How to use customer personas in Meta ads now you can’t target them
Personas didn’t become irrelevant when interest targeting went away. They moved.
You used to build a persona, translate it into an audience segment, and let Meta find people who matched. That lever is disappearing. But the algorithm still needs signals to decide who sees what, and those signals now come from the creative itself. Show an ad featuring someone lacing up trainers before a morning gym session, and Meta learns which users respond to that world. Show a different ad with someone getting ready for a rooftop dinner, and it finds a different audience. Creative is the targeting now, and personas are the input layer that shapes it.
The problem most brands hit is the gap between the persona document and the actual creative. Clients say some version of the same thing: they know they need personas, they’ve probably built them, but when it comes to designing ads or writing the messaging, they have no idea how to structure the testing around those personas.
Persona-led vs product-led: the decision
Before building persona-driven creative, figure out whether a persona-led strategy is even the right call. The alternative is product-led, and the two suit different situations.
Persona-led builds a world the customer sees themselves in. The reaction you want is “that’s me.” It’s emotionally led. You’re reflecting lifestyle, identity, and self-image back at the viewer.
Product-led builds around the problem or function the product solves. The reaction is “that’s the product I need.” It’s functionally led. The creative centres on what the product does, not who uses it.
They’re not mutually exclusive. Most brands end up using both. But when you’re deciding which to lean into, five factors matter:
Category saturation. A saturated market where the brand already knows its customer and wants to build long-term brand equity favours persona-led. There’s nothing new to say about the product category itself, so the differentiator becomes the customer’s identity.
Customer-data maturity. A younger brand that hasn’t identified distinct customer segments can’t build personas from data it doesn’t have. Product-led comes first. You test product-market fit, collect purchase data, and build personas once the data supports it.
Category type. Fashion and lifestyle are emotionally led purchases. People buy because of how something makes them feel. That suits persona. Functional categories, whether that’s homeware, hardware, or performance equipment, tend to be more product-led. The purchase decision is driven by the need, not the identity.
Funnel focus. Testing whether your core offer resonates? Product-led. Working on retention, LTV, defending market share? Persona-led helps you speak to distinct segments with precision.
Budget. Persona-led demands more varied creative: different models, settings, and messaging tracks for each persona. Make sure the spend can support that volume.
Combining the two
Plenty of accounts run both strategies across the year. A brand with a long consideration period and strong seasonal peaks might use persona-led creative in the off-season to nurture intent over months, then switch to product-led conversion creative during peak trading when functional messaging drives the sale.
Even off-season conversion content can be framed around the customer’s world and specific use cases rather than spec sheets. The brand still sells outside of peak. It just sells differently: with conversion messaging wrapped in the persona’s context rather than stripped to pure product.
This hybrid works because the two strategies serve different commercial needs at different moments. The persona-led phase builds familiarity and intent. The product-led phase converts it.
What a mapped persona actually needs
A useful persona requires two layers of input: brand data and a creative workshop. Neither alone is enough.
What the brand supplies
The brand has data the agency cannot fabricate. A demographic character sketch: age, life stage, profession. Buying habits from their ecommerce platform. Post-purchase survey findings that reveal why people buy, not just what they buy. And the platforms, competitor brands, and lifestyle brands their customers use.
Brands that have run customer research arrive with most of this. Brands that haven’t need to start collecting. Post-purchase surveys are the simplest first step.
If a brand has no first-party data but wants to work with personas, market and industry research can approximate the segments. But data-led always beats assumed. Younger brands should collect first and build personas second.
What the creative workshop builds
With brand data as the base, the workshop builds the creative layer: lifestyle associations and use cases, which brand values each persona aligns with, and the motivations and barriers for each segment.
Motivations split into rational and emotional. The rational motivation might be “I need an outfit for a wedding.” The emotional motivation is “I want to feel attractive.” Barriers work the same way: there’s a practical one and an emotional one. That split matters because it gives you two distinct creative angles per persona, per need.
From persona document to creative brief
Most persona work stops at the document. A slide deck gets shared, everyone nods, and the creative team still has no idea what to actually make. The gap between knowing who your persona is and knowing what their ad looks like is where most brands stall.
Bridging that gap takes three layers per persona:
1. A creative lens
One sentence that captures what this persona needs to feel, what their key barrier is, and the single thing to get right. This keeps every brief aligned without turning into a manual nobody reads. Each persona’s creative lens should be meaningfully different. If they’re interchangeable, the personas aren’t distinct enough.
2. One core sentence per funnel stage
What does top-of-funnel creative need to communicate for this persona? What about consideration? What about conversion? Keep each to a single sentence. More than that and it becomes too complex to execute consistently.
3. Do’s and don’ts per creative factor
For each persona, specify what changes across these dimensions:
- Model casting. Who appears in the creative? Age, energy, setting.
- Backgrounds and settings. Where is the scene set? Different personas live in different worlds.
- Copy style. Aspirational vs practical, warm vs direct, playful vs authoritative.
- Formats. Which formats are likely to resonate? Carousel, video, static, UGC.
- UGC briefing. If you’re commissioning user-generated content, what should the brief specify for this persona?
- Repurposing existing assets. This is the budget answer. Rather than shooting new content for every persona, how can existing assets be re-edited, reframed, or recut to serve a different segment?
When two personas each have distinct direction across these factors, the creative team can produce genuinely differentiated output without guessing. When they don’t, every persona ends up looking the same with different copy overlaid.
The differentiation checklist
This is the test that separates real persona differentiation from surface-level messaging swaps.
Strip the messaging off two personas’ creatives. If the ads look the same, they aren’t differentiated.
With Andromeda expanding Meta’s capacity to match creative to individuals, persona differentiation cannot rely on copy alone. The algorithm reads visual and audio signals, not just text. Two ads with the same model, same setting, same format, and same visual style will read as functionally identical to the delivery system, regardless of what the headline says.
Eight factors to vary between personas. All eight ideally. As many as realistic:
- Hook. Different opening moment for each persona.
- Use case. What is this person doing when they use the product? One persona might be heading out for dinner. Another might be doing the weekly shop with kids in tow. Those are different worlds, and they should look different.
- Model. Different person in the creative, reflecting the persona.
- Background and setting. Different physical environment.
- Motivation. What drives this person to buy?
- Barrier. What might stop them?
- Emotional cues. Tone, colour, energy, music.
- Styling. For fashion, even how the same piece is styled can differ between personas. A younger customer might wear it casually. An older customer might dress it up.
The more of these you can make genuinely different, the stronger the differentiation. Two or three varied factors is a start. All eight is the goal.
Where to start when it feels overwhelming
If the checklist feels like a lot, sequence it:
Step 1: Differentiate the message. Make sure each persona’s motivations, barriers, lifestyle, and use cases are reflected in the copy and concept. This is the minimum baseline.
Step 2: Run the message across formats. Take that differentiated messaging and execute it as an image, a video, and a carousel. Same message, different executions. This multiplies your creative volume without requiring new concepts.
Step 3: Get granular with concept variables. Now start testing the visual and contextual elements. Would an outfit check work better than a get-ready-with-me UGC format for this persona? Indoor or outdoor? Model A or Model B?
That sequence matters. Messaging first, then format variety, then concept depth. Skip to concept testing before the messaging is differentiated and you’re testing variables on top of a weak foundation.
Personas are now infrastructure
Persona strategy used to live in a slide deck that sat alongside the media plan. The media buyer selected interests. The creative team made ads. The persona document was a reference, not an operational input.
That split doesn’t work when creative is the targeting mechanism. In a broad-targeting world, the creative tells the algorithm who to find. The persona document is the input layer that shapes what the algorithm receives. If your creative isn’t differentiated by persona, you’re running one ad strategy for your entire addressable market and leaving the algorithm with nothing to optimise against.
The brands getting this right treat persona work as ongoing infrastructure, not a one-off exercise. The persona informs the brief. The brief shapes the creative. The creative tells the algorithm who to find. And the performance data feeds back into the next round of persona refinement. That loop is where the compounding happens.
FAQs
Can you still do interest targeting on Meta?
Interest-based targeting options still exist, but Meta has been steadily reducing their scope and encouraging broad targeting. The direction of travel is clear: the algorithm handles audience selection, and the signals it uses come primarily from the creative. Building a strategy that depends on interest targeting means building on ground that’s shifting underneath you.
How many personas should you test?
There’s no fixed number. It depends on how distinct your customer segments genuinely are and whether you have the creative capacity and budget to serve each persona with differentiated content. Two or three well-differentiated personas will outperform six that all look the same in execution. Start with the segments your data supports and expand as your creative operation can handle it.
Ready to see what we can do for your brand?
Book a call with our team to get know more about it.
