The Challenge
Our client had a killer product (AI-generated videos), a working ads funnel, and aggressive targets: 6,000 new paying users in 90 days. Or roughly 7x their current customer base.
The problem? (You know, besides the massive goal and short deadline...)
They hit a plateau. They scaled quickly to $50k/mo. in spend. But all efforts to push beyond that point led to CAC ballooning on them.
Wouldn't it be nice if you could just keep turning up the budget dial and have the numbers hold? Trust us, they tried.
Here's the thing: ad platforms run on auction dynamics. When you push into broader audiences, you're competing with advertisers who have more relevance in those segments. If your offer is less relevant to that audience, you pay a premium to show up. The wider you go, the more that premium compounds.
Problems, meet solutions
1️⃣ THE TARGET ROAS TRAP
They were using Target ROAS, which is often a smart move, but it actually throttles spend when you're trying to reach new audiences. Target CPA gives the algorithm a clear acquisition cost and lets it optimize for volume instead of holding back.
2️⃣ BRAND VS NON-BRAND CONFUSION
They were lumping branded and non-branded traffic together. Certainly made their aggregate numbers look nice. But it was hiding the truth. We split them out. Uncomfortable in the short term because the "real" non-brand CPA was higher than they thought. But you can't fix what you can't see.
3️⃣ NO COMPETITOR CAMPAIGNS
They weren't bidding on competitors like Runway and Sora. People searching for your competitors are already in buying mode. CPCs are higher, but conversion rates more than justify it. This is almost always worth testing and almost always overlooked.
4️⃣ GEO WASTE
They bid the same across all markets despite wildly different customer values. We structured their campaigns into Tier 1 (high LTV), Tier 2 (scalable), and Tier 3 (test markets) to maximize returns.
5️⃣ CREATIVE BURNOUT & NO STRATEGY
At $2-3k/day, you can refresh ads sporadically. At $16k/day, you need a structured, rolling creative system. We built a process where 20% of spend always went into fresh ads to prevent fatigue. And we implemented our creative testing framework to systematically identify winning concepts and scale them into the core budget.
The Results
We scaled from $2k/day to $16k+/day in controlled phases, keeping CPAs within ~20% of baseline. Plateau busted. The high-level breakdown:
→ Phase 1 ($62k/month) -- Established baselines, pivoted bid strategy, implemented light creative testing framework.
→ Phase 2 ($106k/month) -- Launched competitor campaigns, geo split.
→ Phase 3 ($246k/month) -- Full Performance Max rollout, upgraded to a more sophisticated creative testing program.
→ Phase 4 ($493k/month) -- Optimized budget allocation, advanced automation and optimization roll-out, stabilized CPAs.
Worth noting: none of these phases were "set it and launch the next one." Each required recalibrating bids, budgets, and creative as the account dynamics shifted. Scaling is active, ongoing work.
👋 QUICK NOTE:
We've got a few open spots at our growth agency. Our strategists have 10-15+ years of experience designing and scaling paid acquisition engines. If you've hit a similar plateau or are scaling but aren't sure how to get to the next level, head on over to Demand Curve and get in touch. |
Why This Worked
1️⃣ MARKET TIMING
AI-generated video was exploding. They had what we call a Market Growth Catalyst on their side. Not exactly something that can be engineered. But it's absolutely something that should be identified and leveraged.
2️⃣ CONTROLLED SCALING
Budgets increased in 15-20% increments, avoiding the CPA spikes that come with aggressive budget jumps.
3️⃣ MULTI-CHANNEL SYNERGY
Performance Max generated demand. Search captured it.
4️⃣ GEO PRIORITIZATION & LTV OPTIMIZATION
We accepted higher CPAs in high-LTV regions because the unit economics justified it.
5️⃣ CREATIVE EVOLUTION
We mapped ad fatigue patterns and refreshed on a structured rotation, not gut feel. And our creative testing system enabled more personalized user experiences, which was critical for combating the ol' relevancy problem discussed earlier.
The Takeaway
8x in 90 days sounds like a headline.
In reality, it was the result of a methodical sequence of decisions made in the right order: zoom out, understand the problem, identify leverage points, develop hypotheses, experiment, learn, systematize, and repeat.
No comments:
Post a Comment