brand_radar_12082026_0142.md

Standout observations:

House730.com is generating 17,710 total AI impression volume across 217 distinct questions. Nearly all of that visibility (14190, or ~80%) comes from Google AI Overviews. That is a wildly lopsided distribution — Copilot appears in 111 of the 217 tracked questions yet only drives 730 impressions (4% of total). Perplexity shows up in 9 questions, delivering just 320 impressions. The platform is virtually invisible on Copilot, Perplexity, and Gemini for any user-facing traffic that matters.

This tells us House730's AI visibility is not broad-based. It is almost entirely dependent on Google's AI ecosystem. That's not necessarily a problem if the volume holds, but it makes the brand vulnerable to a single-point failure — a Google ranking or AI Overviews algorithmic shift would gut the channel.

On the positive side, House730 is the undisputed market leader in its tracked brand set. It holds 1.0 share-of-voice with zero deviation over 365 data points. No other tracked brand appears in this radar. This is a monopoly position in whatever competitive set Ahrefs is monitoring. The consultant needs to verify whether that set covers the actual competitors (28Hse, Centanet, Midland, Spacious) or if this radar is tracking House730 in isolation. If competitors aren't tracked, the 1.0 SOV is meaningless and should be retired before the proposal.

Cited domain analysis tells a better story. House730.com dominates with 188 responses across 169 pages — more than double the next competitor. Centanet (89 responses), 28Hse (49), Midland (40), Ricacorp (38). This is a strong lead, but not a comfortable one. The gap from 188 to 89 responses is significant, but the presence of 194 total domains being cited in this space means AI models are pulling from a very broad set of sources. House730 owns roughly 19% of cited pages (160 of 844 total pages). That's decent for a single brand in a fragmented category, but 81% of cited content belongs to other domains — competitors, aggregators, news sites, Wikipedia.

The top-10 cited pages reveal a risk: House730's most-cited page is a category-level rental listing (/rent/t1b8/), with 9 responses. The second-most-cited page across ALL tracked domains is a 28Hse short-term rental page (7 responses). Spacious.hk's blog content on short-term rentals gets 5 responses. Weave Living and RunHotel each get 4. The AI models are disproportionately citing content about short-term rentals — possibly the most competitively contested topic in the dataset. House730 has a presence there, but it is not dominant on that specific topic. Competitors (28Hse, Spacious, Weave) are outperforming House730 on short-term rental content specifically.

Observations on the top questions:

All 10 high-volume questions are Hong Kong property pricing and livability queries — all from Google AI Overviews. The language is traditional Chinese, the volumes range from 500 to 1300 per query. These are high-intent, location-specific queries: "Tseung Kwan O HOS flats price?", "Hong Kong latest property price?", "How much to buy a village house?".

The top cited own pages are almost entirely category-level or individual listing URLs — not content-rich educational pages. The exception: one property-tips article on renting costs (/property-tips/article/1963/...). That article only gets 2 responses. The AI models are not citing House730's editorial or guide content at meaningful volume. They're citing listing pages that answer basic pricing questions.

This is a content-gap signal. House730's AI visibility is driven by its listing inventory, not by thought-leadership or informational content. If competitors start producing structured data or content that better answers "how much does property cost in X district," House730's listing-driven citations could erode.

A concerning pattern: the volume-weighted data sources show Google AI Overviews delivering 14190 impressions from only 40 questions, while Copilot needs 111 questions to generate 730 impressions. The disparity in efficiency is enormous. Either Copilot's user base is tiny, or Copilot citations don't drive meaningful traffic, or Ahrefs' Copilot measurement is of lower quality. Whichever explanation is correct, the actionable takeaway is the same: efforts to optimise for non-Google AI platforms should be deprioritised relative to Google AI Overviews unless strategic reasons override.

Key data point for the consultant to verify: Ahrefs Brand Radar is modelled data updated weekly. The share-of-voice of 1.0 with no tracked competitors is a red flag for configuration error. This must be validated before the proposal frames this as a competitive moat.

Hypothesis to verify (requires additional data):

• The short-term rental topic cluster is disproportionately cited by AI models. If House730 invests in authoritative content on short-term rentals (regulations, cost comparisons, district guides), it could close the gap with Spacious and Weave on this high-visibility topic. The 28Hse page with 7 responses is a clear target to study.
• The listing page dominance suggests structured data or schema markup on House730 is effective. This should be confirmed via a technical crawl — if House730's property schema is superior to Centanet's, that explains the domain lead and is a defensible advantage to invest further in.
• The absence of educational content in top cited pages (one article, 2 responses) may indicate a significant untapped opportunity. Competitors aren't dominating this either — Wikipedia and news sites (hk01.com) fill the gap. House730 could capture informational queries like "how much does it cost to buy a village house" with structured guides, potentially expanding beyond listing-driven AI citations.

Key signals:

• Google AI Overviews drives 80% of House730's AI impression volume (14,190 of 17,710 total) across only 40 questions. This single channel is the entire AI visibility strategy.
• House730 leads all cited domains with 188 responses across 169 pages — more than 2x the nearest competitor (Centanet, 89 responses) — but own-pages share is only 19% of total cited content.
• The top 10 AI questions are all Cantonese-language Hong Kong property pricing queries (1,300–500 volume each), exclusively from Google AI Overviews; zero informational or guide-style questions appear.
• House730's most-cited page is a rental category listing (9 responses); its editorial content is nearly invisible (one guide article, 2 responses). Competitors (28Hse, Spacious, Weave) out-cite House730 on short-term rental topics.
• The Brand Radar shows 1.0 share-of-voice with zero change over 365 days — but "tracked_brands": 0 or 1 depending on the endpoint, suggesting the competitive set may be misconfigured. This figure cannot be used as evidence of market leadership without verification.

Potential featured finding: House730's AI citation strength is almost entirely listing-driven and Google-dependent. The 19% own-pages share and the absence of informational content in top citations point to a single root cause — House730 is not being cited as an authority on property knowledge, only as a data source for current prices. If Google AI Overviews shifts its sourcing model to favour structured authority content over listing pages (a plausible evolution), House730's 14,190-impression channel is at risk. Closing the content gap — specifically on short-term rental guides and district-level pricing explainers — would diversify the citation base and build defensibility against that shift.