Search and answer engines
Search & AEO
How Group performs when a buyer looks for coffee - in a search box and, increasingly, in an AI assistant. Organic trading is measured for 25 Jul 2026 to 21 Aug 2026; visibility, readiness and technical scores are the current state of the estate.
$124,962
55%
55/100
$2,004,649
Visibility by answer engine
Share of monitored prompts where a Group brand appears, and where it is actually cited as a source
| Engine | Prompts | Visibility | Cited | Avg position | Sentiment | WoW | Read |
|---|---|---|---|---|---|---|---|
Google AI Overviews Highest volume surface and the one that removes classic organic clicks fastest | 420 | 57% | 24% | 0.8 | Positive | +5.6% | Watch |
ChatGPT search Recommendation intent, cites retailer and review pages more than brand pages | 420 | 42% | 18% | 2.8 | Positive | +5.3% | Off track |
Perplexity Cites structured comparison content, currently favours independent roasters | 300 | 41% | 10% | 2.0 | Neutral | +10.8% | Off track |
Google Gemini Leans on the Knowledge Graph, so entity consistency is the lever | 300 | 32% | 13% | 0.8 | Mixed | -5.2% | Off track |
Microsoft Copilot Smaller volume in AU but growing inside enterprise accounts | 180 | 26% | 9% | 1.7 | Neutral | +8.4% | Off track |
Claude Cites owned documentation and editorial rather than product pages | 180 | 26% | 7% | 2.3 | Neutral | +11.6% | Off track |
AEO readiness
What decides whether a model quotes the Group or a competitor
Name, address and brand relationships differ between the three brand sites, Google Business Profiles and directories
Product markup is partial, Organisation and Breadcrumb are missing on two of three sites
Pages lead with brand narrative, not with the extractable answer a model can lift
Few mentions on the review and comparison sources these engines actually cite
GPTBot and PerplexityBot are allowed, but key templates render client side
Published and modified dates are inconsistent, which suppresses recency weighting
Prompt clusters
The questions buyers actually ask assistants, and who gets cited today
| Prompt cluster | Prompts | Visibility | Competitor cited | Opportunity |
|---|---|---|---|---|
Best coffee beans to buy in Australia Publish a maintained, dated buying guide with explicit criteria and prices | 86 | 41% | Independent roaster | High |
Best coffee subscription Comparison page with plan, price, roast date and skip policy in structured markup | 64 | 33% | Subscription pure play | High |
Where to buy single origin coffee Origin pages per lot, with harvest, altitude and cupping score as facts | 52 | 53% | Marketplace | Medium |
Wholesale coffee supplier for cafés Service page per state with equipment, training and lead times stated plainly | 44 | 66% | Regional supplier | High |
Office coffee solutions ABC enterprise page answering machines, service SLA and per-cup cost | 38 | 27% | Equipment reseller | High |
How to choose a coffee grinder Already cited, extend with a comparison table that engines can lift | 34 | 73% | Publisher | Low |
Coffee roasters in Melbourne LocalBusiness markup and consistent NAP across all Padre venues | 29 | 81% | Local directory | Medium |
Sustainable and direct trade coffee Publish sourcing data as a citable, dated dataset rather than brand copy | 26 | 34% | Certification body | Medium |
Uncaptured commercial demand
Terms with real buying intent where the Group has no ranking page, sized on annual revenue at current conversion and basket
| Term | Intent | Monthly searches | Difficulty | Current rank | Currently held by | Revenue at stake |
|---|---|---|---|---|---|---|
specialty coffee beans online Build a dedicated commercial landing page | Transactional | 27,100 | 31 | Not ranking | Independent roaster | $666,451 |
coffee subscription australia Build a dedicated commercial landing page | Commercial | 22,200 | 34 | Not ranking | Independent roaster | $377,964 |
single origin coffee Build a dedicated commercial landing page | Transactional | 14,800 | 26 | Not ranking | Independent roaster | $363,966 |
wholesale coffee supplier Build a dedicated commercial landing page | Commercial | 9,900 | 24 | Not ranking | Regional supplier | $168,552 |
espresso beans for home Build a dedicated commercial landing page | Transactional | 6,600 | 19 | Not ranking | Independent roaster | $162,309 |
best coffee beans australia Brief a guide, then link it into the matching collection | Informational | 18,100 | 38 | Not ranking | Marketplace | $119,840 |
coffee beans melbourne Build a dedicated commercial landing page | Commercial | 5,400 | 29 | Not ranking | Local café group | $91,937 |
office coffee machine hire Brief a guide, then link it into the matching collection | Informational | 8,100 | 22 | Not ranking | Equipment reseller | $53,630 |
Technical health
What is currently suppressing both classic ranking and AI extraction
| Check | Now | Target | Read |
|---|---|---|---|
Largest contentful paint (field) Ranking and conversion on mobile product templates | 3.1s | Under 2.5s | Off track |
Interaction to next paint Filtering and variant selection on collection pages | 227ms | Under 200ms | Watch |
Cumulative layout shift Late loading promo bars pushing the add to cart button | 0.11 | Under 0.10 | Watch |
Indexable share of published pages Faceted URLs and expired lots consuming crawl budget | 83% | Over 90% | Watch |
Orphan pages Product and lot pages with no internal link path | 90 | 0 | Off track |
Duplicate or templated titles Cannibalisation between roast size variants | 36 | 0 | Off track |
Thin pages under 300 words Suppresses both classic ranking and answer engine extraction | 65 | Under 20 | Off track |
XML sitemap coverage New lots are discovered days later than they should be | 89% | 100% | Watch |
Structured data coverage
Share of eligible pages carrying valid markup
Price, availability and roast date should all be marked up
Establishes the entity for AI answers and the Knowledge Graph
Gives engines the category hierarchy for free
Padre venues and Seven Miles training spaces
Only where genuine, first-party reviews exist
Collection pages become extractable lists for answer engines
Guides and education content with author and date
Genuine customer questions, not manufactured FAQ blocks
Roadmap
Sequenced so the work that unblocks everything else happens first
| Action | Window | Owner | Impact | Effort |
|---|---|---|---|---|
Fix the entity before chasing rankings One canonical Organisation record per brand, consistent naming and parent relationships across the three sites, Google Business Profiles, Wikidata and the major directories. Every answer engine resolves the entity before it decides whether to cite you. | 0-30 days | Technical SEO | AI visibility | Low |
Ship the missing commercial pages 8 transactional terms worth roughly 112k searches a month currently have no ranking page. Collection and subscription pages first, guides second. | 0-60 days | Content and merchandising | Revenue | Medium |
Rewrite the top 40 pages answer-first Lead with the extractable answer, then the brand narrative. Add comparison tables, explicit prices and dated facts so models can quote the page rather than paraphrase a competitor. | 30-90 days | Content | AI citations | Medium |
Complete structured data across all three brands Product, Organisation, Breadcrumb and LocalBusiness at minimum, with roast date and availability kept live from the commerce platform rather than hand written. | 30-60 days | Engineering | Rich results and AI | Medium |
Clear the technical debt suppressing crawl Faceted URL rules, canonical variants, orphan lot pages and thin templates. This unlocks the rest of the programme rather than adding to it. | 0-45 days | Engineering | Indexation | High |
Build the citation footprint Get the brands into the review, comparison and directory sources these engines already cite. Digital PR aimed at citations, not backlinks for their own sake. | 60-180 days | Earned media | AI visibility | High |