Video summary

Video or Static? And How Many Ads Can Your Budget Actually Afford?

Main summary

Key takeaways

Business

Core guidance: Static (Image) vs Video ads (Meta)

  • There is no strict rule for when to use static vs video.
  • General impact
    • Video ads: tend to build trust faster.
    • Static (image) ads: can underperform on trust unless created properly.

Funnel implication (example)

  • TOFU (top-of-funnel): use video ads
  • BOFU (bottom-of-funnel): often image ads work, but not always
  • Example quoted:

    “I have seen a lot of people scale their website just by using image ads” (Static can succeed if quality/fit is strong.)

Budget allocation + Meta learning: Creatives, spend, and experiments

Cost per lead / acquisition targets used

  • Assumed/mentioned CPL/CPA ≈ 90–100 rupees, then rounded to 100 for calculations.
  • Budget referenced: 1,000 rupees/day.

Rule for the number of creatives during testing

If your target cost per acquisition (or lead) ≈ 100:

  • With 1,000/day budget: use ~1 ad per day
  • Why: pushing more creatives can delay reaching conclusive results (slows learning and makes performance differences clearer/clearer more slowly).

Starting new vs scaling: how to split budget

  • If you have nothing already scaling:
    • The account’s “scaling” portion may be 0%, so most budget goes to testing/experimentation.
  • If you already have something scaling:
    • Allocate ~50% (or whatever it can properly return results on) to scaling, and the remainder to testing.
    • Example implication: if scaling takes half, testing may have only ~500/day left.

Spend threshold to conclude experiments faster

  • Experiment requirement:
    • Spend at least ~5× CPA/CPL per day on one experiment
  • Otherwise, conclusions can take much longer.

Numerical interpretation example (from the talk’s figures)

  • If CPA/CPL = 100 rupees, then 5× implies:
    • ~500 rupees/day per experiment minimum to reach conclusions faster.
  • Therefore, with 1,000/day, you can typically support only limited parallel experimentation.
    • This matches the “one ad per day” guidance to keep learning focused.

Practical “playbook” distilled from the talk

Choose format by funnel stage (typical starting point)

  • TOFU → Video ads (faster trust)
  • BOFU → Image/static ads
    • Validate performance
    • Static can scale if creatives are strong

During experiments

  • Keep spend concentrated
    • Aim for ~5× CPA/CPL per day per experiment
  • With 1,000/day and ~100 CPL/CPA
    • Test with ~1 creative per day, or
    • Limit parallel creatives to avoid slow/inconclusive learning

Budget split logic

  • New account / brand-new campaign: mostly testing
  • If something is already scaling: budget shifts toward scaling, reducing funds for tests

Metrics / KPIs mentioned

  • CPL/CPA target: ~90–100 rupees (used as 100 in examples)
  • Budget: 1,000 rupees/day (example)
  • Rule multiplier: daily experiment spend should be ≥ 5× CPA/CPL

Presenters / sources

  • The subtitles do not provide a name for the speaker(s) or any external source.

Original video