Everyone wants to talk about targeting. Audience segments, lookalike audiences, detailed demographics, interest-based matching. The conversation around DTC advertising almost always circles back to the same question: How do we reach the right person?

But here's the thing. That's not really how modern advertising platforms work anymore. Not really.

Meta doesn't "target" in the way most people think. TikTok doesn't either. Neither does Amazon. These platforms operate as prediction engines. For every single impression, the system is running a calculation: What's the likelihood this person takes action? What's the expected value of that action? How much competition exists for this impression? Is this a quality ad experience?

The platform isn't reading your audience parameters and saying, "Okay, we'll show this ad to women aged 25-34 who like yoga and sustainability." That's the old mental model.

What's actually happening is the platform is algorithmically distributing impressions to whoever is statistically most likely to engage with your specific creative, given unlimited auction pressure. The creative itself becomes the targeting mechanism.

This fundamentally changes how you should think about your advertising strategy. And honestly, it informs everything about how we work with clients at Y'all.

Your ads don't reach the right person because of your targeting settings. Your ads reach the right person because your creative message is compelling enough that the algorithm learns who responds to it, and keeps showing it to more people like them. You don't choose who sees your ad. You choose which messages deserve amplification.

I'll walk you through why this matters, how it informs our approach to creative for DTC performance marketing, and what it means for your brand's growth.

The Algorithm Learns from Creative, Not Targeting Parameters

Here's what I've learned running performance creative for DTC brands. When we put a broad targeting parameter in front of Meta with generic creative, the system takes a long time to learn who actually cares. The learning phase is slow. The costs are high. We plateau.

When we put the exact same broad targeting parameter in front of Meta with highly specific, differentiated creative, the system learns fast. It finds efficiency pockets we didn't expect. It figures out which audience behaviors actually drive results, and it optimizes distribution accordingly.

The targeting parameter is almost incidental. The creative is the real signal.

Think about it from the platform's perspective. The system has no inherent loyalty to your demographic targeting. It's purely optimizing for predicted action. If your ad creative works better for a 40-year-old dad than for a 28-year-old woman, the algorithm will learn that. It'll start showing your ad more to that 40-year-old dad, regardless of whether he fits your "target audience."

This is why broad targeting with diverse creative consistently outperforms narrow targeting with limited ads. You're not forcing the system into a box. You're letting it explore, learn, and find the audiences that actually engage with your message.

Targeting still matters, but differently than most people think. Your targeting parameters define the sandbox in which the algorithm learns. They act as initial boundaries and guardrails, especially important for cold start accounts, limited-data situations, and regulated categories where you need structural constraints. But within those boundaries, creative quality is what determines whether the algorithm finds the right people or wastes budget.

Meta's latest push around Andromeda optimization really drives this home. Andromeda is Meta's newer ranking system that places heavier weight on creative performance signals over rigid audience definitions. The algorithm needs creative diversity to operate efficiently. It needs multiple representations of your core message to understand which variations resonate with which behaviors.

Low creative variety equals slow learning. The system sees one message, learns from it conservatively, and when you try to scale, it immediately hits diminishing returns. You run out of people who look like the early responders.

High creative variety means different messages resonate with different behaviors. You get unexpected efficiency pockets. You can scale to multiple audience segments simultaneously because each segment finds a message that speaks to them. The cost per acquisition doesn't necessarily rise when you increase spend.

This is a fundamental shift. Most brand leaders and marketers still approach creative as decoration on top of targeting. We define a concept as a core creative focal point with a specific point of view. From that, we develop multiple creative representations. That's the real work.

Creative Diversity Isn't Optional

At Y'all, we're testing around 5 to 20 concepts a month with our clients, depending on their spend and objective. Per concept, we're usually doing 2 to 4 Andromeda optimized variants, sometimes more depending on scope. This isn't busywork. It's core infrastructure for modern DTC performance.

And we're not really doing hook testing anymore in the traditional sense. Everyone got obsessed with hook testing for a minute. "What if we swap the first three seconds? What if we try this different opening line?" But here's what we actually found: you can keep the same content but think about it structurally from a storytelling standpoint. That's how you get the aforementioned Andromeda optimized variants.

The system cares about story structure, emotional resonance, and authenticity more than it cares about whether you swapped the thumbnail. We're thinking about the narrative arc. How does the message unfold? Does it feel organic or forced? What perspective is being presented?

One client we worked with spent years protecting their brand aesthetic. Very polished, very consistent, very pretty. But in a crowded DTC space, especially for an unknown brand, generic brand ads don't work from a performance perspective. The algorithm doesn't reward consistency unless it contributes to conversion. It cares about whether the creative converts.

This is where using user-generated content to improve ad performance becomes so valuable. UGC creators bring authenticity that polished brand content can't touch. When we brief out to creators, we give them parameters, but we don't over-engineer the output. The results are really cool because they play as organic and actually different from typical DTC ads.

How Persuasion Actually Works Now

This is the distinction that separates top creative agencies for DTC advertising from everyone else.

Most agencies still think their job is to persuade people. Create compelling ads. Win hearts and minds. Make people want to buy. That's still valuable, persuasion hasn't gone away. It has however, changed shape.

What's happening now is system-level persuasion mediated by the algorithm. It's no longer linear or manual. You can't just write one killer ad and expect the platform to run it to everyone (if only life were that easy). The algorithm needs multiple signals, multiple narrative angles, multiple ways of framing the same core message. Your job as an advertiser is to give the platform enough creative data to learn the most effective persuasion angle for each audience segment.

Think about it this way. If you're running 30 different ad variations, you're giving the platform 30 different persuasion signals to learn from. Maybe 15 of them will underperform. Maybe 8 will perform okay. And 7 will unlock something. The platform will identify those 7 and keep optimizing them. But you have to supply the volume and the quality.

High-quality persuasion at scale is what moves the needle. Bad creative at volume doesn't help. You need creative that actually resonates emotionally, tells a coherent story, and speaks to real customer motivations, paired with enough volume to let the algorithm find which segments respond to which approaches.

We're trying to give the platform enough data so the system can optimize. We don't want to over-structure things because we want to give it boundaries without being overly restrictive. You're building conditions for the algorithm to learn what works.

Why Creative Systems Beat Single Campaigns

Here's what this looks like in practice. We'll come into a client account and see a situation where they've been running the same creative for three months. Costs have climbed. Client wants to cut budget or blame the platform.

What's actually happened is the algorithm has tapped out that creative. It's shown the ad to everyone who would respond to it. To keep growing, you need new creative to reset the learning phase.

So we build out a proper creative roadmap. We define what makes this brand unique. We develop multiple narrative angles. We create variations that appeal to different motivations, different use cases, different pain points. Some of them hit immediately. Some of them flop. Some of them perform decently but sit in the middle.

The ones that hit, we double down. We explore variations of the winning angle. We let the algorithm learn from them. Then, three weeks in, we're seeing new cost efficiency because the platform is discovering audience segments that respond specifically to that message.

This is why what makes a creative agency successful for DTC companies is whether they understand that creative is a system, not a project. Most agencies treat creative like a discrete deliverable. You brief it. They make it. You run it for a campaign. Done.

Real performance creative is iterative. It's informed by continuous learning. It evolves as the platform gives you signals about what works. It compounds over time.

Meta's Andromeda Update and What It Means

Meta's recent shift around Andromeda optimization made this even more obvious. They're basically saying: stop micro-targeting. Stop trying to be clever with narrow audiences and highly specific demographics. Give us broad audiences and diverse creative. The system will figure out the targeting.

In practice means leaning heavy into creative diversity and testing personas and messaging angles rather than hard-coded audience variations. The conversation has shifted from what forced audiences perform best to what creative messages unlock new audience segments.

This requires a different internal capability. You need people who understand narrative structure, not just visual design. You need people who can craft authentic messaging, not just on-brand messaging. You need systems for rapid testing and iteration, not annual creative planning.

If you're an agency that specializes in performance creative for DTC brands, this is becoming the core differentiator. Can you build creative that the algorithm can actually learn from? Can you maintain creative supply? Can you pivot when something doesn't work, without losing confidence in the overall approach?

Building a Creative System That Scales

So how do you actually do this at scale? How do you maintain creative diversity without it becoming a chaotic mess?

First, you define your creative strategy. This isn't vague. It's specific. What are the actual customer motivations we're trying to speak to? What outcomes are they trying to achieve? What objections are they wrestling with? What emotional states drive them to action?

From that, you develop multiple narrative angles. Not surface-level variations. Structural differences. Different ways of framing the problem. Different solutions. Different emotional appeals.

Then you create representations of each narrative. One might be beautiful UGC. One might be a demo video. One might be customer testimonial. One might be educational content. Same underlying message, different packaging.

You launch them in parallel. You give the platform volume to learn from. You watch the data. You kill the ones that don't work. You iterate on the ones that do. You keep the supply flowing.

The temptation is to over-optimize this process. To become so data-driven that you're just tweaking infinitesimally. Resist that. The system needs structural variety, not incremental optimization. That's where most teams go wrong.

Frequently Asked Questions

If creative is the real targeting mechanism, why do targeting parameters matter at all?

They do matter, but differently than most people think. Your targeting parameters define the sandbox in which the algorithm learns. They act as initial boundaries and guardrails, especially important for cold start accounts, limited data situations, and regulated categories. But within those boundaries, the algorithm is optimizing for predicted action based on your creative. Broad targeting gives the algorithm more room to explore and learn. Overly narrow targeting limits its ability to discover high-value audience segments that don't fit your demographics but do respond to your message. Generally, we lean toward wider audiences and let the creative do the differentiation work.

How many ad variations is actually useful? At what point does testing become wasteful?

It depends on your spend level. Under $50k per month, you're probably looking at 5-10 core concepts per month with 2-3 andromeda optimized variants each. In the $50-150k range, step that up to 10-20 concepts. If you're spending $250k or more monthly, you might need 20+ concepts to keep the algorithm learning. These numbers aren't laws. They vary depending on how well you understand your audience and how predictable your winners are. The key is that this isn't a set-it-and-forget-it number. Early on, you might test more. As the account matures and you get better at predicting what works, you might test less.

What's the difference between UGC and in-house creative when it comes to algorithm performance?

Both can work. The real differentiator is authenticity and narrative structure. UGC creators often bring a level of authenticity that's hard to manufacture in-house. But if you're outsourcing UGC without maintaining direct control over messaging and creative direction, you might end up with content that looks cool but doesn't drive conversions. We prefer to maintain direct creative control regardless of whether the content is UGC or in-house production. The sourcing method matters less than the ability to iterate and optimize the underlying message.

How do you know when creative exhaustion has set in versus normal platform fluctuation?

Look at your cost trends over time. If you're running the same creative and costs are climbing steadily week over week, that's typically exhaustion. If you're launching new creative and costs dip back down, that's confirmation. Real platform fluctuation tends to be shorter-term and less directional. Exhaustion is usually a slow climb. As soon as you see that pattern, it's time to refresh creative supply.

Should we throw out our brand guidelines in favor of performance?

Not exactly. Your brand identity matters. But there's usually a way to express it through a performance-first lens. The problem is when brand guidelines become so restrictive that they eliminate authentic messaging or narrative variety. A good creative system respects brand identity while prioritizing performance signal. Sometimes that means pushing back on internal stakeholders about what's actually working. That's uncomfortable, but it's necessary.

What's the one thing we should focus on if we're just starting with this approach?

Creative supply. Just commit to consistent creative output. You don't need perfect systems or sophisticated testing frameworks at first. You need volume and diversity. Define a few narrative angles that make sense for your business. Produce variations. Launch them. Pay attention to what the platform rewards. Iterate. Most brands underestimate how much creative they need to produce and overestimate how much a single piece of creative can accomplish.

Does this approach work the same on TikTok as it does on Meta?

Similar principles, different execution. Both platforms are prediction engines optimizing for engagement and conversion. Both reward authentic, differentiated creative. But TikTok has a stronger emphasis on organic feel and native content format. Your creative system might look the same, but the actual content might skew more toward genuine UGC or native TikTok formatting on platform, while Meta might accommodate more polished variations. The underlying philosophy is the same.

What This Means

Most growth stalls in DTC advertising happen because creative runs dry, not because audiences are gone. If you're running performance marketing and feeling like your current approach isn't sustaining, the issue is probably creative supply, not targeting strategy.

Building a sustainable creative system requires discipline and resources. Most people skip it because the work is unglamorous. But creative supply is the actual lever that moves platform performance.

Brands that have nailed how to scale in 2026 treat creative like infrastructure. They fund it like they fund media spend. They measure it like they measure channel performance. They evolve it based on what the algorithm rewards. That's the difference between brands that hit a ceiling and brands that keep scaling.

If you want to talk through what a creative system looks like for your brand, or if you're thinking about how to build this internally or with an agency partner, reach out. We work with DTC brands on exactly this stuff.

Fill out an inquiry form to chat with us or check out our case studies to see how this plays out with our clients.