Short answer: In 2026, effective Facebook Ads targeting in Australia relies less on narrow interest-based audiences and more on broad targeting paired with strong creative, first-party data lookalikes, Advantage+ Shopping campaigns, Conversions API (CAPI) retargeting, and geo-targeted local campaigns - adaptations made necessary by Apple's App Tracking Transparency changes.
Facebook Ads targeting has changed dramatically since the iOS 14.5 privacy update. Many of the hyper-granular interest-based audiences that made Facebook advertising so powerful have been degraded or removed entirely. But businesses that have adapted to the new reality are still generating exceptional ROAS - they have just shifted their approach.
What Changed and Why
Apple's App Tracking Transparency framework allowed users to opt out of cross-app tracking. In Australia, roughly 80% of iPhone users chose to opt out. This meant Facebook lost visibility into much of its audience's off-platform behaviour, reducing the effectiveness of interest-based and behaviour-based targeting.
Targeting Strategies That Still Work
1. Broad Targeting with Creative Differentiation
Meta's algorithm has become remarkably good at finding your ideal customers - if you give it compelling creative. Rather than narrowing audiences to 50,000 people with specific interests, top advertisers are now using audiences of 1 - 5 million and letting the algorithm optimise delivery based on creative performance.
2. Lookalike Audiences from First-Party Data
Upload your customer list, website purchasers, or high-value leads and let Meta find similar users. First-party data lookalikes remain one of the most powerful targeting tools because they are built on your actual customer data, not third-party tracking.
3. Advantage+ Shopping Campaigns
For e-commerce businesses, Advantage+ campaigns use Meta's machine learning to automatically find the best audiences, placements, and creative combinations. Performance varies by account, so treat Advantage+ as a structured test against your current campaign structure rather than a guaranteed improvement.
4. Retargeting with CAPI
Server-side tracking via the Conversions API (CAPI), part of a well-managed Facebook Ads setup, recovers much of the data lost through iOS privacy changes. Businesses implementing CAPI alongside the pixel generally see more attributed conversions and better retargeting performance.
5. Geo-Targeted Local Campaigns
For local businesses, radius-based targeting remains highly effective, complementing broader digital marketing services. Target users within 10 - 30km of your location with locally relevant creative. This works particularly well for restaurants, retail, medical practices, and trade services across Australian cities.
Creative Is the New Targeting
In 2026, your ad creative does most of the targeting work. Meta's algorithm serves your ads to the users most likely to engage - but it determines that based on the content of your creative, not just your audience settings. This means investing in high-quality video, compelling copy, and diverse creative formats - a discipline our design and creative team supports - is more important than ever.
The Bottom Line
Facebook Ads targeting in 2026 is less about finding the perfect audience and more about creating the perfect message. The businesses winning are those combining broad targeting with exceptional creative and robust first-party data strategies.
Need help navigating the new targeting landscape? Our social media advertising team manages Meta campaigns for Australian businesses across every major industry. Book a free strategy call.
How to structure a Meta account when targeting does less of the work
The account structure that suits broad targeting looks different to the one most Australian advertisers built in 2019. Fewer ad sets, more budget per ad set, and creative variety carried inside the ad rather than across a dozen narrow audiences. Splitting a modest monthly budget across eight interest stacks starves each one of the conversion volume Meta needs to leave the learning phase, so delivery stays unstable and reporting stays noisy.
A workable shape for a service business spending a few thousand dollars a month is one prospecting campaign with a single broad ad set, one retargeting campaign covering site visitors and engagers, and a small testing campaign where new creative concepts run before graduating into prospecting. That is three moving parts, which is roughly what one person can review weekly without guessing.
Exclusions matter more than inclusions now. Exclude recent purchasers, current customers from your CRM list, and job applicants if you advertise roles. Those exclusions do more to protect efficiency than another interest layer ever will.
Getting first-party data into a state Meta can actually use
Lookalikes are only as good as the seed list. A list of every email address you have ever collected produces a lookalike of everyone vaguely near your category. A list of customers who bought twice, or leads that became quoted jobs, produces something far more useful. Segment before you upload: high-value customers, repeat customers, and qualified leads should be three separate sources, not one export.
Match rates depend on the fields you send. Email, phone in international format, first and last name, and postcode all improve matching. Australian phone numbers exported as local format with a dropped leading zero are a common reason a list matches poorly, and it is worth checking that once rather than assuming the platform handled it.
If you run a lead-generation model, the highest-leverage feed is offline conversion data — sending back which leads actually became customers, rather than optimising to form fills that your sales team quietly deletes. That closes the loop between ad spend and revenue and is usually the difference between a channel that looks fine in Ads Manager and one that holds up in the accounts. It is the same discipline we apply in lead generation work generally.
Privacy obligations that apply here in Australia
Uploading customer data to Meta is a disclosure of personal information, so it sits under the Privacy Act 1988 and the Australian Privacy Principles. The Office of the Australian Information Commissioner publishes guidance on the APPs, and the practical implication is straightforward: your privacy policy should disclose that you share hashed customer data with advertising platforms, and people should have a way to opt out. Check yours before your first upload rather than after a complaint. See the OAIC's Australian Privacy Principles guidance for the source text.
Creative testing without burning the account
Because creative now carries the targeting load, the testing process needs to be deliberate rather than opportunistic. Test concepts, not cosmetics. Changing a button colour or swapping a stock photo tells you nothing. Changing the angle — problem-led versus outcome-led, founder to camera versus customer footage, price-forward versus risk-reversal — tells you something you can reuse across every channel.
Run one variable at a time and give each concept enough spend to accumulate a meaningful number of conversions before judging it. If your cost per lead is around fifty dollars, a concept needs a few hundred dollars behind it before a difference is real rather than noise. Judging a creative on twenty dollars of spend is how accounts end up thrashing.
Keep a creative library with the hook, the format, the offer and the result for every concept you have run. Six months in, that library is more valuable than any audience you have saved, because it tells you which messages your market responds to regardless of which platform you run them on.
What to ask an agency about targeting
- How many ad sets will my budget be split across, and why that number?
- Which of my first-party lists are you using as lookalike seeds, and how were they segmented?
- Is the Conversions API implemented server-side, or is this pixel-only with a deduplication claim?
- What is the creative testing cadence, and who produces the assets?
- What happens to leads after the form fill, and does that data return to Meta?
- Which metric are we agreeing to be judged on, and over what window?
The last question separates operators from vendors. An agency comfortable naming a metric and a timeframe in writing is one that expects to be measured.
A worked example of the trade-off
Consider two set-ups spending the same monthly budget. The first splits it across six interest-based ad sets, each receiving a small daily amount, with one video and two static images reused in every ad set. The second runs a single broad ad set with the full prospecting budget and six genuinely different creative concepts rotating inside it. The second learns faster, because every conversion feeds one delivery system rather than six competing ones, and it produces cleaner creative learnings, because performance differences can be attributed to the ad rather than the audience.
The trade-off is control. Broad targeting means accepting that you cannot fully explain who saw the ad. For most Australian service businesses that is an acceptable exchange for stability, but for a business with a genuinely narrow buyer — a specific licence, a specific piece of equipment — narrow targeting still earns its place. Judge it on outcome, not on how the settings look.
Measurement in a post-attribution world
Platform-reported conversions and your own records will not agree, and chasing perfect reconciliation wastes time. Pick one source of truth — usually your CRM or accounting system — and use Meta's numbers directionally. Track cost per qualified enquiry rather than cost per click, watch the trend across weeks rather than days, and run occasional geo or spend-level holdouts if the budget is large enough to justify it.
Where paid social and search overlap, decide in advance how you will treat assisted conversions instead of arguing about them later. The relative strengths of each are laid out in Google Ads versus Meta Ads for Australian businesses, and the placement question specifically is covered in Instagram Ads versus Facebook Ads.
If you want the targeting and creative side handled together rather than in separate silos, that is how our social media marketing engagements are structured, with landing pages and post-click experience covered under conversion rate optimisation so the click has somewhere worth going.
