goes in your product.
A score that sits on a listing, ranks a shortlist or flags what needs a look, with its reasoning one tap away.
One number your team and your customers read the same way.
On every listing
Show the place's score beside the listing, with the dimensions behind it a tap away.
Rank a shortlist
Score every place in a customer's list or a dataset and sort by it, with the same weights for all of them.
A pass mark
Set a threshold and let the score flag the places that need a closer look.
Against a customer's portfolio
Score a place against the customer's own cohort of areas, not the nation, when that is the comparison they care about.
Frequently asked
Scores, answered.
How is the score worked out?
Each of the seven categories is ranked among comparable neighbourhoods, weighted by the preset or by your weights, and combined into one number from 0 to 100. The response lists every dimension with its contribution, its confidence and the reasoning.
What are the four presets?
Moving home, choosing a business location, weighing up an investment, and a general research read. Each one weights the categories for that job.
Can I change what the score cares about?
Yes. Send your own weights with a single request, or save a mix for your organisation so every call scores the same way.
Does an AI decide the score?
No. The score is computed from the data underneath, the same way every time, and every change to the method is published.
Will the same place always get the same score?
Yes, on the same data and the same methodology version. When a publisher releases new data or we release a method change, the score can move, and the version on the response says which method produced it.
Can I see why a place scored the way it did?
Always. Every score opens into its dimensions, how much each counted, how confident we are in each and why.
How do I use it?
One API call returns the score and its breakdown. The MCP server exposes the same scoring to the AI features in your product. Or book a call and we run it on your own places.
One score your product can show and explain.
0 to 100, weighted for the purpose, the same on the same data and version, with the reasoning attached. Start in the free sandbox.