There’s a shift happening in social advertising that most ecommerce brands are still catching up to, and the brands that adapt fastest are going to have a significant performance edge.
The short version: the targeting work you’ve historically done inside platforms like Meta? The AI is doing it for you now. What it needs from you instead is better creative.
Here’s what that means in practice, and how to build an approach that actually takes advantage of it.
The AI Has Taken Over Targeting And It Doesn’t Want Your Help
Every major social ad platform – Meta, TikTok, Pinterest, X – is now leaning heavily on AI to determine who sees your ads. And that AI has gotten very good at learning on its own.
What’s driving that learning? Your creative.
The visuals, the text overlay, the people in your videos, the context of your content – all of it is being read by the algorithm to figure out who’s most likely to convert. You don’t need to tell Meta to find women ages 25–44 who are interested in skincare and yoga. Meta is going to infer that from the ad itself and go find those people.
In fact, restricting the AI with narrow targeting parameters tends to backfire. CPMs go up. Efficiency drops. The algorithm hates being boxed in.
The implication is significant: the time you used to spend building out interest stacks and audience segments inside the platform needs to move upstream into your creative strategy.
You Still Need to Know Your Audience. The Execution Just Changed.
This isn’t permission to skip audience research. You still need to know who you’re selling to: their demographics, their motivations, what they respond to, what drives their purchase decisions. That work is still essential.
What’s changed is how that knowledge gets applied.
Before, you’d build audience segments in Meta and feed them generic creative. Now, you leave targeting broad and let your creative do the targeting. If you know your primary buyer is a 35-year-old woman trying to solve a specific problem, you build creative that speaks directly to that reality, from the style and format to the pain point and emotional hook. The algorithm picks it up from there and finds the right people.
The targeting happens in the creative process now, not the campaign settings.
Creative Volume and Variety: What the New Model Actually Requires
If creative is doing the work that targeting used to do, the bar for creative output has gone up. Here’s what that looks like in practice:
Aim for 4–8 new ads per week, loaded in on a consistent cadence. The specific number scales with your budget. Higher spend demands more volume to keep frequency from becoming inefficient. But the cadence matters more than the count. A steady, predictable pipeline of fresh creative gives the algorithm more to work with and more opportunities to find winners.
Understand the difference between variation and iteration. These aren’t the same thing, and conflating them makes creative production harder than it needs to be.
- Variation is a new concept or angle, like a UGC testimonial, a founder video, a lifestyle shoot, or a product pain-point graphic. Different approaches to communicating your value.
- Iteration is taking one variation and creating multiple versions of it: different lengths, different text overlays, different background colors, captions on or off, graphics added or removed. Each counts as a distinct ad in the platform.
In practice, this means one concept can yield a full week’s worth of creative. A two-minute video becomes six ads when you cut it to 90 seconds, 60 seconds, 15 seconds, add captions to one, add bullet point overlays to another. That’s your week.
Small changes matter more than you think. Changing only the background color on a static ad, nothing else, can be the difference between a ROAS of 0.7 and 4.2. Don’t underestimate what minor iterations can reveal about what resonates.
The Framework for Knowing What’s Working
With volume and variety baked into your approach, the ad account becomes a testing engine. Here’s how to run it:
Let ads spend before you judge them. A useful rule of thumb: let an ad spend at least 10x your target CPA before making any call. If your target CPA is $25, let it spend $250. Then evaluate.
Kill low performers, scale winners, and make more of what works. This sounds simple, but it requires discipline, especially when you have a lot of creative running. Turn off what’s not performing after it’s had a fair shot. Move what’s working into a scaling campaign. Then produce new creative that mirrors the attributes of your winners.
Use frequency as a diagnostic tool. If a new customer is seeing your ad more than 4–5 times in a 30-day window, you probably don’t have enough variety in the account. That’s your signal to add more.
Don’t turn off what’s working just to make room, but keep the pipeline moving.
Where to Find Creative Inspiration (Without Starting from Scratch)
Running out of ideas is a real problem for lean teams. Two sources that consistently deliver:
The Meta Ad Library. It’s free, open to anyone, and shows you every ad your competitors are running: how many variations they’re testing, how long ads have been live, and how they’re iterating. It’s a direct window into what’s working in your category.
Your own product reviews. If a customer wrote a five-star review explaining exactly why they love your product, that’s ad copy. Build a simple template, pull five different reviews, and you have five ads ready to go. They perform well because they function as testimonials, and Meta knows how to serve testimonials to people who respond to social proof.
The Brand vs. Performance Tension (And Why You Should Let It Go)
One of the bigger mindset shifts this requires: loosening your grip on brand consistency in ads.
Marketers care deeply about brand guidelines: fonts, colors, tone, and visual standards. That makes sense across most channels. But in paid social, users aren’t evaluating whether your ad font matches your website. They’re deciding in half a second whether to keep scrolling.
An ad that stops the scroll and speaks directly to a real need is more valuable than an ad that’s perfectly on-brand but feels like a polished corporate ad. The most effective ads often look like content someone created themselves: authentic, relatable, specific.
This doesn’t mean abandon brand standards entirely. It means weighing resonance above visual perfection when the two are in conflict. In ads, being felt matters more than being recognized.
The Bottom Line
The shift to AI-driven targeting in social platforms isn’t coming. It’s already here. Brands that keep investing time in granular targeting parameters are working against the algorithm and leaving efficiency on the table.
The new model is simpler in some ways, harder in others. You need less time in campaign settings and more time thinking about creative strategy: who you’re talking to, what angle you’re taking, and how to generate consistent volume without burning out your team.
Test more. Iterate fast. Scale what works. And let the algorithm do what it’s actually very good at.
Watch the Full Conversation On Demand
If you want to go deeper on any of this, the full conversation with Jaryd is available to watch on demand. He covers the creative volume framework, how to structure your ad account for testing, and what separates the brands that are scaling right now from the ones that are stalling.
These insights come from a recent conversation with Jaryd Wilson, paid media strategist and longtime collaborator with LimeLight Marketing. Jaryd has spent over a decade in the social media space, working closely with ecommerce brands on paid and organic social strategy across Meta, TikTok, Pinterest, and beyond.
Want to pressure-test your current creative approach or figure out what it would take to build a sustainable creative system for your ad account?
