Topical authority is the depth and breadth of a site's coverage of one subject area, as opposed to domain authority, which is a generic site-wide score. Depth on the topic beats raw site-wide strength on long-tail queries.
The standard advice is HubSpot's pillar-cluster model: one pillar page, 8 to 20 supporting articles, all internally linked. The advice is correct. The mistake almost every Australian service business makes is applying it at the wrong scope.
The pillar-cluster model was designed for B2B SaaS, where topics are horizontal and buyers search nationally. A service business in Highett trying to cluster around "plumbing services" is competing against directories with thousands of citations and Google's own local pack. Six months of cluster-building gains nothing, because the scope was wrong before the first article was written.
The shift: cluster at buyer-intent scope, not service-category scope. Not "plumbing", not even "Melbourne plumbing". The cluster is "blocked drains Highett 3190", with 15 supporting articles answering the questions a local resident types after a blocked toilet at 9pm on a Sunday. Hot-water-system failure modes by brand. Council compliance for stormwater connection. The local suppliers that stock common parts. That scope beats the national portal because the portal doesn't have local depth, and both Google and ChatGPT reward sources that own a specific niche over generalists who skim it.
UnderCurrent Automations' audit of 146 Australian service-business sites surfaced the same pattern repeatedly. Sites trying to rank for the category sat at page 3 with thin engagement. Sites with 30 to 50 hyper-local pages around one suburb cluster were getting cited in AI Overviews and Perplexity answers across the long-tail query set. Semrush's research on topical authority and Ahrefs' content depth analysis both confirm depth beats domain. Neither names the scope mistake.
What good looks like: a Brisbane physio building a cluster on "physio for runners in Paddington and New Farm", 15 articles covering injury types, the local running routes that aggravate each one, and clinic-specific recovery protocols. The cluster doesn't compete with Healthline. It owns the only query set that matters to a runner in that postcode.
Classical SEO and Answer Engine Optimisation both compound when the cluster sits at the right scope. UnderCurrent Automations plans cluster scope first, content second, as part of SEO & AI Visibility. Skip the scope question and the content itself won't save you.