Christchurch Hotel AI Visibility Case Study (June 2026)

The owner of a luxury accommodation property in Cashmere, Christchurch had a problem she could not see. Her property had earned consistent five-star reviews. Her garden setting in the Port Hills offered something no city-centre hotel could match. Her rates were competitive for her tier. And yet when a couple in Auckland opened ChatGPT and typed "where should we stay in Christchurch for a luxury escape," her property did not appear. Not once. Not on any of four AI engines.

On 9 June 2026, we ran a RevPARGenius Hotel AI Visibility check across 18 valid engine checks on the Christchurch luxury accommodation market. The subject property — a real Cashmere, Port Hills luxury accommodation, anonymised at the owner's request — scored a visibility score of 16/100 with a 6% mention rate across ChatGPT, Perplexity, and Grok. The George appeared in 67% of the same checks. Hotel Montreal appeared in 44%. The Classic Villa in 39%. The subject property appeared exactly once — a neutral mention at position 8 on Perplexity for the single prompt most closely matched to its own positioning. Here is what the scan showed, why the gap exists, and what to do about it.

Is a 6% mention rate in Christchurch actually a problem?

The honest answer is yes — with context. The Cashmere luxury accommodation is not a weak property. It has a genuine Port Hills garden setting that no inner-city hotel can replicate, consistent five-star feedback, and a positioning in the boutique luxury tier that is genuinely differentiated from the city-centre competition. But hotel AI visibility in 2026 is not a product quality competition. It is an information footprint competition. A 2024 Skift Research survey found that 56% of US travellers already use AI tools to plan trips — and that share is growing in high-mobile-adoption markets like New Zealand and Australia. When a couple in Sydney asks Perplexity "best place to stay in Christchurch for a luxury weekend," the hotels Perplexity recommends capture the booking intent at its highest point. A hotel invisible at that moment is invisible to that guest entirely.

The 6% mention rate is a problem because of what sits above it. The George at 67%, Hotel Montreal at 44%, The Classic Villa at 39% — these are not just properties with bigger marketing budgets. They have deeper, wider, more structured information footprints across the sources AI engines cite. The Cashmere property's product is genuinely competitive at its tier. Its AI presence is not.

What does the citation source breakdown reveal?

The scan analysed 128 citations across all engine checks. Of those: 41.4% came from "other" sources — editorial content, travel blogs, and destination guides. OTAs accounted for 24.2%. Review platforms for 14.1%. The property's own website: 11.7%. Reddit and community sources: 7.8%. Trade press: 0.8%.

The 41.4% "other" category is where the gap lives. The George appears in national and international editorial sources — Lonely Planet New Zealand, NZ Herald Travel, Condé Nast Traveller features, Tourism New Zealand's official recommendations, and travel blogs across English-language markets. These are exactly the sources AI engines weight most heavily when selecting which hotels to recommend. The Cashmere property has minimal presence in any of them.

The OTA citation share of 24.2% is worth reading carefully. It sits well below the Cloudbeds 2025 industry baseline of 55.3% — meaning the property is less OTA-dependent than typical, which is commercially healthy and a sign of strong direct booking performance. But it also means AI engines are not finding the property through the channels they most commonly cite. The own-site figure of 11.7% is above the Boracay case (6.7%), suggesting the website has more presence — but not enough authority to function as the primary source AI engines draw from when describing this property.

Why does The George appear in 67% of checks when this property appears in 6%?

The George's dominance is structural, not arbitrary. The Princeton/IIT Delhi GEO study (Aggarwal et al., KDD 2024) found that content with named, dated statistics generates 31% higher AI citation rates, and named expert quotes generate 41% higher rates. Those lifts compound across every authoritative source that covers a property. The George Christchurch is referenced in Lonely Planet New Zealand, Condé Nast Traveller, Vogue Australia Travel, TripAdvisor with thousands of verified reviews, and multiple Tourism New Zealand official recommendation pages — all sources AI engines are trained to weight heavily. That depth of structured, corroborated, editorially validated information is what produces a 67% mention rate.

For the Cashmere luxury property, the information landscape is significantly thinner across every dimension. The editorial coverage gap is the most material: with 41.4% of all Christchurch hotel AI citations coming from the "other" category — travel editorial, guides, and destination content — a property with no presence in that category is structurally excluded from the largest single citation source AI engines draw from. The property's own website accounts for 11.7% of citations, which is above average, but that share cannot compensate for the editorial gap when the editorial category is five times larger.

Importantly, the gap is not review quality. The Cashmere property's review scores are genuinely strong. What drives the citation volume differential is review quantity — The George and Hotel Montreal have accumulated review volumes across TripAdvisor, Google, and Booking.com that signal social proof depth to AI models, regardless of whether individual ratings differ materially. Volume drives citation frequency in AI recommendation engines, not just rating average.

What should a Christchurch luxury property actually do about this?

The scan that showed the gaps also generated four prioritised recommended actions — two rated HIGH for content, one HIGH for reviews, and one LOW for schema. These actions are sequenced in order of estimated impact and time investment, and together they map the full fix path for a property in this position.

1. Close the "best boutique hotels in Christchurch" and "where to stay for couples" visibility gaps. These are the two highest-priority content actions — both rated HIGH, each estimated to take 1 hour, and each projected to deliver +2 to 4 percentage points on mention rate within 30–45 days. The fix involves creating or updating content on the property website that directly and specifically answers these query formats in natural language: what makes it the best boutique option, why couples choose it, what the Cashmere setting provides that city hotels cannot. AI engines do not find a property on these prompts because the property's content does not yet answer the question these prompts are asking.

2. Upgrade the neutral mention to positive on the bullseye query. The property's single appearance — position 8, neutral, on "luxury hotels in Christchurch, New Zealand" on Perplexity — is the starting point, not the problem. A neutral mention means the AI is aware of the property but not confident enough to recommend it positively. The recommended action is a review velocity campaign specifically targeting authenticated guests: a post-stay message linking directly to TripAdvisor and Google Review for guests who rated their stay at 9 or 10. The estimated impact is +5 to 10 points on Visibility Score as the sentiment component improves — this is the single highest-leverage action on the current position.

3. Implement LodgingBusiness + FAQPage structured data. Rated LOW by the scan (meaning it is the fastest and easiest fix, not the least important), this 15-minute schema implementation is described by the RevPARGenius action system as "the single highest-ROI technical action for AI visibility." LodgingBusiness schema identifies the property type and key attributes to AI crawlers; FAQPage schema feeds the direct-answer format AI engines prefer to extract. ChatGPT and Perplexity explicitly prefer pages with correct schema markup. The absence of this structured data is confirmed in the scan and is directly responsible for part of the citation gap. For a New Zealand luxury accommodation targeting an international visitor market, this fix should be done this week.

4. Earn editorial coverage in the sources AI trusts for New Zealand travel queries. Stuff.co.nz travel section, NZ Herald Travel, the Jetsetter luxury travel directory, and regional tourism board features are regularly cited by Perplexity and Gemini for New Zealand accommodation queries. A single feature on any of these produces a persistent citation source AI engines draw from repeatedly. This is the slowest of the four actions — editorial placement takes months — but it addresses the 41.4% "other" category gap directly and creates compounding returns that the faster technical fixes do not.

What does this case study reveal about the Christchurch AI visibility market overall?

The Christchurch scan revealed a market with clear stratification. The George at 67% and Hotel Montreal at 44% occupy a dominant tier that reflects decades of editorial coverage and review accumulation. A mid-tier group — The Classic Villa (39%), The Observatory Hotel and The Mayfair (both 33%), Crowne Plaza and Adina Heritage (both 28%) — holds the middle ground. Below that, Otahuna Lodge at 22% represents the highest-performing boutique independent, and the tracked independents including Rydges Latimer and Sudima Airport at 11% occupy the lower tier. The Cashmere property at 6% sits at the base of the tracked set — not because of product weakness, but because of information footprint structure.

The Christchurch AI visibility category is still early-stage as a managed discipline. The George and Hotel Montreal are not actively running AI visibility strategies — their dominance is a legacy of their information depth, not the result of deliberate GEO investment. That means the window for independent and boutique properties to close the gap, before AI visibility becomes a formally managed and funded category for established hotel brands, is open now. The Cashmere property's distinctive positioning — a luxury garden setting in the Port Hills, genuinely differentiated from the city-centre competition — is an asset that AI engines will reward once it is properly documented, structured, and cited. The challenge is not the product. It is making the product legible to the systems that now make the recommendation.

Frequently Asked Questions

What is a good hotel AI visibility score for a Christchurch property?

Based on RevPARGenius scan data from June 2026, the Christchurch market stratifies into three tiers: established luxury and chain-affiliated hotels at 28–67% (The George, Hotel Montreal, The Classic Villa, The Mayfair, Crowne Plaza, Adina Heritage), boutique independents at 11–22% (Otahuna Lodge, Rydges Latimer, Sudima Airport), and the majority of independent boutique properties below 10%. A realistic 90-day target for a property currently at 6% is 20–28%, achievable through the four recommended actions from this scan: content gap fixes, review velocity, structured data implementation, and one or two editorial placements in New Zealand travel media.

Which AI engines matter most for New Zealand hotel visibility?

ChatGPT and Perplexity have the broadest traveller adoption for trip planning queries in the Australian and New Zealand market. Gemini is growing rapidly on Android, the dominant mobile platform in New Zealand, though this scan recorded Gemini errors across all six prompts — meaning the current score reflects three engines only. Grok has a smaller audience but is growing among technically engaged travellers. For a Christchurch property targeting Australian visitors and international travellers, appearing consistently on ChatGPT and Perplexity is the highest priority, with Gemini becoming increasingly important as its Android-driven adoption grows.

Why did Gemini return errors in this scan?

This scan recorded Gemini errors on all six prompts (displayed as "—ERROR" in the results table), reducing the valid check count from 24 to 18. Gemini errors in RevPARGenius scans are typically caused by rate-limiting at the API level or temporary service interruptions — they do not reflect the property's actual visibility on Gemini. A follow-up scan will include Gemini results and provide a fuller picture of the property's AI visibility, potentially increasing or decreasing the mention rate depending on Gemini's responses.

How is the RevPARGenius Hotel AI Visibility score calculated?

The score runs tracked prompts across ChatGPT, Perplexity, Gemini and Grok and counts how many responses mention the property by name. The mention rate is appearances divided by total valid checks (18 in this Christchurch scan, excluding the 6 Gemini errors). The visibility score combines raw mention rate with a sentiment weighting — positive mentions score higher than neutral, and neutral higher than negative. The score breakdown of 3 + 13 reflects 3 points from the mention component and 13 from the base/sentiment component. The score is designed to be trended across multiple runs rather than interpreted as a single absolute figure — the platform note "run a second check to see the score trend" reflects this design.

Can a boutique Cashmere property compete with The George on AI visibility?

Not on overall mention rate — The George's footprint is built over decades of editorial coverage and review accumulation. But competing on specific high-intent prompts is realistic and achievable. "Boutique luxury accommodation Christchurch," "where to stay in Christchurch for couples," and "luxury garden accommodation Christchurch New Zealand" are all prompts where an independent property with strong reviews, structured schema, and targeted editorial coverage in NZ travel media can appear consistently. The Cashmere setting — genuinely distinct from city-centre hotels — is a competitive advantage on those specific prompts, provided the property's content makes that distinction legible to AI models.

The first step in fixing a property's AI visibility is knowing exactly where it stands — which prompts it appears in, which engines are citing it, and what competitors are capturing that it is not. The Hotel AI Visibility Checker runs that scan automatically, tracks it over time, and shows the specific gaps and recommended actions that matter most for your property and market.

Sources: RevPARGenius Hotel AI Visibility Checker scan, 09 June 2026, 14:31 UTC (18 valid checks across ChatGPT, Perplexity and Grok; Gemini returned errors on all 6 prompts this run; 128 citations analysed); Cloudbeds 2025 "The Signals Behind Hotel AI Recommendations" (OTA citation baseline 55.3%); Princeton/IIT Delhi GEO Study (Aggarwal et al., KDD 2024) on +31% citation lift from dated statistics and +41% from named expert quotes; Skift Research 2024 AI Travel Planning Adoption Survey (56% of US travellers use AI for trip planning). Subject property anonymised at owner's request. Competitor names and mention rates are observational data from AI engine responses. Gemini error on this run is attributed to API rate limiting — not a reflection of the property's actual Gemini visibility. Last reviewed June 2026. RevPARGenius is an independent hotel market intelligence platform — not affiliated with any OTA, revenue management system, or hotel chain.