I typed "Find a good care home near me" into ChatGPT – then lost the rest of my afternoon
You know those moments where you try something just to see what happens – and suddenly you're three hours in, furiously taking screenshots and muttering "Why that home?!" under your breath?
No? That's because you're not a hopeless nerd like me. But, for the rest of you, read on...
When I searched for 'care home near me' across three AI chatbots, the results left me scratching my head.
Some of the highest-rated homes on carehome.co.uk and Google barely got a mention, while a 3.8-star home dominated AI responses.
Here's what I found – and what that might mean for the future of care sector search.
But first, let's get our bearings on what we're actually dealing with here.
No? That's because you're not a hopeless nerd like me. But, for the rest of you, read on...
When I searched for 'care home near me' across three AI chatbots, the results left me scratching my head.
Some of the highest-rated homes on carehome.co.uk and Google barely got a mention, while a 3.8-star home dominated AI responses.
Here's what I found – and what that might mean for the future of care sector search.
But first, let's get our bearings on what we're actually dealing with here.
The different types of generative AI search: AI overviews and chatbots
Generative AI search is taking multiple forms, each influencing how people find information online in subtly different ways.
One of the most visible is Google's AI Overview – those snippet-like features that appear at the top of search results, summarising key information with source links.
These overviews are becoming increasingly common, appearing in up to 20% of UK and US search results, especially for long-form, informational queries.
What's more, Google just rolled out AI Mode in Search – a new conversational tab that reasons across the web and your Google data – to all users in the US. No UK date yet, but big implications for the future of SEO (and what will ads look like??)
In parallel, AI chatbots like ChatGPT, Gemini, and Claude represent a more conversational type of search.
Users can ask nuanced, multi-part questions, and these platforms pull together aggregated responses from various sources.
Each chatbot uses different training data and underlying search engines:
This results in genuinely different outputs even for identical queries – which explains why my care home search results varied.
- ChatGPT uses Bing
- Gemini uses Google Search
- Claude uses Anthropic’s own proprietary AI model system.
This results in genuinely different outputs even for identical queries – which explains why my care home search results varied.
My experiment: Three platforms, one head scratcher
Right, back to my lost afternoon tinkering with this.
I searched ChatGPT, Claude, and Gemini for both "care home near me" and "home care near me" to see what each platform prioritised.
I searched ChatGPT, Claude, and Gemini for both "care home near me" and "home care near me" to see what each platform prioritised.
Searching for local care home using ChatGPT
ChatGPT gave me a map with several homes highlighted:
Here's what stood out:
- Direct Google reviews integration: The Barchester home reviews were pulled directly from Google reviews (4.9 with 41 reviews). Same with the BUPA home, Ashley Park (4.7, 13 reviews).
- Inconsistent data pulling: The Carebase homes (Claremont Court – 5.0, 5 reviews and Queen Elizabeth – 3.8, 19 reviews) appeared on the map, but their ratings weren't pulled through.
- Strong correlation with traditional search: The homes highlighted in Google search (below) versus ChatGPT are very similar – considering there are dozens of homes in this area.
Sources cited: The top 5 sources were listed as the homes' websites themselves, with no mention of carehome.co.uk and just one reference to Trustedcare. Lower down, citations came from carehome.co.uk, Trustedcare, Lottie, Caresourcer, Yelp, and Carechoices.
Searching for local care home using Google Gemini
Unsurprisingly, I got a Google Map with four homes highlighted:
- Two Barchester homes: Silver Birch (5 stars – completely ignored by ChatGPT!) and Worplesdon View
- The same two Carebase homes: Claremont Court and Queen Elizabeth
Interestingly, even though the answer is presented through Google Maps, and the only sources listed were Google Maps reviews, these aren't the four highest-rated homes in Google reviews – Queen Elizabeth is only 3.8 stars.
Searching for local care home using Claude
Claude gave me a list of four homes, and this is where things got really interesting (for me, at least).
Queen Elizabeth topped the list. Remember, it's rated 3.8 on Google. On carehome.co.uk it has a 9.7 rating, with 102 reviews – putting it10th on that platform.
Guildford House, an Avery home, came second.
Sources: Claude was happy to cite a much more diverse range of sources, seemingly weighing carehome.co.uk data more heavily than the other platforms.
Summary of my care home searches
The Queen Elizabeth Park Mystery
Why do all three chatbots love Queen Elizabeth Park so much?! Here's what I think is happening.
Breadth trumps peak performance in AI chatbot results. This care home appears to have:
- Strong carehome.co.uk ratings with substantial review volume (9.7/10, 102 reviews)
- Consistent mentions across multiple platforms and data sources
- A robust local digital footprint beyond just Google reviews
- Structured data that AI systems can easily parse and understand
It's not the highest-rated home on any single platform, but it's consistently mentioned and well-rated across the diverse digital ecosystem that AI chatbots crawl. This suggests that distributed online presence might be more valuable than peak ratings on individual platforms when it comes to AI chatbot visibility.
Home care results: The franchise factor
The home care search threw up different patterns entirely.
Searching for home care using ChatGPT
This time, no map – just a top ten list dominated entirely by the big franchises.
Searching for home care using Gemini
Gemini only gave me two providers.
What it did well: mentioning Trinity Guildford's Outstanding CQC rating prominently. There was a little info about each provider, but not much depth.
Searching for home care using Claude
Claude provided a list of four providers (no map) and also highlighted Trinity Homecare Guildford's Outstanding rating, putting them at the top.
It offered bullet points for each provider, ranging from useful ("Top 20 Home Care Group National Award for four consecutive years") to stating the blindingly obvious ("allows individuals to remain in familiar surroundings" – thanks for that, Claude).
Summary of home care searches
- ChatGPT served up a top ten list dominated entirely by big franchises, pulling data from around 20 sources with homecare.co.uk listed first.
- Gemini gave me just two providers but helpfully mentioned Trinity Guildford's Outstanding CQC rating.
- Claude picked four providers, putting Trinity Homecare Guildford at the top – again, that Outstanding rating seemed to carry a lot of weight.
Where CQC ratings factor
Their influence varied by platform. Both Gemini and Claude specifically highlighted and prioritised Trinity Homecare Guildford's Outstanding rating, with Claude placing them at the top of home care results largely based on this credential. Whereas for my care home search, it rarely gets a mention.
It seems that Outstanding ratings can boost visibility and positioning in AI chatbot results, but they're not universally displayed as a ranking factor. It's going to come down to what you do with that rating (more on this later).
It seems that Outstanding ratings can boost visibility and positioning in AI chatbot results, but they're not universally displayed as a ranking factor. It's going to come down to what you do with that rating (more on this later).
Where generative AI pulls its information from (and how it aggregates it)
It's important to understand how these systems actually work, if we want to influence their results.
When a user enters a query, the AI engine breaks it down into component searches, consults relevant sources, and compiles a new response using predictive text modelling. These models don't copy and paste – they synthesise information into something contextually relevant and human-like.
Engines also factor in user personalisation settings, past interactions, and trusted citations. This can include third-party reviews, PR features, and data-driven content. This complex aggregation is why brand presence across multiple credible sources is so important for appearing in generative answers.
The key point: It's not just about being the best – it's about being consistently mentioned across the platforms that AI systems trust.
Gen AI is different from SEO, but there's a lot of crossover
Well, that's a relief. No need to bin your entire SEO strategy just yet.
Whilst SEO and generative AI optimisation (GEO – yes, that's apparently what we're calling it) are intertwined, they operate quite differently. Traditional SEO focuses on ranking webpages based on keywords, backlinks, and on-page optimisation.
In contrast, generative AI optimisation is about influencing the AI-generated answers themselves – through brand mentions, citations, and overall authority. Instead of just trying to appear in top search rankings, the goal is to be included in the AI-generated summary or conversation.
However, the foundations remain largely similar: content needs to be high quality, relevant, and aligned with user intent. So whilst GEO introduces new nuances, many SEO practices – like writing expert bios, structuring content clearly, and keeping it up to date – are still highly relevant.
Gen AI Is only going to grow
The influence of generative AI on search is accelerating faster than most of us anticipated. Google's AI Overviews have undergone a global rollout, and AI-powered platforms like ChatGPT have seen explosive growth – reaching 400 million weekly users by February 2025, up 300% from just over a year earlier.
This trend isn't slowing down. New features like Google's AI Mode, which offers entirely AI-generated search results, signal where things are heading.
Here's the reality check: Gen X-ers and older demographics currently search for care using traditional methods, so this isn't impacting the care sector massively yet. But that's changing rapidly as AI search becomes the default option across all age groups.
As AI chatbots continue to shape user behaviour and search expectations, businesses need to start treating AI chatbot visibility as a parallel strategy to traditional search optimisation – not a replacement, but a crucial addition.
Takeaways: Four strategies to optimise for generative search
1. Start measuring what matters (before you optimise anything)
Before diving into AI chatbot optimisation, you need to establish whether this is actually a priority for your business yet. The first step is setting proper KPIs and deciding what performance actually matters to you.
The practical starting point: Update your GA4 reporting to track what traffic is coming from generative AI sources. This lets you see whether AI chatbot traffic is increasing over time or whether this isn't yet a priority for your care business. Charlie Marchant from Exposure Ninja gives some good tips on how to go about that in this webinar.
Look for referral traffic from chatgpt.com, claude.ai, gemini.google.com, and perplexity.ai in your analytics. If you're seeing consistent growth month-on-month, AI chatbot optimisation deserves attention. If those numbers are still minimal, you might be better off focusing on traditional local SEO for now.
Test AI chatbot results for your services – search for your own care home or agency across ChatGPT, Claude, Gemini and other to see what appears.
2. Get strong on the SEO basics (yes, still)
If you want to show up in AI chatbot search, you absolutely cannot skip the fundamentals – ranking well in traditional search still provides the foundation that AI systems build upon.
AI chatbots like ChatGPT and Gemini often draw from the same pool of websites that rank well on Google. If your care home or home care business isn't already visible for local searches like "care home [town]" or "home care in [area]," you're starting from a disadvantaged position in AI chatbot results too.
But here's where it gets interesting: integrating PR data, statistics, and quotes into your SEO content is now more important than ever. Another Charlie Marchant quote: "AI systems absolutely love well-structured data and authoritative quotes – it's catnip for these algorithms."
What this means practically:
- Your location pages shouldn't just list services – they should include satisfaction statistics, CQC outcome data, and staff qualification rates
- Blog posts should incorporate industry statistics and expert quotes, not just generic advice
- Press coverage and awards should be woven into your website content, not buried in a separate news section
Local SEO fundamentals remain the foundation, but explore how they can be more data-rich and quote-heavy to perform in both traditional and AI search.
3. Keep that focus on reviews and data (but not just where you think)
My experience is that providers can sometimes obsess over carehome.co.uk reviews. But these reviews, while still important, may not dominate AI chatbot results the way they dominate traditional local search.
The Queen Elizabeth Park case study proves this point perfectly. Despite not having the best reviews, it consistently appeared at the top of AI chatbot results because of mentions across multiple other platforms.
The key point: AI chatbots seem to weight review diversity and volume across platforms more heavily than traditional search.
What this means practically:
- Don't neglect carehome.co.uk and Google reviews, but don't obsess over them exclusively
- Other industry-specific review platforms may now carry more weight
- Volume of reviews across platforms seems to matter more than perfect ratings on one platform
4. Brand is more important than ever
Just as crucial is how your brand is positioned and talked about online – and this is where traditional PR becomes incredibly valuable for AI optimisation.
If your care business is frequently mentioned in local press, wins regional awards, or has partnerships with local organisations, that sentiment builds a consistent digital footprint that AI tools can pick up and trust.
The lesson: It's not just about being "Good" or even "Outstanding" – it's about being talked about consistently across multiple platforms and contexts.
Practical PR integration for AI optimisation:
- Include staff statistics and satisfaction data in local media interviews
- Create newsworthy reports with original data about your local care market
- Partner with local organisations and ensure those partnerships are mentioned online
Pro tip: Make your achievements searchable and quotable
Got an Outstanding CQC rating? Don't just stick it on a banner. Get that rating into your metadata, your blog posts, your press releases, and your local news coverage. Include the specific quote from your CQC report. Make it part of your brand story across every platform, complete with the data that backs it up.
Plus simple ways to get started
For smaller providers handling marketing and SEO in-house, here are the immediate action steps:
For local search, prioritise Google reviews as well as comparison sites like carehome.co.uk and Autumna.
Start documenting and publishing your data. CQC ratings, staff retention rates, resident satisfaction scores – whatever positive metrics you have, weave them into your website content with specific statistics and quotes.
Build relationships with local media and organisations. The more your brand name appears in credible local contexts, the stronger your AI search presence becomes.
Conclusion: The rules are shifting – but success still follows familiar patterns
Here's the reality: the rules of engagement in care sector marketing are shifting beneath our feet. But they're not entirely unfamiliar.
Providers who take search seriously, invest in their brand story, and actively build trust across multiple platforms will still stand the best chance of being seen. That's not just on the first page of Google anymore – it's in the AI-generated answer that increasingly becomes the first (and sometimes only) result potential clients see.
The future of care sector search is already here. Yes, it's going to change a lot over coming years. But, as with everything in this space, fortune favours those who get stuck in now and start learning – rather than those with a "wait and see" mindset.