Video summary

How to create killer ad creatives (to blow up your business)

Main summary

Key takeaways

Business

Core problem (what most brands miss)

  • Many e-commerce teams run high volumes of ads but can’t explain what creative “wins” really demonstrated or what to do next—they end up guessing between formats (UGC vs static vs carousel) rather than learning from customers.

The process: build “killer” ad creatives from customer evidence (not from format)

Framework / playbook

  • Start before production: don’t open Canva/film until you’ve defined the client problem with evidence.
  • Find the “buying reasons” using customer + market signals:
    • Existing reviews
    • Customer comments on social
    • Support tickets
    • Sales call objections (for SaaS/info products)
    • Post-purchase survey responses
    • Reddit + competitor review threads
    • Competitors’ ads themselves (to infer what they’re saying/promising)
  • Extract 3 points from that research:
    1. What the customer wants
    2. What prevents them from getting it
    3. What they already tried
  • Choose an argument (angle), not a creative type:
    • Decide the single message the ad must communicate.
    • Ensure every element supports that argument (avoid tangential claims like warranties/returns).
  • Prove claims visually:
    • If you claim “compact,” show size.
    • If you claim “fits under desk,” show it under the desk.
    • If you claim “easy to store,” show real storage (e.g., closet/apartment, moving ability).
    • If you cite numbers, use the real quantified data.

Creative-direction example (compact treadmill)

  • Customer wants: to move more while working from home
  • Constraints include:
    • Sitting 8 hours and barely moving
    • No room for a treadmill
    • Need quiet during calls
  • Creative angle becomes: “Take more steps while you work without needing a giant, noisy machine.”
  • Proof in ad content:
    • Under-desk use
    • Quiet during work/video calls
    • Folds/stores away

Why this beats “feature dumping”

  • Avoid turning the ad into an “infomercial” by listing all value props at once (returns, warranty, motor, dimensions, etc.).
  • For cold traffic, you must paint the value picture to build desire; landing-page details can handle policies and broader specs.

Creative testing: learn quickly, isolate variables, and scale what’s proven

Testing philosophy

  • Build a creative machine through iteration:
    • Each round should answer what to do next.
    • Winners become the foundation for more variations.
  • Test one variable at a time:
    • If you change hook + audience + landing page + offer + creative all at once, you learn “nothing” about which change caused improvement.
  • Run “level playing field” tests:
    • Give competing ideas (e.g., UGC vs carousel) equal visibility/budget, then analyze differences fairly.
  • Don’t confuse “sales while spending” with creative success:
    • An ad may look “winning” because of volume/placement/exposure—not because the creative itself worked.

Budget and timeline guidance for learning (small accounts)

  • You must match creative test volume to budget:
    • If spending is low (e.g., $30–$100/day), you can’t run unlimited new ads and expect clear learning.
  • Prefer focused batches and let the platform collect enough signal before deciding.
  • Target: find winners in the first or second week.

Concrete case example (new account spending $30/day)

  • Launched: 13 ads
  • Initial print/limit: 1,000
  • Observed CPMs: $50–$60–$70
    • Conclusion: a claimed $17 CPA was “unrealistic” for that context.
  • Adjusted print limit to: 800 to finish efficiently without burning budget.

Creative interpretation

  • One ad spent heavily from day one
  • Facebook would otherwise keep spending until very high spend tiers ($150–$200–$400)
  • Reason it kept getting budget: high CTR, link click performance, and lower-than-average CPM

Key takeaway

  • The team didn’t assume performance—they diagnosed why it worked.

Metrics and KPIs: don’t just look at CPA

Business KPI stack (customer-journey view)

  1. Attention
    • CTR (ad click-through rate): did it stop the scroll?
  2. Intent / traffic quality
    • Link CTR and the drop from CTR → link clicks
  3. Conversion
    • Purchases and downstream conversion rate (site/landing/funnel performance)

Example diagnostic findings

  • Ad A
    • CTR: 10.9% (high engagement)
    • Few purchases
    • Interpretation: visually compelling, but customer didn’t understand what’s being sold or why they should buy (poor message-to-landing alignment).
  • Ad B
    • CTR: 6.3% (healthy but lower attention)
    • Generated purchases
    • Showed the routine/before-after/how to apply (clear product/value communication).

Additional diagnostic detail

  • Ad A: >3% drop from ad-level CTR to link click CTR → losing people after they realize the ad isn’t communicating enough.
  • Ad B: <2% drop → people clicked because they were already clearer and more buy-ready.

Core concept

  • Treat performance like a two-part puzzle:
    • Cost to acquire attention (top-of-funnel)
    • Conversion rate (mid/bottom-of-funnel)

Warning

  • A “good CPA” may be misleading if it only works at low spend or depends on retargeting.

Goal

  • Build ads that can support scaling by performing across funnel stages, not only at small budgets.

Ultimate aim

  • Consistently achieve return ≥ 3x (ROAS-like framing) before expanding.

What to do next after each test (turn results into briefs)

Playbook

  • “Every creative test should be your next creative brief.”
  • If an angle worked:
    • produce more content around it (e.g., 3 different video bodies with the same hook).
  • If a format/style underperformed (e.g., UGC review style):
    • pivot toward creatives that better prove value (e.g., more product demonstrations).
  • Example “next-iteration” decisions:
    • Replicate a winning competitor angle that customers discuss negatively (use it to address objections/differentiate).
    • Instead of “AI creating more images,” create based on what customers responded to last week.

Concrete instruction

For the next creative round:

  • Choose a client problem
  • Write the focus as a single sentence
  • Make every line of the ad support that sentence
  • Change one important thing at a time

Presenters / sources

  • Presenter: ADVT (the speaker / consultant), name not provided in subtitles

Original video