A compliant website rewrite changes far less of your site than you think, and far more of the part that sells. On 9 September 2026 we ran Australian healthcare advertising rules across 21 live clinic pages from four professions and five cities. Of 1,360 sentences, 105 carried a clear-cut compliance problem. That is 7.7%. But 86% of those 105 were one thing: a superlative. And superlatives live in your page titles and hero lines.
What did we actually measure?
Practice owners ask the same question before they sign anything. How much of my site are you going to change?
Nobody publishes an answer. So we measured one.
We took 21 live Australian clinic pages: six physiotherapy, five dental, four psychology and six chiropractic or podiatry. They span Melbourne, Sydney, Brisbane, Perth and Adelaide. Some are homepages, some are services or about pages. We fetched all 21 on 9 September 2026 with an ordinary browser user agent. All 21 returned a page.
Then we split every page into sentences and kept only those of five words or more. That rule is what drops menu items, buttons, phone numbers and address fragments out of the count. We also removed sentences repeated inside the same page, so a tagline in both a header and a footer is counted once. That left 1,360 sentences.
Each sentence was then checked against the term register that our own compliance tooling uses every day. We imported that register rather than retyping it, so the article and the working checklist cannot drift apart.
We split the results into two piles, and the split matters. Some categories are unarguable on a general clinic page: superlatives, outcome and cure claims, availability absolutes, rebate absolutes, and skip-your-GP framing. A rewrite changes those without needing to ask the practice anything. Other categories always need a human read, because the same word is innocent in one sentence and a problem in another. Every headline number below is the clear-cut pile only. The looser number is in the table.
How much of a clinic page does a rewrite touch?
| What we measured | Result |
|---|---|
| Live clinic pages checked | 21 |
| Pages that returned content to a plain request | 21 |
| Sentences of five words or more, deduplicated | 1,360 |
| Sentences with a clear-cut compliance problem | 105, or 7.7% |
| Sentences flagged including the judgement-call categories | 136, or 10.0% |
| Median page | 6.8% of its sentences |
| Cleanest page in the sample | 1.2% |
| Worst page in the sample | 31.2% |
| Pages with zero clear-cut problems | 0 of 21 |
| Pages above 10% | 8 of 21 |
Read the two ends of that table together. Not one of the 21 pages was clean. And yet the typical page needs about one sentence in fifteen touched.
Both facts are true at once, and clinics tend to believe only one of them. The practice that thinks its site is fine is wrong. The practice bracing for a rebuild is also wrong.
Does it differ by profession?
| Profession | Pages | Sentences | Clear-cut share |
|---|---|---|---|
| Physiotherapy | 6 | 306 | 12.4% |
| Dental | 5 | 328 | 10.4% |
| Psychology | 4 | 369 | 4.6% |
| Chiropractic and podiatry | 6 | 357 | 4.5% |
Physiotherapy and dental pages carry roughly two and a half times the rate of the psychology pages. The psychology pages in our sample were longer, more clinical in tone, and spent more words explaining a method than praising a team. That is a writing choice, not a rule, and any practice can make it.
Which lines does it touch?
This is the part that surprised us. It is not spread across the rulebook. It is one category.
| Category | Hits | Share of clear-cut hits |
|---|---|---|
| Superlative and specialist claims | 90 | 86% |
| Medicare and health fund absolutes | 14 | 13% |
| Treatment, outcome or mechanism claims | 9 | 9% |
| Availability absolutes | 1 | 1% |
| Skip-the-GP framing | 1 | 1% |
Shares add to more than 100% because a sentence can carry two problems at once.
Narrow it further and it is four words.
| Word | Times used | Pages using it (of 21) |
|---|---|---|
| best | 33 | 15 |
| specialist | 16 | 6 |
| expert | 11 | 6 |
| leading | 10 | 6 |
| Medicare rebate or rebates | 11 | 2 |
| elite | 5 | 1 |
| restore | 5 | 4 |
| top | 3 | 3 |
All 21 pages used at least one superlative. Every single one. Whatever else a compliant rewrite does, it starts here.
Why is it always the same four words?
Because they are the default vocabulary of every marketing template ever sold, and because two separate rules make them expensive.
The first is section 133 of the Health Practitioner Regulation National Law. It says a person must not advertise a regulated health service in a way that:
- is false, misleading or deceptive, or is likely to be misleading or deceptive
- offers a gift, discount or other inducement without stating the terms and conditions
- uses testimonials or purported testimonials about the service or business
- creates an unreasonable expectation of beneficial treatment
- directly or indirectly encourages the indiscriminate or unnecessary use of regulated health services
The maximum penalty is $60,000 for an individual and $120,000 for a body corporate.
An unverifiable superlative has to survive the first limb and the fourth. If you cannot show the working for best, it is hard to argue it is not likely to mislead.
The second rule is narrower and heavier. Section 115 makes it an offence to take or use a specialist title for a recognised specialty unless you are registered in that specialty. The maximum there is $60,000 or three years imprisonment for an individual. That is not a marketing rule. That is a title protection.
Sixteen uses of the word specialist across six of our 21 pages is not sixteen offences. Most will be loose writing rather than a claim to a protected title. But it is sixteen sentences that somebody has to read carefully, and that is exactly what a rewrite is for.
What does the rewrite look like, line by line?
Here is the actual work. Each pattern below is one we found repeatedly across the 21 pages, rewritten into a version that says more and claims less.
These are constructed examples, not quotations. They are written in the shape of what we found, because we do not name or quote the clinics in our sample. We are also not saying any page in the sample breaches a rule. That is a decision for a regulator on the facts, and it depends on evidence we do not have.
| The line we keep finding | What it engages | What we write instead |
|---|---|---|
| The best physiotherapists in Melbourne | s133(a) misleading, s133(d) unreasonable expectation | Musculoskeletal physiotherapy in Melbourne, six days a week |
| Our expert team of specialists | s115 title protection, s133(a) | Four physiotherapists, each with more than ten years in practice |
| A leading dental practice, trusted by thousands | s133(a), and an unsubstantiated volume claim | A general and cosmetic dental practice, open since 2009 |
| We will restore your smile and heal your pain | s133(d) outcome claim | Treatment aims to reduce pain and improve function. Results vary between people. |
| Medicare rebates available | Australian Consumer Law, an absolute where eligibility varies | A Medicare rebate may apply if you have a current care plan from your GP. We can check before you book. |
| Same-day appointments, no waitlist | Australian Consumer Law, availability absolute | We hold same-day slots for acute injuries. Call before 10am and we will tell you what is free today. |
| See us without a referral | s133(e) encouraging unnecessary use | You do not need a referral to book. If a rebate matters to you, a GP care plan is worth asking about first. |
Look at the right-hand column. Every replacement is longer and more specific than the line it replaces. That is the whole trick. The rule does not stop you saying good things about your practice. It stops you saying unprovable ones.
Does a compliant rewrite make the page bland?
This is the real objection, and it deserves a straight answer rather than reassurance.
The risk is real, and it comes from the wrong reflex. If you delete the best physiotherapists in Melbourne and replace it with quality physiotherapy services, you have made the page worse. You removed a claim and put nothing in its place.
Now look at where the flagged sentences sit. Superlatives are 86% of the problem, and superlatives do not appear in the middle of a paragraph. They appear in page titles, hero lines and headings. A rewrite touches 7.7% of the sentences and a much larger share of the sentences a visitor actually reads.
So the volume is small and the prominence is high. That is not a reason to be nervous. It is the reason the job is worth doing properly, because those are the lines that were carrying your offer while saying nothing checkable.
A practice has plenty that is both true and specific: how many years it has been open, how many practitioners it has, what happens in a first appointment, what it does not treat, how long a wait usually is. None of that needs a superlative, and all of it is more persuasive than one. We wrote more about that tension in why compliant copy gets called boring.
Why a scanner is not a rewrite
We should show you the limits of our own instrument, because this is where an automated compliance check goes wrong.
The scanner flagged 27 sentences for the term CBD, which is on the list because cannabidiol is a scheduled substance. We read all 27 by hand.
Every single one was the central business district. Melbourne CBD. Sydney CBD. Brisbane CBD. Perth CBD. Not one was cannabidiol. Twenty of those sentences had no other flag on them at all, so a practice acting on the raw output would have rewritten twenty perfectly good lines about where its clinic is.
The other false positives are just as instructive. One page was flagged for the word herbal. It was describing the herbal tea in the waiting room. Three pages were flagged for holistic, a word that carries real risk on a cannabis-adjacent page and none at all on a psychology page describing a scientist-practitioner model. And five mentions of AHPRA were flagged, of which several sat in exactly the place naming the regulator is appropriate: a line telling patients to consult a registered practitioner.
This is the difference between a list and a judgement. The list gets you to the 136 candidate sentences quickly, which is genuinely useful. Deciding that 105 of them need changing, and that 27 do not, is the work.
If you want to run the list yourself first, our AHPRA self-audit checklist walks through it in order, and the advertising guidelines explainer covers what each limb of section 133 actually means.
What this check does not show
Four things, and they matter.
Twenty-one pages is a sample, not a census. It covers four professions and five cities. It does not cover cosmetic, telehealth, weight or hormone clinics, which are the highest risk categories in Australian healthcare advertising and would almost certainly score worse.
We checked one page per site, not whole websites. A homepage is usually the most marketing-heavy page a practice has, so this may overstate the rate across a full site. It also means we saw none of the treatment pages, where outcome claims tend to live.
Keyword matching cannot see everything. It does not catch a testimonial that never uses the word testimonial, an implied comparison, a before and after image, or a missing disclaimer. Every one of those is a real compliance problem and none of them would appear in our 105. The true figure is higher than 7.7%, not lower.
We are not saying any page in our sample breaches anything. Section 133 turns on whether an advertisement is likely to mislead, judged on evidence. We have counted words on pages. That is the beginning of the question, not the answer.
