UK Heat Pump Search Trends & AI Visibility Report 2026
How are UK consumers researching heat pumps — and which types of heat-pump content are now being surfaced by generative search?
- 709 unique search terms analysed
- Approx. 17,700 Google generative AI impressions
- Approx. 19,800 Microsoft AI citations
Key Findings
Heat Pump Guide UK has analysed 709 unique heat-pump-related search terms alongside first-party visibility data from both Google Search Console and Bing Webmaster Tools.
Over the three months to mid-October 2026, Heat Pump Guide UK recorded approximately 17,700 impressions in Google’s generative AI search features and approximately 19,800 citations across Microsoft’s supported AI experiences.
The two platforms measure different things, so the figures should not be added together. Taken separately, however, they provide an unusually useful view of how specialist UK heating content is being discovered and used across emerging AI search experiences.
The research combines three different types of evidence: UK search-demand data, Google generative-search impressions and Microsoft AI citations.
709
unique UK heat-pump-related search terms analysed
~17,700
Google generative AI impressions over the three months to mid-October 2026
~6,730
Google generative AI impressions in August 2026, the strongest completed month
~19,800
Microsoft/Bing AI citations over the same three-month period
The clearest pattern
Practical decision-making content performs strongly across Google’s generative search data. Product comparisons, smart-heating suitability, property-specific installation costs and common home-comfort problems all feature among Heat Pump Guide UK’s most visible pages. On Microsoft’s platforms, air-to-air heat pump, air-conditioning and cost guides are cited most.
The data does not prove that these subjects are universally preferred by AI systems. It shows which Heat Pump Guide UK pages were most frequently surfaced within the first-party datasets analysed.
What Search Data Was Analysed?
The UK search-demand research originally contained 757 keyword rows gathered across five heat-pump research exports.
After duplicate removal, the working dataset contained 709 unique search terms.
| Dataset Measure | Result |
|---|---|
| Original keyword rows | 757 |
| Unique search terms | 709 |
| Duplicate rows removed | 48 |
| Terms appearing across multiple exports | 43 |
| Keywords with at least 500 monthly searches | 42 |
| Keywords with at least 50 monthly searches | 523 |
| Keywords showing zero reported volume | 109 |
| Keywords without volume data | 35 |
This keyword dataset was compiled in August 2026 and has not been refreshed for this update.
The dataset spans homeowner questions, product comparisons, installation costs, running costs, air conditioning, air-to-air heat pumps, commercial systems, industrial applications, refrigerants and heat-pump technology.
Google Generative AI Visibility: About 17,700 Impressions in Three Months
Google Search Console’s Generative AI performance report recorded approximately 17,700 impressions for Heat Pump Guide UK across the three months to mid-October 2026.
Google defines these impressions as occasions where links to the site were shown to users within supported generative AI features on Google Search.
The report currently includes:
- AI Overviews
- AI Mode
Month by month
Completed months recorded approximately 4,290 impressions in July, 6,730 in August and 4,170 in September 2026. Daily impressions rose again during the first half of October.
The report is particularly valuable because it provides page-level visibility, allowing us to see which subjects are actually being surfaced within Google’s generative-search experiences.
Which Heat Pump Guide UK Pages Appear Most in Google Generative Search?
The leading pages over the three months to mid-October 2026 reveal a broader pattern than heat-pump installation information alone.
| Page | Google Generative AI Impressions | Main Search Need |
|---|---|---|
| Best Air Source Heat Pumps UK | 5,282 | Product comparison and selection |
| Best Smart Thermostat for Combi Boilers | 1,148 | Heating-control product selection |
| Best Smart Thermostat for Old Houses | 935 | Property-specific heating controls |
| Why Are My Upstairs Rooms Too Hot? | 929 | Home heating and comfort problem |
| Heat Pump Cost for a 4-Bed House | 807 | Property-specific installation cost |
| Are Smart TRVs Worth It? | 734 | Purchase-value assessment |
| Heat Pump Cost for a 3-Bed House | 729 | Property-specific installation cost |
| Why Is One Room Colder Than the Rest of the House? | 521 | Heating and home-comfort problem |
| Best Smart Thermostat for Gas Boilers | 505 | Heating-control product selection |
| Hive vs Nest | 436 | Product comparison |
What the Google Data Suggests
The strongest Google generative-search pages do not belong to a single conventional SEO category.
Instead, they fall into several practical information needs.
Product Selection
Best heat pumps and smart thermostats help users narrow down specific products or technologies.
Property-Specific Costs
Three-bedroom and four-bedroom heat-pump cost guides answer a more precise question than a generic installation-cost page.
Direct Comparisons
Head-to-head content such as Hive vs Nest helps users compare competing options.
Real Home Problems
Questions about rooms being too hot or too cold show that generative search visibility extends beyond conventional heat-pump keywords.
Our interpretation: The common characteristic is not simply the phrase “heat pump”. These pages directly resolve a defined decision, comparison or household problem.
The Best Air Source Heat Pumps Guide Is the Clear Google AI Leader
The strongest individual page in the Google dataset was our Best Air Source Heat Pumps UK guide.
It recorded 5,282 generative AI impressions during the three-month period.
That represents approximately 30% of the site’s roughly 17,700 property-level Google generative AI impressions.
This percentage should be treated as an indicative comparison rather than a strict share calculation because Google applies different aggregation rules to property-level and page-level performance data.
Nevertheless, the scale of the difference is clear: the page recorded more than four times the impressions of the second-highest page shown in the report.
Specific Questions Can Outperform Broader Heating Topics
One of the most useful findings is the strength of highly specific content.
For example:
- Best Smart Thermostat for Old Houses recorded 935 impressions.
- Why Are My Upstairs Rooms Too Hot? recorded 929.
- Why Is One Room Colder Than the Rest of the House? recorded 521.
These are not broad category pages. They answer narrowly defined problems that a homeowner might naturally describe in conversational language.
This is relevant to generative search because AI interfaces are particularly well suited to longer, more natural-language questions where the user is looking for an explanation rather than simply a list of websites.
Property-Specific Heat Pump Cost Content Is Also Prominent
Two property-size heat-pump cost guides appear close together within the Google data:
- Heat Pump Cost for a 3-Bed House — 729 impressions
- Heat Pump Cost for a 4-Bed House — 807 impressions
Microsoft’s data shows the same pairing, with 314 citations for the 3-bed guide and 268 for the 4-bed guide.
The similar visibility of these pages supports the case for addressing installation cost in the context homeowners actually use when researching a project.
A homeowner is often less interested in an abstract national average than in questions such as:
- What might a heat pump cost for a house like mine?
- Would a larger house require a larger heat pump?
- Could radiator or cylinder work increase the cost?
- How much difference could a grant make?
Actual costs vary substantially by heat loss, system design, required output, emitter sizing, hot-water requirements and installation complexity, so property size alone should never be used as a firm quotation.
Microsoft/Bing Recorded Around 19,800 AI Citations
The Google figures measure impressions. Microsoft’s Bing Webmaster Tools measures a different signal: citations.
Across the same three-month period, Heat Pump Guide UK recorded approximately 19,800 citations within Microsoft’s AI Performance reporting.
The site also averaged approximately 21 unique cited pages per day.
Microsoft defines a citation as a website URL displayed as a source within an AI-generated answer across its supported AI experiences.
These currently include Microsoft Copilot, AI-generated Bing experiences and selected partner integrations.
Important: Bing states that citation counts do not indicate ranking, authority or the position of the source within an individual answer.
Google Impressions and Microsoft Citations Are Not the Same Metric
The two datasets should not be combined into a single headline number.
| Platform | Metric | What It Measures |
|---|---|---|
| Google Search Console | Generative AI impressions | How often links to Heat Pump Guide UK were shown to users within Google’s supported generative search features. |
| Bing Webmaster Tools | AI citations | How often Heat Pump Guide UK URLs were displayed as sources within supported Microsoft AI-generated answers. |
For that reason, it would be misleading to describe the figures as “37,500 AI appearances”.
The value comes from having separate first-party measurements showing that the site is being surfaced across both major search ecosystems.
What Are Grounding Queries?
Bing Webmaster Tools also exposes samples of grounding queries.
These are phrases used during the retrieval process when an AI system searches for information to support an answer.
The retrieval query therefore does not necessarily have to match the original wording entered by a user.
Traditional Search
A user submits a search and chooses between search results.
AI Search
An AI system can interpret a user’s request, retrieve supporting information and generate a combined answer with cited sources.
This creates a useful additional research layer: traditional keyword data helps reveal what people search for, while grounding information can provide some visibility into how AI retrieval systems identify the information needed to support an answer.
Google and Microsoft Surface Different Heat Pump Guide UK Content
The two platforms surface different parts of the site. Google’s generative-search impressions concentrate on product selection, heating controls and specific home problems. Microsoft’s AI citations concentrate on air-to-air heat pumps, air conditioning, installation costs and switching from gas.
Between them, the site is being surfaced across almost every major topic it covers. On Microsoft’s platforms alone, an average of 21 different pages were cited each day.
| Rank | Google: most-surfaced pages | Google impressions | Microsoft: most-cited pages | Approx. Microsoft citations |
|---|---|---|---|---|
| 1 | Best Air Source Heat Pumps UK | 5,282 | Air-to-Air Heat Pumps UK | 2,900 |
| 2 | Best Smart Thermostat for Combi Boilers | 1,148 | Home Air Conditioning UK | 1,900 |
| 3 | Best Smart Thermostat for Old Houses | 935 | Best Portable Air Conditioners UK | 1,500 |
| 4 | Why Are My Upstairs Rooms Too Hot? | 929 | Air Source Heat Pump Cost UK | 1,200 |
| 5 | Heat Pump Cost for a 4-Bed House | 807 | Gas Boiler to Heat Pump | 1,100 |
What Microsoft’s grounding queries show
The grounding queries behind those citations point the same way. The most frequent was “air to air heat pumps uk”, with approximately 1,700 citations. Other common retrieval phrases included:
- heat pump grants uk: approx. 390
- air to air heating systems: approx. 375
- best air source heat pumps uk 2026: approx. 375
- air conditioning unit: approx. 295
- are heat pumps worth it uk: approx. 250
- heat pump thermostat: approx. 235
- air to air heat pump cost: approx. 165
Our interpretation: This supports the earlier finding that people describe air-to-air technology in air-conditioning language. It also suggests a publisher’s visibility can differ substantially between AI platforms, so findings from one ecosystem should not be assumed to apply to another. We have not tested why the two platforms differ. Their retrieval systems, user bases and reporting methods are all different.
Commercial and Industrial Heat Pumps Form a Separate Search Market
The wider search-demand research also identified a significant body of heat-pump terminology that sits outside conventional homeowner research.
The working dataset contained:
60
direct commercial heat-pump terms
25
industrial and process-heat terms
142
terms within the broader technology and refrigerant category
Microsoft’s AI citation data shows these specialist pages are also being used as sources: Commercial Air Source Heat Pumps (321 citations), Commercial Heat Pumps UK (268) and Commercial Heat Pump Installation (209).
This supports treating commercial heat pumps as a specialist information area rather than simply scaling up homeowner advice.
A hotel, warehouse, factory or industrial process can involve very different temperatures, loads, operating profiles and design considerations from a residential heat-pump installation.
Consumer Language Does Not Always Match Industry Language
The keyword research also reinforces the importance of terminology.
Consumers do not necessarily describe technology in the same language manufacturers, engineers or policymakers use.
A good example is air-to-air technology.
A reversible split air-conditioning system is effectively an air-to-air heat pump, but a homeowner researching summer cooling may search for phrases such as home air conditioning, air con or split air conditioning rather than air-to-air heat pump.
That is why Heat Pump Guide UK maintains separate resources covering home air conditioning in the UK and air-to-air heat pumps.
Microsoft’s grounding data reflects this. “Air to air heat pumps uk” was the most frequent grounding query behind citations of the site, while “air conditioning unit” also appears among the most common.
What This May Tell Us About AI Search Content
The Google results contain an important pattern.
The pages receiving the strongest generative-search visibility generally do at least one of four things:
- Compare options.
- Quantify a cost for a defined situation.
- Answer a specific household problem.
- Help a user decide whether something is suitable or worthwhile.
That does not mean publishers should manufacture thousands of narrowly targeted pages.
It suggests that useful content should clearly identify the real question being answered and provide sufficient context for the answer to stand on its own.
This also aligns with Microsoft’s guidance that content used by AI systems benefits from clear structure, supporting evidence, depth and up-to-date information.
AI Visibility Is Not the Same as Website Traffic
Another important distinction is that an AI impression or citation does not necessarily result in a website visit.
A user may see enough information inside an AI-generated answer without clicking through to the cited source.
AI visibility therefore needs to be evaluated alongside conventional measures such as:
- organic search impressions;
- organic clicks;
- website sessions;
- return visitors;
- affiliate engagement where relevant; and
- external citations and backlinks.
For publishers, the emerging question is not only whether a page ranks in conventional search, but whether its information is considered useful enough to be retrieved and referenced within AI-generated answers.
Note on Google’s August 2026 Data-Logging Issue
Google reported a data-logging issue that decreased recorded impressions on the Generative AI performance report in Search between 13 and 17 August 2026.
Google states that the issue affected data logging only, and that on 21 August 2026 the missing data was restored and the report now shows complete metrics.
What this means for this report: The figures in this October 2026 update were recorded after Google restored the affected data, so the August 2026 total is treated as complete. The first edition of this report, published in August 2026, used figures recorded before the data was restored.
What This Research Does Not Prove
AI visibility data is new and should be interpreted carefully.
- Google impressions are not AI rankings. They measure eligible link visibility within supported generative search experiences.
- Bing citations are not authority scores. A citation count does not show ranking, prominence or importance within an answer.
- Google and Bing figures cannot be added together. They use different metrics and aggregation methods.
- This is not a competitor benchmark. Comparable first-party AI visibility data for competing UK heat-pump publishers has not been analysed.
- Some Microsoft figures are rounded. Bing Webmaster Tools displays larger totals in rounded form, so citation figures above 1,000 are approximate.
- Search volume is estimated. Keyword tools model demand and should not be interpreted as a precise census of every UK search.
- AI search systems continue to change. Their retrieval methods, interfaces and reporting tools are still evolving.
Methodology
UK Search-Demand Dataset
Heat Pump Guide UK combined five keyword-research exports covering residential, commercial, industrial and technology-related UK heat-pump searches.
The combined dataset contained 757 rows. After duplicate removal, 709 unique search terms remained. This dataset was compiled in August 2026.
Terms were reviewed and grouped by subject and intent to identify broader areas of demand. Keyword volumes are treated as estimates rather than exact search counts.
Google Generative AI Data
Google data was taken from the Generative AI performance report within Google Search Console for HeatPumpGuideUK.co.uk using a three-month reporting window to mid-October 2026.
The report showed approximately 17,700 total generative AI impressions. Completed months recorded approximately 4,290 in July, 6,730 in August and 4,170 in September 2026.
Google states that this report currently includes AI Overviews and AI Mode. Property-level totals use property aggregation, while page-level data is grouped by URL.
Google disclosed a logging issue affecting 13 to 17 August 2026 and reported on 21 August 2026 that the missing data had been restored.
Microsoft/Bing AI Performance Data
Microsoft data was taken from Bing Webmaster Tools AI Performance reporting for HeatPumpGuideUK.co.uk, using its three-month view. Page-level citations and grounding queries were taken from the same report.
The analysed period recorded approximately 19,800 citations and an average of approximately 21 unique cited pages per day.
Bing states that total citations show how often URLs are displayed as sources within supported AI-generated answers and do not indicate ranking, authority or source position.
Rounding and Interpretation
Figures marked as approximately or around are rounded from the values displayed in the first-party reporting tools.
Google generative-AI impressions and Microsoft AI citations are reported separately because they represent different measurements.
Changes Since the August 2026 Edition
The first edition of this report recorded 9,290 Google generative AI impressions and approximately 5,400 Microsoft citations over its earlier three-month window. This October 2026 update replaces those figures with the three months to mid-October 2026.
How to Cite This Research
Journalists, publishers, manufacturers, researchers and industry organisations are welcome to reference statistics from this report with attribution and a link to the original research.
Suggested attribution
Source: Heat Pump Guide UK, UK Heat Pump Search Trends & AI Visibility Report 2026.
When referencing the AI data, please keep the Google and Microsoft metrics separate. Google figures relate to generative-AI impressions, while Microsoft figures relate to AI citations.
For media enquiries, methodology questions or research collaborations, visit our contact page.
About Heat Pump Guide UK Research
Heat Pump Guide UK is an independent UK information resource covering residential heat pumps, heating controls, home cooling and specialist commercial heat-pump technology.
Alongside practical guides, we analyse search behaviour and emerging AI-search visibility to better understand the questions consumers and industry users are asking about heating technology.
Heat Pump Guide UK does not install heat pumps. Optional installer matching may be available separately from our independent research and editorial content.
Research Sources & Further Reading
This report combines Heat Pump Guide UK’s own research with first-party search-platform reporting and documentation.
- Heat Pump Guide UK: 2026 UK heat-pump keyword research dataset.
- Google Search Console: Generative AI Performance data for HeatPumpGuideUK.co.uk.
- Google Search Console — Generative AI Performance Report
- Google Search Console — Data Anomalies
- Bing Webmaster Tools: AI Performance data for HeatPumpGuideUK.co.uk.
- Microsoft — Introducing AI Performance in Bing Webmaster Tools
- Microsoft — Elevating the Role of Grounding on the AI Web