If you've logged into Meta's Ads Manager recently, you've probably run into an insistent banner nudging you toward "Advantage+". Meta wants nearly everyone to use this type of campaign, and that's no accident: it benefits them as much as, in many cases, it benefits you. The problem is that "in many cases" isn't "in every case", and that's worth stopping to think about before clicking the button.
Advantage+ is Meta's name for a family of campaigns where artificial intelligence decides almost everything you used to decide by hand: who sees the ad, in which placement (Feed, Stories, Reels, Messenger), with which of the creatives you uploaded, and at what bid. You set the objective, the budget, the catalogue or the creatives, and the system learns on the fly who is most likely to buy, click, or hand over their details.
What problem it actually solves
For years, building a good Meta campaign meant creating dozens of ad sets: one per audience, another per age bracket, another per placement, each one competing for the same budget and fragmenting the data the algorithm needs to learn. With luck, after weeks of tweaking, you'd find the winning combination. The problem is that the very act of splitting the budget so finely meant each individual ad set received too little data, and a system starved of data learns slowly and poorly.
Advantage+ flips that logic: instead of splitting, it pools. A single ad set with access to a broad audience (or with no manual restrictions at all) receives the full budget and the full volume of data at once, and Meta's algorithm, which has seen billions of purchases across its platform, decides in real time who is worth showing what. For a store with decent sales volume, that usually translates into faster learning and, often, a lower cost per result than a well-built but fragmented manual campaign.
The two faces of Advantage+ worth telling apart
It isn't one single tool, it's several sharing the same surname, and it's worth not mixing them up. Advantage+ shopping campaigns (formerly automated catalogue campaigns) are built for ecommerce: you connect your product catalogue and Meta dynamically assembles the ads and decides the targeting. Advantage+ audience is a lighter layer you can switch on inside regular lead generation or traffic campaigns: you give it a starting audience as a suggestion, but allow Meta to step outside it if it finds better results elsewhere. They're cousins, not twins, and the amount of control you give up is different in each.
When it genuinely makes sense
Meta's automation performs better the more data it has to learn from. An online store already generating at least 30-50 purchase conversions a month per campaign is the natural candidate: with that volume, the algorithm has enough material to find patterns that would take a human manually adjusting things weeks of trial and error. We've seen cases with clients running catalogues of 200-300 products where switching from manual catalogue campaigns to Advantage+ shopping cut cost per purchase by 15% to 30% within the first two months, simply because the system stopped splitting budget between audiences that in practice overlapped.
It also makes sense when your team (in-house or agency) doesn't have the time to review and adjust campaigns every week. Advantage+ doesn't replace strategy, but it does reduce the day-to-day maintenance load, and for a small business without a full-time specialist that's worth real money, even if it never shows up as a line on an invoice.
When it's worth being wary
The first clear case is businesses with very low volume. If your campaign generates five or ten conversions a month, the algorithm doesn't have enough data to learn anything useful, automated or not, and in that scenario Advantage+ doesn't fix the underlying problem. The second case is businesses with a very specific, narrow target audience, for example a manufacturer of specialised medical equipment that only wants to sell to a particular kind of clinic. There, giving the AI full freedom to look for "whoever is most likely to buy" can end up showing your ad to people who match a purchasing pattern but aren't your real customer, because the system optimises for behaviour, not business judgement.
The third case, more subtle, is brands that need to control the context their ad appears in for image reasons. Advantage+ decides placements automatically, and although you can exclude some, you lose much of the fine control over "I want this to show on Reels but never on Audience Network". If that matters a lot to you, a manual campaign gives you more room.
How to test it without betting the whole budget
The sensible way to test Advantage+ isn't to switch off what already works overnight, but to run both in parallel for three or four weeks with comparable budgets and look at the real cost per result, not just the one Meta itself reports (which tends to claim credit for conversions that would have happened anyway). A useful trick: use Meta's native A/B test, which splits traffic randomly between the two variants in a statistically cleaner way than comparing two campaigns running separately over time.
Give each campaign at least one or two weeks before drawing conclusions. Meta's algorithm goes through a "learning phase" during which performance is usually worse than normal; judging Advantage+ by its first five days is the most common way to write off something that actually just needed more time to settle.
Mistakes we keep seeing
The most frequent one is uploading only two or three creatives and expecting the system to work magic. Advantage+ needs variety (several formats, several message angles, video and static) because part of what it optimises for is precisely which creative pairs best with which person. With little material, it has nothing to work with.
The second mistake is never checking the results breakdown by age, gender or location once the campaign has been running. Even though the system decides for you, the breakdown reports are still available, and it's worth glancing at them now and then: if 80% of your buyers turn out to be between 55 and 65 and your product actually targets people around 25, something in your creative or your offer is attracting the wrong crowd, and no automation is going to flag that for you.
The third mistake is treating Advantage+ as an excuse to stop thinking about the offer and the message. Automation improves how budget is split between people, not the quality of what you're offering them. A mediocre ad, automated, is still a mediocre ad, just shown more efficiently.
A practical case: from manual campaigns to Advantage+ at a home goods store
An online kitchenware store with a catalogue of around 150 products had spent a year and a half running eight separate ad sets on Meta, segmented by age, interest and product type. Performance was acceptable but stagnant: they'd gone months without getting cost per purchase below 18 euros, despite adjusting targeting every few weeks. The team tested Advantage+ shopping for six weeks, running it in parallel with a third of their usual budget while keeping the rest on the manual structure so they could compare.
The first two weeks were, as expected, worse than the already-settled manual structure: cost per purchase hovered around 24 euros while the system learned. From the third week onward it started stabilising, and by the sixth week the Advantage+ campaign's cost per purchase had dropped to 13.50 euros, 25% below what they'd been achieving for months with the manual structure. The most telling part wasn't just the final figure, but the buyer pattern: the system had found a segment of buyers aged 35 to 44 interested in interior decor that none of the eight manual ad sets had isolated as its own audience, simply because nobody had thought to cross those two specific variables.
The final decision wasn't to eliminate the entire manual structure, but to migrate 70% of the budget to Advantage+ and keep a small, very specific manual ad set for a repeat-purchase audience (customers who had already bought before), where fine control still delivered more than automation. That hybrid, rather than an either-or choice between one model and the other, is the pattern that has worked best in most of the cases we've managed.
Signs your account is ready to make the jump
Not every Meta account is equally prepared to get the most out of Advantage+ from day one. A favourable sign is already having at least 90 days of conversion data history on the account, because the system can draw on that accumulated history in addition to whatever the new campaign generates, speeding up the initial learning phase. Another favourable sign is having a well-structured product catalogue, with categories, attributes and prices correctly loaded, because Advantage+ shopping depends heavily on the quality of that data to decide which product to show to whom.
Conversely, a newly created account, with no conversion history, or with a poorly tagged product catalogue (generic categories, thin descriptions, low-quality images) will struggle to get automation performing well from the start, regardless of how well the campaign itself is built. In those cases, it's usually worth spending a few weeks first cleaning up the catalogue and building up some history with a simpler structure before making the full jump to automation.
Common mistakes when switching to Advantage+ (and how to avoid them)
The first mistake, already mentioned in passing, is worth expanding on: switching on Advantage+ and abruptly switching off the entire manual structure the same day, with no overlap period at all. This leaves the business with no reference data to compare against, and if initial performance is poor (to be expected during the learning phase), there's no way to know whether it's worse than before or simply going through the normal rough patch of any new campaign. Keeping the manual structure active, even with a reduced budget, during Advantage+'s first three or four weeks gives a real point of comparison.
The second mistake is setting expectations for immediate improvement based on other people's cases, without accounting for the fact that every account starts from a different history. An account that already had a very finely tuned, efficient manual targeting setup may see a modest improvement switching to Advantage+, while an account with poorly structured manual targeting may see a huge improvement. Comparing your own result with another business's, without knowing that other business's starting point, leads to mistaken conclusions in either direction.
The third mistake, more subtle, is not reviewing "Advantage+ creative" (the automatic generative enhancements to text and images Meta can apply) without understanding what it involves. This feature can automatically generate variations of your own images and copy, which works well for many brands but can drift from a carefully maintained tone or visual identity if left unchecked. It's worth turning this feature on deliberately, reviewing the output during the first few weeks, instead of leaving it on by default without checking what it's actually generating.
Frequently asked questions
Does Advantage+ work as well for lead generation as it does for ecommerce sales?
Not exactly. It was born in, and still performs best in, ecommerce with a catalogue, where purchase signals are clear and plentiful. For lead generation (contact forms, for example) it works, but it's worth watching lead quality closely, because the system can optimise for the volume of forms filled in without distinguishing well between a qualified lead and a curious one who will never buy.
Can I exclude audiences even while using Advantage+?
Yes, you can exclude your current customers, your own database, or very specific age ranges if you have legal or business reasons to. What you lose is the fine control over who to prioritise within the audience you do include.
Do I need less budget with Advantage+ than with manual campaigns?
Not necessarily less, but it usually performs better per euro, especially if your budget used to be split across too many small ad sets. If your monthly budget is very low (under roughly 300-400 euros), the algorithm probably won't have enough time or volume to learn well, automated or not.
Can I go back if Advantage+ doesn't work for my business?
Yes, at any point you can go back to creating manual campaigns with targeting and placements controlled by you. It isn't a permanent decision, and there's no exit cost beyond the time it takes to rebuild the structure.
Does Meta push Advantage+ because it benefits them more than me?
Partly yes: fewer manual decisions mean less friction and smoother spend flowing through the platform, which benefits Meta. But that doesn't invalidate the fact that, for businesses with enough volume, the result is also better for the advertiser. It's worth deciding based on your own data, not just on what Meta recommends by default.
What happens to creative control if I let the AI decide the combinations?
You're still the one uploading the copy, images and videos: Advantage+ combines and prioritises among the material you give it, it doesn't generate content on its own unless you explicitly enable the generative enhancements (like text or image variations), which you can also turn off if you'd rather keep full control over every piece.