ThinkWithPrompt by SynotticThinkWithPrompt

Analyze mobile game ad creatives across ad networks

Compares creative performance network by network, separates fatigue from network effects, and tells you what to scale, iterate, or kill.

Marketing & BrandFew-ShotStandardAnalyzeFill-in TemplateInteractive

The prompt

Analyze the performance of my mobile game ad creatives across ad networks. Game and genre: [e.g. "casual puzzle, F2P with ads + IAP"] Data: [PASTE export - creative ID, network, country/tier, date, spend, impressions, installs, CTR, IPM, CPI, D1/D7 retention, ROAS D7 or other payback metric] Goal: [e.g. "scale to $X/day at D7 ROAS above Y%"] Creative notes: [format, length, hook/concept, playable or video, launch date for each creative] Rules: - Judge creatives relative to their network and geo, not in absolute terms. A creative can win on one network and fail on another because auctions, audiences, and formats differ. - Separate top-of-funnel signals (CTR, IPM) from quality signals (retention, ROAS). A high-IPM creative with weak retention is a warning, not a winner. - Check for fatigue: falling IPM or rising CPI over time at similar spend. - Flag results with too little spend or too few installs to trust. Return: 1. A summary of what's working and where. 2. Table: Creative | Network | Key metrics | Trend | Verdict (scale / iterate / pause) | Why. 3. Patterns across creatives (hooks, formats, lengths, concepts) that predict success, with the evidence. 4. 5 new creative test ideas based on those patterns, and which networks to test them on first. 5. Data gaps or tracking issues that could be skewing the read. Use only the numbers I give you; don't invent benchmarks. If you cite typical ranges, label them as rough and ask me to confirm.

Highlighted [brackets] are blanks to replace with your own details.

Fill in the blanks and use this promptMore Marketing & Brand prompts

What to fill in

  • e.g. "casual puzzle, F2P with ads + IAP"
  • e.g. "scale to $X/day at D7 ROAS above Y%"

How this prompt is built

This prompt follows the Few-Shot structure (Few-shot prompting). A few-shot prompt includes one or more worked examples of the input and the output you want, so the model can copy the pattern instead of guessing it.

Related prompts

Find more prompts like this