Past Sessions

• Event: MAU Vegas 25
• Date: Thursday, May 22, 2025
• Speakers: Hannah Parvaz, Founder, Aperture
• Estimated read time: 6-7 minutes

Quick Read Summary

If your paid social strategy feels like it lost its steering wheel, you are not imagining it. Targeting has weakened, costs have risen, and the temptation is to respond with more campaigns, more audience slices, and more “clever” setup. Hannah Parvaz argues that this is exactly how teams burn budget without learning.

The practical shift is to stop treating ads as a distribution lever first, and start treating them as a learning system. In her framing, ad creative is not decoration around targeting, it is the mechanism that helps platforms find the right people, while revealing which problems, motivations, and language actually move buyers.

The teams that regain momentum do two things consistently. They build creative diversity so they are not pinned to one message, and they run structured creative testing that isolates variables, anchors results to a clear control, and compounds learnings over time.

The outcome is not just lower cost per acquisition, it is a clearer route to product market fit, because every creative is a small experiment that tells you what the market will pay attention to, and what it will pay for.

The real problem is not targeting, it is relevance

For years, performance marketing rewarded the teams that could build neat targeting buckets, lookalikes, interest stacks, and intricate campaign structures. Parvaz’s contention is that this era is fading, and the operational symptoms are familiar, higher CPMs, lower returns, less control, more guessing, and audience tactics that decay faster than teams can replace them.

When targeting becomes less deterministic, many teams respond by multiplying complexity. Parvaz shared an account she took over that was running 142 live campaigns and had never been profitable in five years. The activity looked like progress, but it produced “zero learnings” because it recycled the same creative with different targeting, and it had not launched a new creative in six months.

Her point is blunt, more campaigns does not mean better performance. It usually means more noise, more reporting, more time spent convincing yourself something is working.

“You can’t outsmart bad creative with campaign structure and you can’t optimize your way out of irrelevance.” Hannah Parvaz, Founder, Aperture

That idea reframes the job. If the market is saturated and attention is scarce, the first-order question is not “which audience do we pick,” it is “why would anyone care.”

Product market fit is not a vibe, it is purchase intent

Before ads can become a tool for growth, Parvaz argues teams have to be precise about product market fit. She referenced Mark Andreessen’s definition, being in a good market with a product that can satisfy that market, then sharpened it with a more practical filter, you have found a need and built a solution customers actively want to buy.

The word “actively” matters. Clicking is not the same as committing. In her view, product market fit shows up when customers reach for their wallet, Apple Pay, or card and choose you with intent, not curiosity.

This is where advertising can either accelerate learning or distort it. If you measure the wrong thing, or allow the platform to allocate spend unevenly across early creatives, you can end up scaling what is merely “getting delivered” rather than what is truly resonating.

Parvaz’s alternative is to treat creative testing as a disciplined path toward product market fit, because it forces you to articulate hypotheses about what people want, then validate those hypotheses against market behaviour at speed.

Ads are experiments, creative is the interface to the algorithm

Parvaz’s central thesis is that the algorithm is not your enemy, it is the most powerful research engine you can rent, if you give it clean inputs.

“Your creatives are not just ads, they are experiments.” Hannah Parvaz, Founder, Aperture

In her framing, each creative is “a tiny customer interview disguised as a video or headline.” It tests whether a pain point lands, whether a motivation moves people, and what message triggers action across the funnel. This is where creative testing becomes more than an optimisation tactic, it becomes product insight.

She also argued that platforms “know your audience” at a depth that makes traditional targeting less valuable over time. She cited a 2014 Stanford and Cambridge study of 86,000 people, saying that at 300 or more interactions, Facebook could predict a person better than their partner, then challenged the room to consider how much stronger the inference engines are now.

You do not have to accept every implication of that claim to use the practical lesson. The platform’s job is pattern recognition. Your job is to feed it patterns worth recognising, which means creative that is specific, varied, and grounded in real customer language.

Creative diversity beats message monoculture

One of the most actionable parts of Parvaz’s talk is the diagnosis that many products are not being rejected, they are being misunderstood.

She illustrated this with a supplement brand that led with one fixed message, essentially “we have vitamins,” repeated in different colours and fonts. The issue was not that nobody wants vitamins, it was that different buyers care about different outcomes, skin, hair, vegan credentials, sleep, feeling like themselves again. A single message only reaches the people who already prioritise that one angle.

Her conclusion is a strategic one for paid social strategy, if your messaging has low diversity, your reach is artificially constrained, even if your audience targeting is broad. The creative itself needs to do the segmentation, because the algorithm can infer who to show an ad to based on what the ad is saying.

This is also where “creative velocity” becomes non negotiable. Parvaz described a team with around 20 million users that launched one new creative per month. Her assessment was that it is “just not enough” to keep pace with rising frequency and shrinking effective reach.

A simple creative testing loop that compounds learning

Parvaz laid out a three step loop that turns creative into a system rather than a one off production task.

Her first insistence is that creative cannot be generated from abstractions. She said 75 percent of teams she speaks to have not spoken to a customer in the last year, and called this the core failure mode.

She recommended five to 10 deep conversations, not hundreds, and listening for simple but high signal inputs, what they want, what is holding them back, what they fear, what made them act. Features are rarely the story, benefits and lived experience are.

A helpful method she gave is mapping what you hear into two columns, goals and struggles, then selecting the themes that are distinct and actionable. She suggested you often end up with eight to 10 solid directions, each treated like a creative bet.

She described a practical campaign structure designed to isolate messaging variables, one campaign, one ad set per creative theme, and one or more ads per ad set. The point is to give every message an equal chance and make performance comparisons clear.

The detail that matters here is spend discipline. If the platform pours budget into one idea early, you are not running a fair test. She recommended allocating the same budget per theme at this stage to remove spend as a variable, and to move toward statistical significance.

She shared an example from a FinTech app called Juno, where the team started with simple statics and small changes, even changing one word to test different angles such as salary negotiation, investing, and budgeting. In that test, the budgeting angle came out on top and reduced CPA by 27 percent, then became the control that informed what they scaled.

Once a theme wins, Parvaz’s approach is not to declare victory, it is to deepen the learning by iterating formats and funnel steps while keeping a benchmark.

She described testing different designs against the control, again with equal budget allocation. In one round, a design reduced CPA by 23 percent, while another variant increased CPA by 74 percent. The result is not just a performance delta, it is a signal about what the audience finds credible, appealing, or off putting.

She then described moving from design into additional copy and emotional angles, such as empowerment, humour, and rebellion, with the warning that these are not “just headlines,” they are strategic bets grounded in customer insight.

When the team brought in a quantified benefit, she described using a stat of 4,400 more per year, which led to a 38 percent drop in CPA, while a savings based variant underperformed. The lesson is that even “obvious” value props need proof in the market before they earn a place in your evergreen system.

Finally, she described moving into UGC video once statics validated the message. She noted that more than 90 percent of viewers do not get past the first three seconds, which is why she recommends hook testing with many options, then holding the body and call to action constant. She shared a small but telling example where adding a sigh at the start of a video doubled costs versus a version that cut straight to the line.

Put together, the loop moves from messaging, to design, to creators, to hooks, with every round informing the next. Parvaz reported that this structured momentum led to a 93 percent reduction in CPA in the financial education case study she described.

The profitability unlock is learning discipline

The headline promise of “stop burning budget” is ultimately about managerial discipline, not secret tactics. In Parvaz’s view, teams lose money when they spend without learning, and they regain profitability when they consolidate, increase creative velocity, and treat creative as an insight engine.

Her story about cutting 142 live campaigns down to four, including one awareness campaign, then reaching profitability within three months for the first time, is a reminder that simplification can be a performance strategy when it restores signal, focus, and iteration speed.

If you want a practical way to stress test your own system, ask one question at the end of every week, what did we learn that changes what we do next week. If the answer is unclear, you are probably busy, but not progressing toward product market fit.

Conclusion

Advertising is no longer a reliable shortcut to growth, unless you use it as a learning machine first. Parvaz’s argument is that creative testing, anchored in real customer language and run with disciplined structure, is how teams find product market fit faster and scale profitably.

The stakes are straightforward, if you are spending without learning, you are not growing, you are just burning.

Speaker

Hannah Parvaz, Founder, Aperture.  Hannah is an award winning growth expert, international keynote speaker, and founder of Aperture, named Most Innovative Growth Agency of the Year. She has worked with hundreds of mobile apps, from early stage startups to household names, leading growth strategies that have taken products from zero to millions of users. Her work spans performance marketing, creative strategy, mobile attribution, and product analytics, rooted in customer psychology and data. Through Aperture, she partners with teams to unlock growth that scales across channels, product, and brand.