Pakistan
USA
Canada
UK
Dubai
Australia
Germany
- Advantage+ audience uses your inputs as suggestions, not boundaries.
- Location, minimum age and exclusions still work as hard controls.
- Restrict only for legal, geographic or genuine B2B reasons.
- Test the two set-ups with a proper A/B test, not side by side.
How does each option work?
With Advantage+ audience, you can give Meta suggestions such as an age range, interests, or a custom audience of past customers. Meta starts near those signals and then looks further if it predicts better results elsewhere. The system leans heavily on your conversion data, so the quality of your pixel and Conversions API set-up matters more than the interests you pick.
With the original audience options and detailed targeting, you define the audience yourself. Delivery stays inside it, although for some objectives Meta may still expand beyond your interests if Advantage detailed targeting is switched on. Check the setting in each ad set, because it is easy to assume you are targeting narrowly when you are not.
Audience suggestions vs audience controls
The difference that matters most is which inputs Meta can override.
Input | Advantage+ audience | Detailed targeting |
|---|---|---|
Location | Hard control | Hard control |
Minimum age | Hard control | Hard control |
Excluded custom audiences | Hard control | Hard control |
Interests and behaviours | Suggestion only | Limit (unless expansion is on) |
Custom and lookalike audiences | Suggestion only | Limit |
Gender and maximum age | Suggestion only | Limit |
When does broad targeting win?
Broad delivery tends to suit businesses with a clear conversion event and enough volume for Meta to learn from. Ecommerce stores, local services with simple enquiries and consumer apps are the usual fits. Here, narrow interests often just make reach more expensive without making buyers more likely. Broad does not mean careless; hard controls still apply.
Broad also helps when your audience is hard to describe with interests. A Lahore bakery's best customers may not have "baking" as an interest at all. Meta's system can find them from behaviour signals you cannot select. If you already use lookalike audiences, try them as suggestions inside Advantage+ audience rather than as the only audience.
When should you restrict the audience?
Keep tighter controls when spending outside a group would be wasted or not allowed. A clinic that can only serve one city should lock location. A B2B software firm selling to finance directors may need job-related signals, though detailed targeting by job title is often thin. Regulated categories such as housing, employment and financial products already limit targeting, so read about special ad categories first.
Restriction also makes sense when your conversion data is weak, for example a new pixel with few events or a lead form with no feedback from sales. Meta's broad delivery depends on knowing who converted, so it can drift towards cheap but poor results. In that case fix tracking first, because neither option performs well when Meta cannot see what a good result looks like.
How do I test them fairly?
Run a clean comparison instead of two ad sets competing in the same account.
- Use the A/B test toolSet up an A/B test in Ads Manager so audiences are split and do not overlap.
- Keep everything else equalSame objective, optimisation event, budget, creative and schedule in both arms.
- Pick one success metricChoose cost per purchase or cost per qualified lead before the test starts.
- Give it enough timeLet both arms leave the learning phase before reading results.
- Check lead quality tooFor lead campaigns, compare what the sales team does with the leads, not just the cost.
What does the creative have to do?
When targeting is broad, the ad does the targeting. The hook, the visual and the first line of text decide who stops scrolling. A wedding photographer should show couples and venues, not generic camera shots. A B2B firm should name the role and the problem in the first frame. Weak creative cannot be rescued by settings.
Give Meta several genuinely different concepts so it can match each one to a different pocket of people. A problem-led video, a customer review and a clear offer will each attract a slightly different buyer. One concept per ad set leaves the system with little to work with, however good the audience settings are, and it tires faster.
Common mistakes with audience settings
These are the problems we find most often when reviewing ad sets.
Stacking many interests
Ten narrow interests rarely beat a broad audience and usually raise costs.
Forgetting exclusions
Existing customers keep seeing acquisition ads when no customer list is excluded.
Assuming targeting is narrow
Expansion settings can be switched on by default, so delivery is wider than you think.
Comparing in the same account
Two live ad sets compete for the same people, so the result is not a fair test.
Changing settings weekly
Each change can restart learning, which hides what is actually working.
How we handle this for clients
We start every account by checking tracking: pixel, Conversions API, event deduplication and which event the campaign optimises for. Next we review audience settings ad set by ad set, including whether expansion is switched on without anyone noticing, and whether customer lists are excluded where they should be. This tells us whether a change of targeting is likely to help or whether the real problem sits elsewhere.
If data is healthy, we usually move to Advantage+ audience with only the controls the business genuinely needs, such as service area and customer exclusions. We then run a structured test against the existing set-up and decide on the metric agreed in advance. Creative planning runs alongside, because broad delivery depends on it. Ongoing work is covered under our Meta Ads management service.
Pakistan, Gulf and UK advertisers
In Pakistan, most buyers are on mobile and many prefer to message rather than fill in forms. Broad audiences pair well with WhatsApp or Messenger destinations, as long as the ad language matches the audience. Roman Urdu text often reads more naturally than formal Urdu script for younger buyers. Test both scripts if unsure.
In the UAE and Saudi Arabia, a large expatriate population means language is a better filter than interests. Run Arabic and English ads separately and use language as a control where relevant. In the UK, location controls are useful for local service areas, and regulated sectors such as finance follow UK advertising rules on top of Meta's.
Junaid has more than 18 years in SEO, paid media, web development and custom software. He founded SEOnumber1, builds WordPress plugins, CRMs and web applications, has served 500+ clients in 48 countries, and holds Google Ads, Analytics and Google Blue and Green Belt certifications. He is a National Level Award Winner in Digital Marketing. Read his story or ask a question about this article.