On the podcast I talk with Eric about how measurement dysfunction paralyzes growth, why diversifying channels for the sake of diversification actually hurts performance, and the futility of trying to interpret why ads win.
Top Takeaways:
đ Broken measurement kills growth
The biggest pitfall isnât creative or channel choiceâitâs disorganized measurement. When finance, product, and UA each use different models, growth stalls. The fix isnât another dashboard; itâs alignment. Build one coherent, incrementality-aware framework everyone trusts, with clear definitions of success and outputs that meet each teamâs needs.
đ Donât diversify just to diversify
Spreading budget across more channels feels safer but often reduces performance after integration, creative, and reporting overhead. Start with a waterfall method: max out your primary channel until ROAS hits your threshold, then move to the next. Diversify for scale or cross-channel effectsânot optics.
đČ Stop asking why an ad worked
Winners often defy tidy explanations. Treat individual ad outcomes as stochastic and largely uninterpretable. Put your energy into the system: feed diverse concepts, automate prospecting/synthesis, and measure whether your process is increasing the rate of wins over time. Learn from inputs and processânot post-hoc stories about outputs.
⥠Ship speed over certainty early
You wonât have fully baked LTV or incrementality in week one. Push spend methodically: kill obvious losers immediately, let plausible winners age, track cohort ROAS at day-7/30/60, and widen budgets as curves support it. Iterative frontier-pushing beats premature âterminal LTVâ guesswork.
𧩠Engineer better signals
Algorithms optimize to the signals you send. Create intentional, high-intent events (light âhurdlesâ that correlate with LTV) and send those back to platforms. Better signals shift spend toward durable users and compound efficiency, especially as automation on major platforms accelerates.
About Eric Seufert:
đšâđ» Quantitative marketer, media strategist, investor, and author.
đ Eric shares expert advice on the Mobile Dev Memo blog and is an investor at Heracles Capital.
đĄ âThe way I approach creative testing is trying to identify losers as quickly as possible. The winners take time to prove out, but the losers are pretty quick to prove out.â
đ LinkedIn
Follow us on X:
Episode Highlights:
[1:00] Intelligent design: How to effectively incorporate AI into your business strategy.
[4:52] I, Robot: Machine learning =/= generative AI.
[8:36] AI Pitfalls: AI works best for automating tasks and coming up with ideas â not generating brilliant creative assets.
[17:29] Predictive AI: Brand-specific, full-fidelity video ads generated by AI could be a reality within 18 months.
[33:25] Risky business: How to effectively diversify across advertising channels to optimize ROAS-adjusted spend.
[37:43] Measure of success: Above all, make sure your measurement system is coherent and has cross-team alignment.
[42:04] Tortoise vs. hare: To balance speed and efficiency, identify your ad âlosersâ as quickly as possible.
[44:43] Missed opportunity: Good marketing comes down to embracing some uncertainty and minimizing the rest.
[49:23] Human touch: Why generative AI creative tools probably arenât a worthwhile investment right now.
