Google Ads in Australia: The 2026 Operator's Guide

The honest summary

Australia is the market where good account structure is worth the most and gets the least attention. It is small by population, concentrated into a handful of coastal cities, and dominated in most categories by a short list of large national players who all advertise. Auctions are dense and clicks are expensive, and the usual levers — bid tweaks, ad copy rounds, another negative keyword — move less than the two things nobody wants to spend an afternoon on: where your ads are allowed to serve, and whether your campaigns are large enough to learn from.

I have run Australian accounts. That is worth stating plainly, because most of what is published about this market is written from somewhere else and it shows in the advice. The practitioner sections below reflect what I have watched go wrong in Australian accounts; where I have relied on published research instead, I say so.

Why Australian clicks cost what they do

The structural explanation is concentration. Australia has roughly the population of a large American state spread across a continent, and in most consumer categories that population is served by a small number of national brands plus a long tail of local operators. Two or three grocery groups. A handful of banks. A short list of insurers, telcos and comparison sites. Those companies have real budgets and they are all bidding.

So the top of the results page fills with well-funded advertisers more consistently than in a larger, fragmented market where demand is spread across hundreds of regional competitors who are not all in every auction. Fewer players does not mean less competition. In paid search it usually means more, because every one of the few is serious.

On the one published cross-country index that exists, Australia ranks among the most expensive markets measured, close to United States levels rather than comfortably below. That index is built from Keyword Planner forecasts, so it is directional rather than observed — but the direction matches what an Australian account feels like from the inside, and it cuts against the assumption that the United States is in a cost class of its own.

The benchmark data does not exist

If you have looked for Australian cost benchmarks you have probably found a table of cost per click and cost per lead by industry, on an Australian domain, in Australian dollars. It is not Australian data.

Tracing these, the pattern was consistent: the same industry categories in the same order, the same industry count as a well-known United States study, and figures matching that study line for line with the currency symbol swapped. The publisher of that American report has no Australian edition. Nobody does. The tables are a re-badge.

The second source is worse, because it is not even a re-badge. A programmatic template running on at least one domain produces a country-by-industry benchmark page for a long list of countries, generated from Keyword Planner forecasts and refreshed on a schedule — its own recoverable description of its method says as much. Several of those pages return a 404 when fetched directly, so what circulates is a search-result summary of a page that no longer resolves.

I documented this whole mechanism for another market where the same generator operates. The Nepal cost investigation shows the generator, the 404s and the citation chain in detail. Australia is the more surprising case, because it is a mature, well-funded market with active industry bodies — and still nobody has published an observed-spend benchmark study.

Geography is the setting that decides everything

Australia is the extreme case for location targeting among the markets I work in. The population sits in a few coastal cities separated by distances that make service delivery impossible rather than merely inconvenient — Sydney to Perth is further than London to Moscow. A radius that looks generous on a map of a European country covers most of a state here, and half of it is uninhabited.

Two settings do most of the damage, and both are defaults.

1. Location options. Google offers a choice between serving to people present in or regularly in your targeted locations, and serving to people present in or interested in them. The broader one is the default and it lets in people researching your city from elsewhere. For a service business with a van that is not incremental reach, it is a different audience.
2. Radius drawn around the office rather than the service area. The office is where you sit; the service area is where you will drive on a Tuesday in traffic. Different shapes, and only one belongs in the account.
3. Excluded locations left empty. With this much empty space between markets, exclusions do more work than inclusions. If you will not travel past a point, say so in the account rather than filtering it by hand later.

The related decision is whether a business genuinely serves nationally or is four city businesses wearing one brand. If delivery, pricing or capacity differs by state, the account should reflect that. If not, splitting by state divides the conversion signal for no benefit — which is the problem that follows every small market.

Budget starvation, the small-market failure mode

This is the failure I see most often in Australian accounts, and it almost never gets diagnosed correctly.

A national account gets split by state, because states feel like the natural unit. Then by service line, because reporting is cleaner. Then brand and non-brand are separated, correctly. Then someone adds a campaign for the high-value service and another for remarketing. Twelve campaigns now share a budget that would have supported three.

Each is getting a handful of conversions a week, or fewer. Automated bid strategies need volume to learn from, and below a certain threshold they do not degrade gracefully — they spend unevenly, chase noise, go quiet for days and then burn a week of budget in an afternoon. The account looks like it has a bidding problem. It has a structure problem.

Australia produces this more reliably than larger markets because there is less demand to go around. The correction is consolidation, which is unpopular because it makes reporting less granular. That is the trade: fewer, larger campaigns the system can learn from, with granularity moved into ad groups, labels and segments where it costs nothing. My bid governance framework sets out what has to be true before handing bidding to an automated strategy, and the volume prerequisite is the one Australian accounts fail most often.

Seasonality runs the other way

Imported playbooks get the Australian calendar wrong in two specific ways, and both are expensive.

The first is obvious once said. Christmas falls in summer, so the retail peak, the school holidays and the long shutdown from late December to late January land on top of each other. Campaigns needing human handling — quotes, bookings, callbacks — should be planned around reduced capacity rather than increased demand. For home services and outdoor trades, summer is peak delivery season and winter is not, the reverse of what most seasonality advice assumes.

The second matters more for business-to-business. The Australian financial year ends on 30 June, which concentrates business purchasing, capital expenditure, software renewals and end-of-financial-year retail promotion into May and June. If your account calendar was written where the fiscal year ends in December, your budget weighting is inverted for the two months that matter most to a large slice of Australian commercial buying.

The features Australia does not get

Feature parity is high. Merchant Center, Shopping, free product listings, local inventory ads, call reporting with forwarding numbers and lead form assets are all available, and there is no Comparison Shopping Service requirement of the kind the United Kingdom has. The real absences:

The Local Services absence quietly invalidates a lot of imported advice. In the United States it is often the single best channel for a home services business, and American content treats it as the obvious first move. An Australian plumber, electrician or law firm reading that content is being told to use a product they cannot access. The Australian equivalent is a properly maintained Google Business Profile plus Search with tight geography — more work, and less automatic.

Microsoft Ads: a desktop conversation or nothing

StatCounter's June 2026 figures put Bing at around nine percent of tracked Australian search referrals overall — a respectable-looking number that has launched many a diversification proposal. On mobile the same data puts it around half a percent. That is the steepest desktop-to-mobile collapse of any market I have compared, and it makes the headline number close to useless as a planning input on its own.

Read those as tracked search referrals from a pageview sample, not a measure of how many people use each engine. The rule that falls out is simple: if your Australian traffic is majority desktop — much business-to-business, some professional services, considered purchases researched at work — Microsoft Ads is worth a test. If it is majority mobile, as most consumer, local and phone-led categories are, you are buying a rounding error and a second platform to maintain.

The metric that lies

In Australian accounts the metric that lies is impression share, and specifically the instinct to chase it.

Because the market is small and the competitor set is short, impression share feels controllable in a way it never does in a market with a thousand competitors. Losing it to a named national brand feels like losing. So budgets rise and bid caps loosen, the gap closes, and cost per acquisition climbs quietly while everyone congratulates themselves on the coverage.

The second number that catches it is cost per qualified lead over the same period, ideally with sales-accepted status fed back into the account. Impression share is a coverage measure, not an efficiency one, and the share you are missing is frequently the share a competitor is overpaying for. In a market with few bidders the marginal impression is expensive precisely because the large competitor above you decided to own it — and matching that is a choice to fight on their terms with your budget.

The prerequisite for any of this is that your conversions are counted correctly in the first place. The conversion tracking audit walks through deduplication, counting settings and the difference between a tag firing and a real event. Every efficiency argument above collapses if the denominator is wrong.

Privacy and consent, operationally only

Australia has a federal privacy regime overseen by a national privacy regulator, and a separate anti-spam regime overseen by the communications regulator. The pairing is the opposite of what people expect, and it is the useful thing to know.

On the tracking side, Australia has no dedicated cookie-consent statute of the sort Europe has. That does not mean nothing applies — the privacy regime still governs how personal information is handled, and the regulator has been active on tracking technologies — but the account-side consent plumbing is a different conversation from a European build. Google also does not extend its own regional user consent policy to Australia, so Consent Mode obligations that apply elsewhere are not imposed here by the platform.

On the email and messaging side the requirements are real and enforced: consent, accurate sender identification, and a working unsubscribe honored promptly. For a funnel that ends in a lead form and a nurture sequence, that is where the operational work belongs — consent captured at the form, stored on the record, visible to whoever sends the next message, and unsubscribe handling that removes the person from every list rather than one.

One structural point worth knowing rather than assuming: Australia has a turnover-based carve-out that puts many smaller businesses outside the federal privacy regime entirely, unlike most comparable markets. Whether it applies to you is not something to work out from a blog post.

Honest scope: when Australian paid search is the wrong tool

Three situations where I would say no on a call.

• When the budget is too small for the category. In a concentrated, expensive market an underfunded campaign in a competitive vertical does not produce a small result, it produces no result, because you never accumulate enough data to improve. Below that threshold, pick a narrower geography or service and own it, or spend the money elsewhere.
• When the service area is nominally national but delivery is not. Australian distances make this a real constraint rather than a logistics detail. If you cannot service Perth, do not advertise in Perth, however good the search volume looks.
• When the offer only works at a price the market will not pay. A small number of large incumbents set price expectations in most categories, and paid search will find you people comparing you to them within days. That is useful information, and it is expensive information.

And a sequencing note rather than a no: if the landing page has not been looked at, fix it before touching the account. In an expensive market a weak page compounds faster, because every wasted click cost more. The landing page work is dull and it is where the leverage sits when clicks are not cheap.

What I bring and what to verify

The practitioner judgments come from running Australian accounts, within a book of work across seven or more countries. Across those markets a lead-to-customer close rate of around twelve percent of generated leads has been typical, lower for consultancies — a cross-market figure from my own accounts, not an Australian benchmark and not something to plan against.

The platform claims come from Google-published country lists I retrieved and read: the Merchant Center supported countries table, the local inventory ads page, the call reporting eligibility list, the jurisdictional surcharge page and the Local Services country selector. Market shares are StatCounter's June 2026 tables — tracked referral shares from a pageview sample, not audience measurement. Market size figures are IAB Australia's press releases, not the gated report.

If you want the method rather than the country specifics, the Signal-to-Revenue Framework sets out how I sequence measurement, structure and bidding in any market; everything above is exceptions applied on top of it. I run that practice through Arcetis, and this guide leads with geography and structure rather than with tactics because in Australia those are the decisions that set the ceiling on everything else.

Related reading