Key result
A$16,861.07 in AI-attributed website purchase revenue across five purchases
AI search was credited with A$16,861.07 in recorded website purchase revenue across five purchases for a property manager business between 1 April and 25 August 2026. All five purchases were attributed to ChatGPT sessions.
That was a 795% increase in AI-attributed revenue compared with the same period in 2025. The reporting period also recorded 758 AI-source sessions, up 184% year on year.
THC manages holiday rental accommodation across the NSW South Coast and Southern Highlands, Australia. Its website helps travellers explore destinations, compare individual holiday homes and book directly.
Reporting period: 1 April–25 August 2026 · Comparison: 1 April–25 August 2025 · All revenue in Australian dollars
Disclaimer: NDA signed, so I can not mention the exact business name.
The results at a glance
| Performance highlight | Result | Measurement period |
|---|---|---|
| AI-attributed booking revenue | A$16,861.07, up 795% year on year | 1 April–25 August 2026 |
| Purchases attributed to AI search | 5, compared with 1 a year earlier | Same reporting window in each year |
| AI-source website visits | 758 sessions, up 184% year on year | 1 April–25 August 2026 |
| Average revenue per AI-attributed purchase | A$3,372 | Five recorded purchases |
| AI-attributed revenue during August | A$13,489.36 across 3 purchases | 1–25 August 2026 |
| Revenue associated with one accommodation page | A$9,614.82 within the AI-attributed results | 1 April–25 August 2026 |
The August and individual-accommodation figures are included in the A$16,861.07 total. They are breakdowns of the same result.
“Booking revenue” here means GA4-recorded website purchase revenue. It has not been reconciled with cancellations, refunds or completed stays. Five purchases are a small sample; the results describe attributed performance rather than the isolated effect of any one optimisation.
The challenge: make individual stays discoverable
THC wanted stronger visibility in AI search for the destinations where it offered accommodation.
For a property management business, the discovery challenge extends across a portfolio. A traveller may be looking for a family beach holiday, a large house for a group, a home with a pool or accommodation in a particular town. Each search needs a clear connection to the relevant destination and available properties.
The opportunity was to make those connections more explicit across the website and external sources, then measure whether AI discovery led to visits and recorded purchases.
Organic search already provided a substantial commercial foundation: 428 recorded purchases and A$1.17 million in attributed revenue during the reporting period. The GEO work addressed an additional way for travellers to discover the business and enter its direct booking journey.
The approach: connect traveller needs, destinations and properties
My GEO methodology starts with what a business can actually provide, then connects those capabilities to the situations in which people search for them.
For THC, that meant connecting holiday rental accommodation, specific destinations and traveller requirements. The monitored prompts included booking directly on the NSW South Coast, finding large group accommodation in the Southern Highlands and looking for holiday homes with pools.
Generative Engine Optimisation, or GEO, is the work of improving how a business can be discovered and represented in AI-generated answers. Here, the implementation combined useful destination content, stronger location pages, off-site references and technical maintenance.
1. Expand information around the holiday
The executive report records 99 new articles during the programme. The project records also document recurring content batches and work on “things to do” content.
This content supports the research that happens around an accommodation choice. Travellers need to understand the area, what they can do during a stay and whether a destination suits their trip.
In my methodology, this is information coverage: providing relevant information around the reasons someone might choose the accommodation. A destination guide has a commercial role when it connects that research to a location or property the business actually offers.
The useful relationship is specific: a traveller’s need, a suitable destination and an accommodation option that meets it.
2. Strengthen the location pages
The programme included content enhancements across 38 suburb and location pages. The project records also show completed work on weather blocks for suburb pages.
Location pages connect broad travel research with individual properties. They help someone move from considering a destination to exploring accommodation available there.
These pages form part of the business’s infrastructure content: the core information explaining what it offers and where. The supporting articles provide context around that offering.
For a portfolio spanning coastal and inland destinations, this structure gives each location its own place in the website. It also provides a clearer route from destination information to relevant holiday homes.
3. Build references beyond the website
The report records 50 off-site citations created, supported by a completed citation-building task in the project records.
This maps to presence and repetition in my GEO methodology. The purpose is to establish references to the business beyond its own website, alongside the information on its location and accommodation pages.
These were off-site business citations. The figure does not represent 50 citations earned inside AI answers; that is a separate measurement.
4. Maintain access to the content
The programme included four website audits, internal-link implementation and redirects for broken pages. An HTML sitemap was also completed during the reporting period.
These changes supported the website’s technical foundation and the connections between its pages. Across a portfolio of locations, articles and accommodation listings, those connections help visitors navigate the site and help crawlers reach the content.
The technical work supported the same objective as the content programme: make relevant information accessible and connect it to the accommodation it describes.
5. Measure the path through to purchase
The reporting followed AI-source visits through to recorded purchases and purchase revenue.
This is the measurement sequence in my GEO methodology: visibility → website engagement → business outcomes. Each stage answers a different question: whether the business appears, whether people visit, and whether those visits produce a recorded commercial action.
For this case, the strongest evidence sits at the purchase stage. The five AI-attributed purchases carried measurable revenue, with source attribution identifying ChatGPT.
From AI discovery to recorded booking revenue
The same reporting window in 2025 recorded one AI-attributed purchase worth A$1,883.25. In 2026, that became five purchases worth A$16,861.07.
| AI-attributed outcome | 1 April–25 August 2025 | 1 April–25 August 2026 |
|---|---|---|
| Recorded purchases | 1 | 5 |
| Purchase revenue | A$1,883.25 | A$16,861.07 |
The 2026 purchases averaged A$3,372 each. During 1–25 August alone, three purchases accounted for A$13,489.36 of the total AI-attributed revenue.
This gives the case study a direct commercial connection: visits attributed to AI search were associated with purchases on THC’s website.
One accommodation page was associated with A$9,615
The report’s accommodation-referrer breakdown associates A$9,614.82 of the AI-attributed revenue with one holiday home’s page — approximately 57% of the total.
That makes individual accommodation pages important to the measurement story. A portfolio-wide traffic figure shows the scale of discovery; property-level reporting shows where recorded revenue is associated with a particular stay.
It also provides a practical direction for further work: examine the destination information, property details and booking journey around the accommodation attracting valuable visits.
What this case demonstrates
AI search became a measurable source of website purchase revenue for THC Across the reporting period, it accounted for 758 sessions, five recorded purchases and A$16,861.07 in attributed revenue.
The implementation connected several parts of the same system: destination articles, location pages, external business references and technical access. These are practical applications of the GEO methodology, with performance measured through to purchase rather than stopping at visibility.
For an Australian property management business, the commercial opportunity is to connect the places and types of stays people ask about with accommodation they can book directly. This case provides a measured example of that connection.
Explore my SEO and GEO approach to see how content, business information and measurement fit into the path to direct bookings.
Measurement note
Results cover 1 April–25 August 2026, compared with the same dates in 2025. Revenue and purchase counts were checked against property-specific GA4 session-source data in AgencyAnalytics. “Booking revenue” refers to GA4-recorded website purchase revenue; it has not been reconciled here with cancellations, refunds or completed stays.
The AI-session series follows the report’s source set across ChatGPT, Gemini, Claude, Perplexity and Copilot. Five purchases are a small sample, and the results describe attributed performance rather than an isolated causal test of any one optimisation.
Delivery counts come from the supplied executive report, with supporting implementation checked against project records. Individual accommodation names have been omitted.