A short explanation of what's changed about search, and why the new site is organised around topics and relationships rather than a list of pages. Three minutes.
For most of the last twenty years, ranking meant putting the right words on the page. If someone searched “HR consultant Melbourne”, you wanted a page that said “HR consultant Melbourne” a reasonable number of times.
That is no longer how it works. Google, and now ChatGPT and its equivalents, don't match words — they try to work out what things are and how they relate to each other. A person. A business. A service. A piece of legislation. Each of those is a thing, and the connections between them are what the system is actually reading.
The industry calls these things entities, and the approach entity SEO. The name is unhelpful; the idea is simple. Instead of hoping the right words appear on the right page, you state plainly what each page is about and what it connects to — and you keep those statements consistent everywhere.
“Does this page contain the phrase someone typed?”
Answered by putting keywords on pages. Worked until everyone did it.
“What is this page about, who wrote it, what do they do, and can I trust them on this subject?”
Answered by structure — the relationships between your pages, stated clearly enough to follow.
Take a real question one of your clients would ask. Today, someone searching it might land on one of your blog posts, read it, and leave — because nothing on that page connects the problem they have to the business that solves it.
Every link in that chain is stated on the page — visibly, for the reader, and invisibly in the page's code, for the machines. Which means when someone asks an AI assistant “who helps Australian employers with unfair dismissal?”, there is a path it can follow from the answer back to you.
Right now that path doesn't exist. Your highest-demand page was found 41,046 times in search and clicked five times, and not one other page on your site links to it. The information is there; the connections aren't.
If there is one rule to take from all of this, it is that rule.
When two of your pages answer the same question, they don't add up — they split. Google sees two candidates, isn't confident which to show, and shows both weakly. You end up competing with yourself.
It is happening across your site right now. Two posts asking whether reference checks are worth it. Three separate interview guide templates. Sixteen pages describing what HR Gurus does. Eight pages thanking someone for filling in a form. 276 addresses doing work that 23 pages should do.
The clearest example is the expensive one: your homepage and your Melbourne HR consulting page both rank for “hr consulting melbourne”. One sits around position 12, the other around position 74. Between them they were found 20,955 times and clicked eleven.
To connect things consistently, you need an agreed list of what the things are. Not categories someone invented in a planning meeting — the actual concepts your industry runs on, named the way the rest of the world names them.
That list is called an ontology. It's a map of a field: the topics it contains, and how they sit inside each other.
Every one of those is a real thing with a name Australian employers already use. When your content is filed against them, a search engine doesn't have to guess what your article is about — the article says so, and says what it sits underneath.
Twelve areas, named the way the Fair Work Ombudsman and Safe Work Australia name them: modern awards, the National Employment Standards, unfair dismissal, general protections, long service leave, workers compensation, enterprise agreements.
It covers everything worth keeping — all 290 articles, all 34 downloads and all 16 HR Advice answers have a place to sit, and so does every service you sell.
Five hundred blog posts with no structure are five hundred separate attempts to rank. The same posts organised under twelve topics, each linking up to its topic page, become twelve substantial bodies of work — and the topic page inherits the strength of everything beneath it.
A topic page that explains a problem and doesn't offer a download or a service is a page that reads well and does nothing. In the new structure every topic has both.
Your templates currently sit under headings almost nobody visits, while the topics that draw thousands of readers have nothing to download. Filing both against the same map fixes that automatically.
People increasingly ask ChatGPT rather than Google. Those tools rely heavily on being able to follow relationships — this business, these people, this expertise, this evidence. It's already happening: AI assistants sent 313 visits and eight enquiries in sixteen months, converting at roughly fourteen times the rate of Google search.
The topic structure suggests twelve topic areas, what sits under each, which of your four services each one leads to, and where the gaps are. Depending on how you want the topic pages to work, this may use one of the additional templates your agreement already allows for. Nothing extra to approve or pay for — but it is your call how these pages should work, so we will put the options to you rather than deciding for you.
What's worth your eye is whether the twelve areas we've suggested are the right ones: whether they match the problems clients actually bring you, whether anything is missing, and whether anything is there that you'd rather not be known for.
Once the structure is settled, the third document works through what happens to every existing page — and the decisions we need from you before Thursday.
Anything at all — a section that doesn't sit right, a question, or a straight yes. It comes to us with the page you were on, so there's no need to explain where you are.