The short answer: check in this order
- 1. Your measurement: are conversions being recorded properly?
- 2. The learning phase: does your ad set get enough conversions to learn from?
- 3. Your optimisation goal: are you asking Meta for buyers or for clickers?
- 4. Your angle and creative: does the ad speak to the right people for the right reason?
- 5. Your landing page: does it deliver what the ad promises?
- 6. Your offer: are the price, product and trust right?
- 7. Your audience and setup: is your audience too small or too fragmented?
1. Your measurement: is it really not converting, or can't you see it?
Always start here. If your store sees sales but Meta doesn't, the problem isn't your ads. A broken purchase event, a cookie banner that doesn't pass on consent properly or an error in the link between pixel and Conversions API can make conversions invisible. Read how to check this in ROAS dropped? Check your tracking first and Conversions dropped after your cookie banner.
2. The learning phase: enough conversions to learn from?
Meta's system learns from the conversions it sees. An ad set gets the status "Learning limited" when it's unlikely to get around 50 optimisation events in the week after its last significant edit (Meta Business Help). According to Meta, that's not a penalty, but a sign the system can't optimise well with the current setup.
Causes Meta lists: a small audience, a low budget or bid, a lot of overlap between ad sets, an optimisation event that rarely happens, or too many ads at the same time. Fixes: consolidate ad sets, broaden the audience, raise the budget, or temporarily optimise for a more frequent event.
Note: significant edits, such as changing the optimisation event, audience or creative, or pausing the set, restart the learning phase (Meta Business Help). If you change something every few days, the system never gets a chance to learn.
3. Your optimisation goal: buyers or clickers?
Meta looks for the people most likely to do what you ask for. Optimise for clicks or traffic and you get people who like to click, not necessarily people who buy. If you want sales, choose a conversion goal. If you don't (yet) have enough purchases for that, an event closer to the purchase, such as "add to cart", can be a stepping stone.
4. Your angle and creative: the right reason for the right people?
Lots of clicks and few purchases can mean your ad sparks curiosity, but not among people who actually want to buy. A strong angle attracts buyers, not spectators. Look at where it breaks down per layer: scrolling past points to the hook, watching without clicking to the concept, clicking without buying to the angle or offer. More in Angle vs concept vs hook.
5. Your landing page: does it deliver on the promise?
- Does the page match the ad? Someone who clicks on "finally no more back pain" should see that at the top of the page, not a generic homepage.
- Does it work well on mobile? Most ad traffic is mobile. Slow loading, small buttons or an awkward checkout cost sales immediately.
- Is there enough trust? Reviews, clear terms, familiar payment methods and a real sender.
6. Your offer: is the deal right?
Sometimes the ad is fine and so is the page, but the offer doesn't convince. Is the price logical compared to alternatives? Is it clear what someone gets? Is there a reason to buy now? A strong ad can't rescue a weak offer; it only makes it visible faster.
7. Your audience and setup: not too small, not too fragmented
Too many small ad sets with overlapping audiences compete with each other and each get too few conversions to learn from. A simpler setup with broader audiences and more variety in your creative often works better, because your creative then finds the right audience itself. Why that is, read our article on Meta Andromeda.
A typical example (for illustration)
A composite example, not a specific client case.
An online store runs five ad sets, each with its own interest audience, optimised for traffic, and tweaks something every few days. Plenty of visitors come in, but few purchases. They merge the sets into one with a broad audience, choose purchases as the goal, add three different angles and leave it alone for a week. The system finally gets enough conversions to learn from, and the visitors who come in are more often buyers.
Where it often really goes wrong
When ads don't convert, the reflex is to make new ads or cut the budget. Sometimes that's right. But often the problem sits earlier in the chain: in measurement, setup or the goal. That's why at SASDAMX we look at creative and measurement together, and start with the question of whether we can even see what's happening properly. Read more about how we approach breakthrough ads.