GEOView by Third City
GEO Audit  |  Prepared for [Brand]

AI Search
Visibility
Audit

How your brand appears across ChatGPT, Google AI Mode, Gemini, and Perplexity. Based on 100 iterations per query across 24 prompts.

Prepared by
Third City — GEO Agency of the Year 2026
Prepared for
[Brand]
Platforms audited
ChatGPT · Gemini · Perplexity · Google AI Mode
Total data points
24 queries × 100 runs = 2,400
4
AI platforms audited
24
query prompts
100
iterations per query
2,400
total data points
01 — Executive summary

[Brand] has strong AI search visibility in UK-focused [sector] queries — particularly around [key message] — but lags behind competitors in [theme] leadership narratives, and risks being anchored to an outdated product story.

[Brand] appears in a top-3 position in the majority of tested queries on ChatGPT and Perplexity. But the narrative accompanying those mentions is thinner than it should be — heavy on [key message], light on [theme], and almost absent of [key message 3] outside direct queries. Google AI Mode is a particular blind spot: its list-style format places [brand] in results without amplifying its strongest messages.

Strengths
Strong association with [key message]. Consistent top-3 presence in ChatGPT and Perplexity. Credible foothold in [theme] messaging across all platforms.
Weaknesses
Low visibility in [key message] leadership narratives. Legacy models surface ahead of current flagships. Google AI Mode deprioritises brand storytelling in favour of list-style buyer guides.
Opportunities
Strengthen presence on high-authority review and editorial sources. Expand [theme] messaging. Optimise [brand].com and structured data to influence what models cite.
02 — Methodology

How the audit
was conducted

The audit was conducted using a UK IP address, incognito mode, on desktop — designed to reflect what a typical UK user would see, free from personalisation or prior search history. Each of the 24 queries was run 100 times to produce statistically meaningful results, not snapshots.

ChatGPT Google AI Mode Gemini Perplexity
Step 01
Query set design
24 UK-focused prompts spanning [key themes], developed from consumer audience research and Google search insights via Answer the Public. Agreed with [brand] before fieldwork began.
Step 02
100-iteration sampling
Each query was run 100 times with natural language variations — providing the statistical robustness needed to calculate meaningful likelihood-of-mention scores rather than individual data points.
Step 03
Structured recording
Results logged in a structured tracker: brand mentions, positions, competing brands, cited sources, publication dates, thematic messages, sentiment, and model.
Step 04
Qual and quant analysis
Quantitative benchmarks (likelihood of mention, positioning) combined with qualitative insight (tone, source credibility, messaging gaps) for a complete strategic picture.
A note on variability
LLMs are generative by nature — they produce different answers each time. Running each query 100 times allows us to identify the likelihood of brand mentions and the most commonly cited sources, giving a statistically meaningful picture rather than a snapshot.
03 — Visibility benchmark

[Brand] in the
AI answers

[Brand] surfaced strongly in ChatGPT and Perplexity across generic [sector] queries. In ChatGPT, where [brand] was mentioned it was often the top-cited brand. Perplexity consistently included [brand] but typically placed it after key competitors. Google AI Mode showed the weakest results, leaning on list-style buyer guides that reference [brand] without narrative depth.

"[Brand] is present in the AI conversation — but often without its strongest storylines attached."
Overall mention rate — % of 100 iterations, all queries
100% 75% 50% 25% 72% 61% 48% 37% ChatGPT Perplexity Gemini Google AI Mode
Avg position when mentioned (1 = cited first)
#1 #2 #3 #1.4 #2.1 #2.6 #2.9 ChatGPT Perplexity Gemini Google AI Mode
Heatmap — likelihood of top-3 mention across 100 iterations
Generic sector queriesChatGPTGeminiPerplexityGoogle AI Mode
Best [product] in the UK78%52%71%28%
Top-rated [sector] brands61%33%58%44%
Which [product] should I buy?82%49%76%31%
[Theme] queriesChatGPTGeminiPerplexityGoogle AI Mode
Best [theme] [product]74%68%55%70%
[Theme] options for [audience]48%41%29%50%
Most [theme] [product] brands69%34%57%46%
[Audience group] queriesChatGPTGeminiPerplexityGoogle AI Mode
[Product] for [audience]80%55%60%73%
Best [product] for [audience use case]67%31%52%48%
[Audience] buying guide58%44%38%41%

High (>65%)  Medium (35–65%)  Low (<35%)

04 — Key message inclusion

What AI is
saying about you

Across all four platforms, [brand] is most consistently associated with [key message 1] and [key message 2]. Google AI Mode is the outlier: its list-style format groups [brand] with competitors without expanding on its strongest narratives.

Key message radar — frequency of message appearing across 100 iterations per platform
ChatGPT
Gemini
Perplexity
Google AI Mode
[Key msg 1] [Key msg 2] [Key msg 3] [Key msg 4] [Key msg 5] [Key msg 6] 25% 50% 75%

Distance from centre = frequency of message appearance across 100 iterations. Google AI Mode (red) consistently scores lower across all themes due to its list-first output format.

[Key message 1]
ChatGPT85%
Gemini80%
Perplexity55%
Google AI Mode25%
[Key message 2]
ChatGPT60%
Gemini78%
Perplexity30%
Google AI Mode50%
[Key message 3]
ChatGPT28%
Gemini52%
Perplexity22%
Google AI Mode18%
[Key message 4]
ChatGPT72%
Gemini48%
Perplexity56%
Google AI Mode20%
05 — Models surfacing

Which products
AI talks about

[Brand]'s AI search presence is concentrated around its [product type] line-up, with secondary mentions of [other models]. [Newer model] has started appearing in Perplexity and Google AI Mode. But the legacy [model] continues to dominate across all platforms — for reasons set out below.

Model mention frequency — % of relevant queries where each model was named
100% 75% 50% 25% ChatGPT Gemini Perplexity Google AI Mode [Legacy model] [Current flagship] [Newer model]

Legacy [model] continues to surface above newer flagships due to the volume of high-authority indexed content from 2018–2021.

Deep dive
Why legacy models still surface — and why it matters
The [model] continues to appear prominently in LLM answers despite newer models being [brand]'s current flagships. This is because [model] was one of [brand]'s earliest mainstream products in Europe, generating a large body of reviews, comparisons, and buying guides from 2018–2021 across high-authority sites. Many of these articles remain well indexed and regularly cited in AI responses, risking anchoring [brand]'s reputation in an outdated entry-level narrative.
Why it persists
High-authority review content from 2018–2021 continues to be cited across all platforms. LLMs draw on indexed sources, not product timelines. Volume of older coverage outweighs newer but sparser content.
Recommendation
Commission fresh long-form content on newer flagships in the same high-authority outlets. Optimise [brand].com model pages with structured data. Target comparison queries where legacy models currently appear.
06 — Competitor landscape

Who else is
in the room

[Competitors] were identified by [brand] as key players to analyse. The clear takeaway: [brand]'s positioning in Google AI Mode is less about competing with [main competitor] and more about outperforming mainstream European rivals in [theme] narratives.

Competitor share of voice — % of all brand mentions, 2,400 data points
31% [Brand] [Brand] 31% [Comp 1] 28% [Comp 2] 19% [Comp 3] 10%
First-mention rate — % of queries where each brand cited first
60% 40% 20% 38% 34% 18% 10% [Brand] [Comp 1] [Comp 2] [Comp 3]
Competitor
Positioning in AI answers
Threat
[Competitor 1]
Dominates [key message] as the first brand mentioned in [theme] queries. Strong association with [key messages]. Durable authority halo in LLM outputs, even in UK-focused searches.
High
[Competitor 2]
Positioned as [brand]'s closest rival in the [theme] space. Strong [key message] credentials. Frequently highlighted as offering [theme]. Less visible in ChatGPT, Gemini, and Perplexity.
Medium
[Competitor 3]
Surfaces mainly in Google AI Mode as a contextual comparison. Almost invisible in other platforms. Limited immediate threat but worth monitoring as content matures.
Watch
07 — Source mapping

Where the
answers come from

The dataset shows a clear weighting towards [source] and [source] sites across all platforms, though the mix varies considerably. Notably, no major UK national news outlets — BBC, Guardian, Telegraph, Times, FT, Independent — were cited in responses across any of the four platforms across 2,400 data points. This is a significant earned media gap.

Citation volume by source type — total citations across 2,400 data points
800 500 250 ChatGPT Gemini Perplexity Google AI Mode Specialist review Consumer media [Brand].com Commercial/listing Other editorial

Perplexity cites [brand].com directly at a much higher rate (217 citations) than other platforms. Google AI Mode shows a heavy weighting towards commercial/listing sources. UK national press is absent across all four platforms.

ChatGPT
  • [Specialist review sites]Dominant
  • [Consumer media]Strong
  • [Brand].comPresent
  • UK nationalsAbsent
Gemini
  • [Specialist review sites]Dominant
  • [Consumer media]Strong
  • [Brand].comPresent
  • UK nationalsAbsent
Perplexity
  • [Brand].comDominant (217)
  • [Specialist review sites]Strong
  • [Consumer media]Present
  • UK nationalsAbsent
Google AI Mode
  • [Commercial/listing sites]Dominant
  • [Specialist review sites]Present
  • Editorial sourcesWeak
  • UK nationalsAbsent
08 — Strategic recommendations

Building a PR strategy
that feeds AI

The audit is clear: [brand] is visible in AI search, but it is not in control of its own story. The models are reaching for whatever content exists — and right now that means legacy reviews, competitor comparisons, and commercial listing sites. A PR-led GEO strategy changes that. It puts [brand]'s strongest messages into the sources AI actually trusts.

Here is how that strategy works in practice, built around the specific gaps this audit identified in the UK electric vehicle market.

How AI search actually works for EV brands

When someone asks ChatGPT "what's the best electric car to buy in the UK", the model doesn't browse the web in real time. It draws on a training set built from indexed content — reviews, features, comparisons, expert commentary — weighted by source authority and recency. The brands that dominate those answers are the ones whose story has been told repeatedly, credibly, and in the right places. PR is the mechanism that makes that happen.

01
Own the range anxiety conversation before competitors do
The audit shows [brand] has weak visibility in real-world range and charging infrastructure queries — exactly the questions UK buyers ask most. This is a PR gap, not a product gap. Commission long-form features in T3, Auto Express, What Car, and Which? centred on independent real-world range testing. Ghostwrite expert commentary from [brand]'s engineering team for placement in The Times, Guardian, and Telegraph motoring desks. These are the outlets AI models cite most heavily for EV purchase intent queries — and right now, [brand] is largely absent from them.
Why this works for AI visibility Perplexity and ChatGPT heavily weight specialist automotive press. A single authoritative long-form review in Auto Express is cited far more frequently than ten pieces of owned content. Third City's relationships with motoring editors mean placement — not just outreach.
02
Build a sustainability narrative that AI can actually find
Gemini and ChatGPT both show [brand] has an opportunity in sustainability messaging — but the content to support it barely exists in indexed sources. The strategy: place a series of expert-led opinion pieces on the lifecycle carbon credentials of [brand]'s vehicles in the Guardian, Independent, and BBC Future. Commission a third-party data story — ideally with an academic or NGO partner — that gives journalists and AI models a citable primary source. This kind of content has a long shelf life in AI training data and builds the association [brand] currently lacks.
Why this works for AI visibility UK nationals are completely absent from [brand]'s AI citation footprint across all four platforms. A single well-sourced Guardian feature on EV sustainability can shift the source mix measurably within months.
03
Replace legacy model coverage with current flagship stories
The [legacy model] continues to dominate [brand]'s AI presence because the volume of older indexed content outweighs anything written about current vehicles. The fix is a sustained PR campaign targeting the specific outlets — Autocar, Driving Electric, Electrifying.com — that produced the original coverage. Long-form comparison pieces, owner reviews, and "best EV for…" buying guides featuring current models are the priority. Pitch these explicitly as updates to existing coverage, which editors are receptive to and which AI models weight as more recent.
Why this works for AI visibility LLMs weight recency alongside authority. New content in the same trusted outlets directly competes with and gradually displaces older citations. This is a 6–12 month play, not overnight.
04
Create the expert voice AI keeps looking for
Across all four platforms, [brand]'s spokespeople are almost entirely absent from AI-generated EV answers. The models default to quoting industry analysts, academics, and independent motoring journalists. Third City's approach: establish [brand]'s head of technology (or equivalent) as a go-to commentator on EV infrastructure, battery technology, and the future of UK motoring. A programme of media briefings, comment placement in response to EV news moments, and op-eds in the FT and Times builds an authoritative voice that AI models begin to associate with the category.
Why this works for AI visibility Named expert commentary is cited disproportionately in AI answers to opinion and recommendation queries. ChatGPT in particular surfaces expert voices when answering "should I buy" and "what do experts say" queries — the highest-intent questions in the EV purchase journey.
05
Make [brand].com the source AI wants to cite
Perplexity already cites [brand].com at 217 citations — more than any other single source. That foothold needs to be built on. Work with the digital team to create a structured resource section: a real-world range calculator, a charging network guide by UK region, a plain-language breakdown of government incentives, and an FAQ that directly mirrors the phrasing of common AI queries. Schema markup on all of these. The goal is to make [brand].com the authoritative reference that Gemini and ChatGPT reach for, not just Perplexity.
Why this works for AI visibility Owned content that directly answers specific queries is increasingly cited as AI platforms prioritise primary sources. This also reinforces the PR messaging — every piece of owned content that gets cited anchors the narrative [brand] controls.

What this looks like as a programme

Months 1–3

Foundation

Media briefings with motoring press. Range anxiety feature pitches. [Brand].com content audit and schema markup. Spokesperson positioning brief.

Months 4–6

Activation

Long-form placements live in specialist and national press. Sustainability data story placed. Expert commentary programme running. Flagship model comparison pieces published.

Months 7–12

Measurement

Re-audit against original 24 queries. Track shift in citation sources, message inclusion rates, and first-mention position across all four platforms. Iterate.

How Third City can help — GEO Agency of the Year 2026 (PRmoment)
THIRD CITY
01
AI Visibility Audit
Statistically grounded picture of your AI visibility across all four platforms — run at 100 iterations per query. Powered by GEOView, our proprietary diagnostic tool.
02
Comms-Led Strategy
A practical plan for your PR, content and digital teams: the trust signals to build, the citations to earn, and the stories that AI models actually pick up.
03
Team Training & Advisory
Jargon-free sessions that get marketing, PR, and SEO teams fluent in how AI search works. Follow-up toolkit and ongoing advisory on tap.
04
Campaign Development
High-impact PR and content built to surface in AI answers: structured explainers, expert-led commentary, and citation-worthy stories.

Ready to act
on this?

Third City can take these findings forward through targeted PR, content, and GEO optimisation work — helping [brand] move from a recognised [sector] presence to a consistently cited authority in the age of AI search.

Mark Lowe
Co-Founder
mark@thirdcity.co.uk
Izzy Shipley
Third City
izzy@thirdcity.co.uk