Meta Ads Strategy
Contents
Something fundamental has happened to Meta Ads over the past 18 months. Not a minor adjustment. Not a new button in Ads Manager. A fundamental change in how the platform distributes ads and who it shows them to.
And yet most e-commerce brands are still running the same strategy they used in 2022. Tight interest-based audiences. Three to five creatives that have been running for months. A single hero video they spent $4,000 producing. And then they wonder why ROAS is falling, CPM is rising and the account feels stuck.
This article is about what has actually changed, why your content strategy is now the most important variable in your entire Meta Ads account, and what you should concretely do differently.
In late 2023, Meta rolled out their Andromeda update. It's the biggest change to the ad distribution system since they moved away from the original auction model. In short: Andromeda enables Meta to evaluate up to 10,000 times more ads per impression than before.
What does that mean in practice? That broad targeting doesn't just work. It is now better than precise targeting in almost every case.
Before Andromeda, the logic was: you gave the algorithm a hint by selecting interests and lookalikes, and it found the best buyers within that segment. That was a sensible approach, because the algorithm's capacity to scan broadly was limited.
Now the algorithm scans Meta's entire user base in real time. It can evaluate billions of signals per user, and it genuinely doesn't want your interest stacks. You can still set them up, but in practice Advantage+ ignores them and goes broad anyway. Meta has confirmed this repeatedly in their performance documentation.
This isn't a theory. It's something we see account after account: broad Advantage+ Shopping campaigns outperforming tight audience setups with 30-60% lower CPA. Not because the ad is better. But because the algorithm has been allowed to do its job.
The consequence: Targeting is no longer your competitive advantage. Everyone is going broad. Everyone is using Advantage+. That means the battle has moved somewhere else entirely. It's now purely about what you show people.
Want to go deeper on the Andromeda update itself? We cover it in depth in this article.
Here's the central argument of this article, and it's worth stating directly: your creative strategy is now the only meaningful variable you can control in Meta Ads.
Think about it. Budget anyone can scale. Advantage+ campaign structure is the same for everyone. Pixel tracking and CAPI are table stakes, not an advantage. Audience targeting is controlled by the algorithm. Placements? Automatic placements are the standard.
What's left? Your creatives. Your hooks. Your messaging. Your ability to reach different people with different angles on your product.
This is a paradigm shift for performance marketing. For five years, Meta Ads was primarily about technical media buying: the right campaign structure, the right audiences, the right budget allocation. It was a spreadsheet job. Now it's a creative job.
That doesn't mean campaign structure is irrelevant. It means it's become table stakes. You don't lose because your structure is wrong. You lose because your creative pipeline is too thin.
When all advertisers bid on the same broad audience, competition per impression increases. Meta has to choose between thousands of ads for each individual user. And it picks the ads with the highest probability of engagement and conversion.
The algorithm rewards relevance. And relevance comes from creatives that match the user's context, needs and stage in the buying journey. A generic product ad with a 20% discount competes against ads that speak directly to the user's specific problem. And the generic ad loses. Every time.
That's why the answer isn't "better" creatives in the traditional sense. It's more and more varied creatives. Because you don't know which message will land with which part of your audience. The algorithm knows. But it needs variations to work with.
Let's make this concrete with some numbers.
Imagine two brands in the same category with the same budget. Brand A tests 5 new creatives per month. Brand B tests 40.
Generally, 10-20% of tested creatives become winners, meaning ads that perform above break-even and can be scaled. It varies, but that's a realistic average for e-commerce on Meta.
Brand A (5 creatives/month): finds 0-1 winners per month. Maybe a good month with 1 winner. Maybe a bad month with zero. They depend on luck and have no pipeline.
Brand B (40 creatives/month): finds 4-8 winners per month. They have a constant stream of fresh winners to scale. When a creative fatigues after 2-3 weeks, there's already a new one ready.
Over a quarter, Brand A has found 0-3 scalable creatives. Brand B has found 12-24. Who do you think scales faster? Who do you think has lower CPA and higher POAS?
A winning creative on Meta typically has a lifespan of 2-4 weeks before performance starts to drop. Some last longer, but that's the exception. Andromeda has paradoxically amplified this, because the algorithm is so efficient at finding the right users that it quickly exhausts the best segments for a given creative.
If you only have 2-3 active winners and one of them fatigues, you suddenly lose 30-50% of your performance. That's the scenario we see again and again with brands that don't have a creative pipeline. They have good periods when they hit a winner, and then flat periods when it dies.
The solution isn't to make "evergreen" creatives that never fatigue. Those don't exist in an Advantage+ world. The solution is to have a pipeline that constantly produces new variations, so you always have fresh ads ready.
Need a steady pipeline of test-ready content? We deliver 10-50 static ads per month at a fixed price.
See pricingIt's not "make a beautiful brand video and run it for 3 months." It's a system. A machine that produces, tests, iterates and scales creative content based on data.
Here are the five pillars we build content strategies on for our clients.
Your product doesn't solve one problem for one type of customer. It solves multiple problems for multiple types of customers. Your content strategy needs to reflect that.
Take an example: a brand selling an ergonomic office chair. Here are five fundamentally different angles:
Each angle reaches a different part of the audience. And you don't know in advance which one will perform best. So you test them all. The algorithm figures out who responds to what.
Different formats perform differently across different placements. A 9:16 Reel works differently from a 1:1 feed image, and both perform differently from a carousel with five slides.
A solid content strategy includes at minimum:
The point isn't to produce everything. The point is to have variation so the algorithm can match the right format with the right user on the right placement.
On Meta you're not competing against other advertisers. You're competing against friends' holiday photos, memes, news articles and dog videos. Your first one and a half seconds (for video) or your top third of the image (for statics) determines whether the user stops or keeps scrolling.
We see it in the data constantly: two identical ads with a different hook can have a 300-400% difference in hook rate. Same product, same offer, same copy. Just a different first line or a different visual element at the top.
That's why you should always test hooks as an independent variable. Take your best-performing ad and make five versions with five different hooks. That's the fastest path to better performance without having to reinvent the entire ad.
The best ad copy doesn't come from brainstorms in a meeting room. It comes from your customers' own words.
Go to Trustpilot and read your reviews. Not to find stars, but to find language. What words do people use when they describe why they bought? What problem did you solve? What surprised them positively?
Go to Reddit, Facebook groups and forums where your audience hangs out. What do they complain about? What solutions are they looking for? What language do they use?
The best copy feels like someone has read your mind. And the reason it feels that way is that it's actually written using the customer's own phrasing. That beats polished marketing copy every time in a performance context.
A content strategy without a feedback loop is just content production. The difference is data.
When you've tested 20 creatives and found 3 winners, you don't stop. You analyze what the three winners have in common. Is it the angle? The hook? The format? The color palette? The copy length?
Then you produce 10 new variations based on the patterns you've found. Maybe your pain angle with short, direct copy consistently outperforms your aspiration angle. So you make five new pain angles with different hooks and visual styles.
That's the iterative cycle that separates brands that scale from brands that throw money at random creatives and hope for the best.
Okay, we're biased. We're called JustStatics. But the data speaks for itself, and it tells a story that surprises most performance marketers.
Static ads are the most underrated format on Meta. Most brands spend 80% of their creative budget on video, because video "feels" more premium and professional. But when you look at cost per acquisition across formats, statics are often on par with or better than video. Especially in e-commerce, where purchase decisions are fast and impulse-driven.
We've written a full article on static ads vs. video ads, so we won't repeat all the arguments here. But the main reasons statics are the backbone of a high-volume content strategy:
The optimal strategy isn't either/or. It's statics as the testing engine combined with video for scaling confirmed angles. You find your winning angle with statics (fast and cheap), then invest in video production of the angle you already know works.
Here's how we do it: Our clients use JustStatics to produce 10-50 static ads per month. They use them as test material in dedicated test campaigns, find winners, and scale from there. Some make video versions of the best statics. Others scale statics directly. Both approaches work.
We talk to e-commerce brands every week. The same patterns come up, and they all cost money.
It's the most common mistake, and an understandable one. It feels safe to hold onto something that "works." But on Meta in 2025, even good creatives die. Creative fatigue isn't a question of "if", it's a question of "when." And the answer is typically 2-4 weeks.
When you run the same ads for three months, you pay more and more for worse and worse results. CPM rises, CTR falls, and ROAS crumbles. You don't notice it day by day, but over a month the drop is significant.
We see it all the time. A brand hires an agency to make a beautiful brand video for $3,000-5,000. The video is gorgeous, well-produced and tells the brand's story. And then it runs as the only creative for 2-3 months.
The problem? First, you don't know whether the video performs at all until you've spent money testing it. Second, you've put your entire creative strategy on one card. If the video doesn't work, you have nothing to fall back on.
Those same dollars could have bought 80 static ads. That's 80 chances to find winners versus 1 chance. The math is clear.
Many brands test "creatives" as whole units. They make three completely different ads and test them against each other. That gives data, but not insight. Because you don't know whether it was the hook, the copy, the image or the format that decided the result.
Systematic testing isolates variables. Take the same image with five different hooks. Or the same hook with five different images. Then you know exactly what's driving performance.
There's a widespread misconception that video always beats images on Meta. That's simply not true in e-commerce. Statics perform consistently well, especially in feed placements, and they're unmatched for quickly testing angles and messages.
Brands that only run video are missing an entire dimension of their creative strategy. They test fewer angles, iterate more slowly and pay more per test.
This is the most frustrating mistake, because it's based on an understanding of Meta that's no longer correct. We see brands spending hours fine-tuning audiences, testing lookalike percentages and stacking interests. All while their ads are the same ones they've been running for two months.
In a post-Andromeda world, targeting optimization is like polishing the wheel caps on a car with a flat engine. It looks fine, but it moves nothing. Shift your focus to creatives. That's where the return is.
Enough theory. Here's what you concretely do, starting this week.
Then repeat. Every month. This isn't a one-time task. It's a machine you build and maintain.
Rule of thumb: For every $1,000 in monthly ad spend you should test a minimum of 5-10 new creatives per month. Running $10,000/month? Then you should be producing 50-100 new creatives. That sounds like a lot. That's why you need an efficient production partner.
Meta Ads in 2025 is a fundamentally different game than just two years ago. Andromeda has removed targeting as a competitive advantage. Advantage+ has standardized campaign structure. The only thing separating winners from losers is the ability to produce, test and iterate on creative content at high speed.
That requires two things: a systematic approach to content production and a partner who can deliver volume without compromising quality.
The brands that understand this and act on it now will gain market share while their competitors are still fine-tuning audiences in Ads Manager.
The question is: which side do you want to be on?
We deliver 10-50 static ads per month, ready to go straight into your test structure. No lock-in. Fixed price. Delivered on a fixed date.
Book a free call