Google is not sending patients to your website

We pulled 653 radiology keywords for Australia, June 2026. Together they account for about 404,500 searches a month.

The interesting part is not the total. It is that the largest, highest-intent slice of that demand never reaches a practice website at all.

Nearly 30% of demand is “near me”

49 queries carry “near me”, and they total 118,560 searches a month. That is 29.3% of everything: 23.8% with no brand attached and 5.5% naming a brand.

Two things worth noticing before the obvious one.

X-ray leads the modalities. Grouped by modality, “near me” demand runs x-ray 25,590, MRI 16,330, CT 7,210, ultrasound 5,600, DEXA 4,400. The largest local demand is for the cheapest study, which is worth remembering before anyone optimises the whole funnel around MRI revenue.

The phrasing is still spreading. “mri centers near me” is up 1,700% year on year and “x ray services near me” 1,157%. Small numbers, steep curves. Patients are converging on this grammar, not drifting away from it.

What Google does with a “near me” query

This is the part that changes the conclusion.

A “near me” search does not return ten blue links. It returns a map with three businesses, each with a star rating, an address, and buttons: Call, Directions, Website.

Three consequences follow. None of them make your website less important, and all of them change what it is for.

What the patient sees is your Business Profile, not your site. The local pack is ordered on proximity, relevance and prominence, Google’s own three factors, where prominence leans heavily on reviews and profile completeness.

Your website is still what earns that placement. This is worth being precise about, because it is easy to get backwards. Google crawls the site linked from your profile to judge whether you are relevant to the query. A practice whose site never says the words “bulk billed” or “ultrasound” is a weak candidate for “bulk billing ultrasound near me” however good the profile looks, and in a competitive metro market you do not rank one without the other. The site is not the destination. It is the evidence.

The most common action is not a website visit. It is Call or Directions. The patient taps the phone number inside the search result and your line rings, having never seen a page you control.

Increasingly there is no click at all. Questions such as “bulk billed MRI near me” are the kind an AI summary answers directly, by reading the indexed text on your domain. If your site states plainly that you bulk bill MRI, you can be summarised and cited without anyone visiting. If it does not, you are absent from the answer and no analytics package will explain why.

Being quoted and being booked through are separate problems. The first needs readable text. The second needs a booking page an agent can actually operate. Most practices fail the first and never discover the second.

So for the single biggest block of demand in this dataset, Google is not a source of web traffic. It is a source of phone calls. The website earns the placement; the handset at reception converts it.

Branded search is 41%, and it is yours to lose

Ninety of the 653 queries name a practice or network, and they total 166,510 searches, 41% of all volume.

Unbranded search is winnable. Branded search is only losable. Those people typed your name, often because a referrer said it, and every one you fail to convert is a patient you already had.

Branded queries do not return ten links either. They return a knowledge panel with your phone number attached. Again: the phone.

The branded query with “near me” attached

22,200 searches a month combine a brand with “near me”. Look at what the phrase assumes.

A patient adding “near me” to a brand is treating that brand as a set of interchangeable sites. They are not deciding whether to use that provider. They are shopping between its branches for whichever suits their Tuesday.

The brand-plus-suburb queries say the same thing more plainly: niddrie x ray 2,900, exact radiology springfield 1,900, kippax medical imaging 1,900, lumus imaging caboolture 1,900, lumus imaging kawana 1,300, qscan upper coomera 1,000.

That is a learned behaviour, and it does not stop at the brand boundary. A patient who has worked out that locations are interchangeable inside a network is one step from working out that they are interchangeable between networks.

Putting the two figures together

The 29% and the 41% are two lenses on the same list, and they overlap: 22,200 searches are both. Split into buckets that do not double count, the demand looks like this.

• Unbranded and local ("near me", no brand) 23.8% 96,360
• Branded and local (brand + "near me") 5.5% 22,200
• Branded, not local 35.7% 144,310
• Everything else (generic, modality and clinical terms) 35.0% 141,630

Google Ads Keyword Planner, Australia, June 2026. 653 keywords.

Read it as two numbers and the picture is plainer: 23.8% of this demand is somebody who has not chosen a practice and is standing in a map pack, and another 5.5% has chosen a brand but not a branch. Together they are the 29.3% of searches that carry “near me”.

Because referrals are portable, and patients know it

In Australia a referral for imaging can generally be presented at any practice that offers the study. There are real exceptions: a named radiologist, a study only one provider runs, priors held somewhere specific. As a rule though, the paper is not a commitment.

Practices have known this for years and mostly behaved as though patients did not. That assumption is expiring.

A referral is a lead, not a booking. It is yours only until somebody makes it easier to be somewhere else.

Cost is a front-of-mind question, and it is unanswered

12,630 searches a month are about money or access, which is a question about Medicare eligibility and bulk billing as much as about price: bulk billing ultrasound near me 4,400, mri bulk billing 4,400, bulk billed mri near me 2,400, bulk billing radiology near me 1,300.

Note where the competition sits. The tiny queries carry the highest commercial pressure: “private x ray near me” has a competition index of 92, “x ray without referral near me” 71, against 38 to 45 for the bulk-billing terms. Somebody is bidding hard on patients who want to skip the referral entirely.

A patient asking whether a study is bulk billed is asking a question your website usually answers with “conditions apply” and your phone answers correctly. That is another vote for the handset.

The demand mix is moving, and DEXA is the tell

DEXA and bone densitometry account for 49,500 searches a month, and they are climbing fast.

Meanwhile “radiology” as a bare term is down 18% year on year, and chest radiography and chest radiology are both down 19%. Side by side, the rotation is the story rather than either number.

dexa scan near me 4,400/mo

up 83 per cent +83%

dexa scan 22,200/mo

up 50 per cent +50%

dxa scanning 22,200/mo

up 50 per cent +50%

radiology 27,100/mo

down 18 per cent −18%

chest radiography 4,400/mo

down 19 per cent −19%

chest radiology 4,400/mo

down 19 per cent −19%

−90%0+90%

Excludes 'dexa bone density scan' (+240%): at 170 searches a month the percentage is noise, not a trend.

Demand is not growing evenly. It is rotating toward bone density while the generic terms soften. If your DEXA capacity, rostering and slot pools look the same as they did last year, this is the number to take to the next operations meeting.

What this dataset does not tell you

Any keyword report is a sample dressed as a market. This one included.

Nearly a third of the rows are not patients. 196 of the 653 keywords are clinical or professional lookups: radiopaedia articles, FRCR, PACS and RIS, teleradiology, and several hundred pathology-plus-radiology searches made by registrars.

Be careful which denominator that is: they are 30% of the keyword variations and 2% of the search volume. The 29% and 41% figures earlier are shares of volume, so this number is not competing with them. Anyone quoting a keyword count rather than a volume is misleading you, possibly by accident.

79 rows have zero Australian volume, mostly American practice names the tool volunteered.

The long tail is bigger than the head and is not in here. Every suburb, every body part, every combination of modality and funding.

And only one query in 653 shows booking intent by name: “benson radiology bookings”, 880 a month. Patients do not search for booking pages. They search for a practice, or for a category near them, and they expect the booking to be possible once they arrive.

Treat these numbers as direction, not as size.

Where this actually lands

Put the three big findings together.

Twenty-nine per cent of demand is local and answered with a map and a call button. Forty-one per cent is branded and answered with a knowledge panel and a call button. A referral does not bind the patient to you. And the fastest-moving demand is in a modality most practices have not re-rostered for.

Your website has work to do, and it is not the work most practices think. It has to convince Google you are relevant. What it does not do is take the booking, because the patient never arrives on it. That part is a capacity and answering problem wearing a marketing costume.

You can buy more visibility. Rankings, ads, listings, review campaigns, all of it works.

Notice what you are buying, though. Because the map pack surfaces a call button rather than a link, spend in this channel does not produce web traffic. It produces telephone traffic. Every dollar that lifts you into the top three rings a phone somebody has to answer, so a reception desk already at capacity converts your marketing budget into hold music.

What you cannot buy is attendance. If a fifth of inbound calls are lost to engaged tones and voicemail, more visibility raises your cost per scan, because you are paying to pour patients into a container with holes in it.

Fixing the container is cheaper than buying more water, and nobody sells it to you, which is probably why it stays broken.

One argument almost nobody makes

Practices compete on distance and price, and the strongest reason for a patient to stay put barely gets mentioned.

Prior imaging. A patient who changes provider leaves their comparison studies behind. A radiologist reading a follow-up without the prior is working with one hand tied, and the patient wears that.

It is a clinical reason to come back to the same practice, it is true, it matters, and it is almost never said on a website or by a receptionist. Practices are fighting on convenience while holding the better argument.

Five things to do this week

Search your own specialty plus “near me” on a phone, on mobile data, and count how many taps it takes to reach a human. Do it on your competitors too.

Audit your Google Business Profile properly, not just the review count. Three things carry weight: the primary category, since a radiologist and a diagnostic imaging centre are different categories and Google treats them differently; secondary categories for each modality you actually run; and the Q&A section, which you are allowed to seed yourself. Given that 12,630 searches a month concern funding, a question and answer about bulk billing belongs there, and the same language belongs in the text of the site behind it.

Ring your own practice at 8:05 on a Monday and time it.

Pull your DEXA volumes for the last four quarters and put them beside that +50%.

Say the prior-imaging thing out loud. It costs nothing and no competitor is using it. A line on the homepage and in the hold message: returning patient? Keeping your follow-up scans where your prior images are held gives the radiologist a comparison. It is true, it is clinically useful, and it is the only retention argument here that does not depend on answering the phone faster than anyone else.

The search data says the decision is still open when the patient starts looking. Whether it closes in your favour has very little to do with search.

Most practices read that 27,100 as demand still to be won. A good share of it is not new demand at all. When your phone rings out, or the hold music starts, the patient opens another tab and types “radiology near me”.