Meta shipped more than 47 policy changes between January and April alone. Two of them take control away from you. One of them hands a real advantage to stores that were previously too small to qualify.
The ability to exclude Audience Network was, for years, the first thing an experienced buyer did on a new account. That lever is gone. You cannot block Audience Network, you cannot block Messenger, and you cannot carve out specific Reels placements. Distribution is decided for you.
The honest read is that most placement exclusions were superstition carried forward from 2018. Audience Network genuinely was a junk-traffic problem once. It has not been that for a while, and the accounts that kept excluding it were usually optimising against a memory.
What you lose is not performance, it is diagnosis. When results move you now have one fewer variable to inspect, which makes creative and offer testing more important rather than less.
Weekly conversions needed to qualify for Advantage+ Shopping
Halving the requirement brings a large band of mid-size stores into a campaign type that was previously out of reach.
| Measure | Value |
|---|---|
| Previously required | 50 |
| Required now | 25 |
Lookalike audiences are being phased out as a default, replaced by predictive models working from aggregated behavioural signals. If your account structure was built on a stack of one percent, three percent and five percent lookalikes off a purchaser list, that structure has an expiry date.
The replacement is not worse. It is less legible. You cannot point at an audience and explain why it was chosen, which is uncomfortable if you report to someone who wants to see the targeting rationale.
The practical consequence is that your first-party data matters more, not less. The models still need something to learn from, and a clean, well-segmented customer list is the highest-quality input you can give them.
Advantage+ Shopping previously needed 50 conversions a week to qualify. It now needs 25. That single number quietly opens the campaign type to a large band of stores that were told for two years they were too small for it.
For a store doing 25 to 50 orders a week, this is the most consequential Meta change of the year, and it arrived with far less noise than the placement story.
Worth saying plainly: qualifying is not the same as benefiting. A store with thin creative and a weak offer will get the same result faster. The campaign type rewards volume of good creative, and that is still the constraint for most brands.
We stopped rebuilding lookalike stacks and started spending that time on first-party list hygiene, because that is the input the predictive models actually use. Segmented by recency and value, not one undifferentiated purchaser file.
We moved every store in the 25 to 50 conversion band onto Advantage+ Shopping tests rather than waiting for them to grow into the old threshold. Several were sitting just under 50 and had been told to wait.
And we front-loaded creative volume, because with placement control gone and audience control going, creative is most of what is left that you actually decide.
The pattern across every change is the same. Meta is taking back the dials and keeping the outcomes. That is uncomfortable if your value was in knowing which dials to turn, and it is good news if your value is in the offer, the creative and the customer data behind it.
Read more on ecommerce advertising. Read more on paid advertising management. Read more on how to structure a Meta campaign.
Sources: Meta Ads product and policy announcements, 2026; Common Thread Collective, Meta Ads change log 2026.
We audit ecommerce accounts against what changed this year, including whether you now qualify for campaign types you were told to wait for.
