If you mean Meta Ads, the short answer is:
For most ecommerce brands with enough conversion volume, Advantage+ Sales (formerly Advantage+ Shopping) is the better default in 2026. Manual campaigns still win in specific situations where you need control that Meta's automation can't reliably provide.
Meta has actually moved further in this direction: Advantage+ Shopping was renamed Advantage+ Sales, and Meta describes it as its end-to-end AI solution for sales, automating audience, placements, budget and other delivery decisions. Meta reports a 9% average improvement in cost per conversion for Advantage+ Sales campaigns, although that's Meta's own aggregate data rather than an independent benchmark.
My practical hierarchy
| Situation | I'd start with |
|---|
| DTC/ecommerce, broad appeal, 50+ purchases/month | Advantage+ Sales |
| Established brand with lots of conversion data | Advantage+ Sales |
| Scaling a proven product | Advantage+ Sales |
| New account / very little conversion data | Manual or hybrid |
| Very narrow geographic audience | Manual/hybrid |
| Strict prospecting vs retargeting separation | Manual/hybrid |
| Need hard audience exclusions/control | Manual |
| Highly different customer segments with different economics | Manual/hybrid |
| B2B/high-consideration purchase | Usually manual/hybrid |
| Testing a specific audience hypothesis | Manual |
| Creative testing | Either — usually keep delivery automated |
Why Advantage+ is winning more often
The biggest change isn't simply "better targeting." It's that Meta has more freedom to decide where the next dollar goes.
With manual campaigns, advertisers tend to say:
"I think these people are my customers."
Advantage+ effectively says:
"Give me the conversion signal and enough room to explore; I'll find the people."
Meta says Advantage+ Audience can expand beyond your selected interests, demographics, lookalikes, etc., while still allowing hard constraints such as location, minimum age and certain exclusions.
And Advantage+ Sales combines that with automated budget allocation, placements and creative optimization. Meta specifically recommends importing eligible existing ads so its system can test creative combinations and deliver the strongest ones.
That matters because audience targeting is becoming less of the advertiser's competitive advantage. Creative, offer, product-market fit, first-party data and conversion quality increasingly matter more.
When I'd still go manual
1. You need a hard audience boundary
This is probably the biggest reason.
If you absolutely need:
- only existing customers
- only non-customers
- a specific age bracket
- a tightly defined geographic area
- a specific prospecting audience
- exclusion of a particular segment
…manual gives you more predictable control.
That said, don't automatically assume you need manual just because you have an audience hypothesis. Advantage+ can use audiences as suggestions rather than hard targeting, which often gives Meta more room to find cheaper conversions.
2. Your business has radically different customer economics
Suppose you sell:
- $40 product → 25% margin
- $400 product → 70% margin
- $4,000 product → huge LTV
You may not actually want Meta optimizing purely for purchase volume.
A manual/hybrid structure can make sense when you need to control spend around different economics or optimize toward a carefully defined business outcome.
Even better, if you have reliable first-party data, feed Meta the actual downstream value rather than forcing the platform to optimize on a simplistic purchase signal.
3. You don't have enough conversion data
This is where I wouldn't blindly copy the "Advantage+ always wins" advice.
If an account is brand new and generating almost no purchases, Meta doesn't have much signal to work with.
You can still use Advantage+, but I would be cautious about expecting sophisticated AI optimization immediately.
In that situation, a relatively simple structure with strong creative and clean conversion tracking can be more useful than building a complicated manual targeting architecture.
4. You have a genuine testing hypothesis
This is an important distinction:
Manual targeting is useful for experimentation even when Advantage+ is better for scaling.
For example:
"Do people interested in marathon running convert 30% better than broad?"
If you put both audiences into an Advantage+ campaign, Meta may simply find the cheapest conversions and blur the distinction you're trying to measure.
A controlled manual test can answer the question.
Just don't confuse "manual is useful for testing" with "manual should be my permanent scaling structure."
5. Your offer is extremely niche
If you're selling something that only makes sense to a tiny population—say, specialized industrial equipment or a highly specific professional service—giving Meta an enormous broad audience may not be advantageous.
This is one of the situations where human knowledge about the market can still materially outperform pure algorithmic exploration.
The setup I'd use for most ecommerce brands
I wouldn't choose between "all Advantage+" and "all manual."
I'd use a hybrid architecture:
1. Main acquisition engine → Advantage+ Sales
Put the majority of prospecting budget here.
Give it:
- strong creative variety
- clean purchase events
- good product/catalog data
- sufficient budget
- broad audience
- good first-party data
- enough time to learn
2. Manual campaigns → strategic exceptions
Use them for things like:
- controlled audience tests
- geographic experiments
- specific customer segments
- highly constrained prospecting
- unusual products/offers
- incrementality experiments
3. Retargeting → don't automatically overbuild it
This is another area where old Meta playbooks can be misleading.
Don't assume you need five manually segmented retargeting campaigns simply because that's how accounts were structured a few years ago.
Let the economics tell you whether separating those users actually creates incremental sales.
The biggest mistake I'd avoid
Don't compare campaigns by ROAS alone.
An Advantage+ campaign can look spectacular because it's very good at harvesting existing demand.
For example:
- Manual prospecting: 2.5x ROAS
- Advantage+: 5.0x ROAS
That doesn't necessarily mean Advantage+ is generating twice as much incremental revenue.
The Advantage+ campaign may be getting a lot of people who were already going to buy.
So when deciding between the two, I'd look at:
Incremental revenue → contribution margin → new-customer acquisition → CAC → LTV → total blended business performance
rather than simply:
Ads Manager ROAS.
My rule of thumb in 2026
If you're an established ecommerce brand and you're asking "Which should be my default?" → Advantage+ Sales.
If you're asking "Should I ever use manual?" → Absolutely. But use manual because you have a specific reason to constrain or test the algorithm—not because manual targeting feels more sophisticated.
And there's an important Meta-specific nuance now: the platform increasingly lets you combine automation with manual inputs rather than forcing a binary choice. Meta explicitly says advertisers can still use manual controls where needed, while its automated systems handle the parts where machine learning has an advantage.
If you tell me your monthly Meta spend, monthly purchase volume, AOV, gross margin, and whether you're primarily trying to acquire new customers or maximize total revenue, I can give you the campaign structure I'd use for that exact account.
