Methodology | FosterScope

What does FosterScope measure?

FosterScope gives each local authority fostering service and regional fostering hub it measures 2 numbers. The Findability Score, from 0 to 100, measures how easy the service is to find on Google and in AI assistants, based on 4 checks of its own website and listings. Share of Local Fostering Voice is the percentage of local fostering Google results that come from the local authority and its regional fostering hub rather than from an independent fostering agency (IFA). The rankings show the Findability Score. We report the share to each client in their dashboard, where we can identify enough of the local results to be accurate.

How is the Findability Score worked out?

The Findability Score adds up 4 checks, each with a fixed share of the total. 60% of the score comes from what families actually see (AI assistant answers and Google results) and 40% from whether the website is built so machines can read it.

Findability Score weights.
Check	Weight	What it measures
Named by AI assistants	30%	Whether ChatGPT, Claude, Gemini and Perplexity name the service when asked how to foster in the area.
Found on Google	30%	How prominently the service appears in Google’s local results when people search for fostering near them.
Website can be read	20%	Whether search engines and AI assistants can reach and read the service’s website without errors or blocks.
Structured data	20%	Whether the website carries the labels in its code (schema.org) that tell search engines and AI assistants who runs it.
Total	100%

We may add more checks in future.

How is Share of Local Fostering Voice worked out?

Share of Local Fostering Voice is the local authority’s portion of the fostering-provider results a family sees when they search Google for fostering in their area. We sort every result by who it comes from: the local authority and its regional hub; IFAs, identified from Ofsted’s register of fostering providers, and the comparison websites that send families to them; other local authorities; and neutral national fostering organisations. The share is the local authority’s part of that total.

Results that do not come from fostering providers, such as social media, animal rescue charities that use the word “foster”, and advice or government websites, are reported separately as background and left out of the calculation. Results from websites run by FosterScope or the company that owns it are left out of the calculation.

We report a share only when 2 conditions are met: we can identify at least 90% of the results, and we have matched at least 75% of the IFAs on the Ofsted register to their websites. Today it covers Google results only. Its definition also includes answers from AI assistants, which we do not measure yet. We do measure whether AI assistants name each service, in the Findability Score's Named by AI assistants check.

Google adverts are shown on client dashboards for information but never count in either number. When an IFA advert appears on a search we measure, we record it and show it, but it does not change the Findability Score or the share. Comparison websites that list several fostering providers are different: when they appear in the ordinary results, they already count within the share, and we do not track where they send an enquiry.

What happens when we cannot measure a service?

Some local authority websites block automated checks, usually through security software. When that happens we cannot complete the structured data check, so we do not publish a Findability Score for that service. We list it as not measured, with the reason, and still show the checks we could complete. We never record a zero, because a zero would wrongly suggest the service cannot be found at all. In October 2026, 14 of the 150 services are not measured: 12 local authorities and 2 regional hubs.

A service we could not measure is different from a service that is not included. A service that is not included has no score and no “not measured” label.

A small number of services are not included at all, because we cannot yet measure a fostering page of their own. We leave these out rather than publish a figure we cannot stand behind. Where one local authority runs fostering for another, we score the one that runs it, so Cornwall’s score covers the Isles of Scilly.

What does FosterScope promise?

Being easier to find brings a local authority more fostering enquiries and a bigger Share of Local Fostering Voice against IFAs. FosterScope claims both directly. We do not claim to deliver more approved foster carers on our own, because turning an enquiry into an approval depends on the local authority’s own assessment and support. Approvals are a shared outcome of working together, not a number FosterScope delivers.

Sources

We identify IFAs from Ofsted’s “Ownership of children’s social care providers in England”, as at 31 March 2025. Each month we collect Google results and answers from 4 AI assistants: ChatGPT, Claude, Gemini and Perplexity. We publish Ofsted’s national figures, with their sources, as fostering statistics for England.

We publish the rankings as files too. For AI assistants says what the files hold and how to cite a score.

When the method changes

We improve the method over time. When we change how we measure, we still show each change in score, and say beside it that part of the change comes from an update to our method.

Each monthly measurement starts at noon on the 1st of the month.