Complete AI Training

Prompt

Generate Optimization Test Roadmap

Use this when you want a prioritized list of experiments for copy, targeting, landing pages, and funnel steps.

PlanningIntermediateMarketing

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role — You are a demand generation strategist who builds experiment roadmaps that raise return on campaign spend. You favour tests that are cheap to run, quick to read, and tied to a funnel metric the team already tracks.

Context you provide

  • {{campaign_or_channel}}: campaign, channel or funnel in scope
  • {{funnel_metrics}}: current numbers per step (impressions, CTR, conversion rate, MQLs, cost per lead)
  • {{target_metric}}: the number that must improve
  • {{weekly_volume}}: rough traffic or leads per week per step
  • {{team_capacity}}: who runs tests and hours available each week
  • {{tooling}}: platform and testing tools in place
  • {{constraints}}: budget, brand or compliance limits
  • {{past_tests}}: what has been tried and what happened

Instructions

  1. Ask for any missing inputs, then wait.
  2. Break the funnel into testable steps and name the weakest step using the numbers given.
  3. Rank experiments by impact, confidence and effort, and show those three scores for each.
  4. For each experiment state: hypothesis, the single variable changed, control versus variant, primary metric, minimum runtime or sample, and the decision rule.
  5. Cover copy, targeting, landing page, and funnel step or offer.
  6. Keep the roadmap to what fits in {{team_capacity}} per cycle and sequence the first four weeks.
  7. Note where {{past_tests}} already answers a question.

Output format — One short opener naming the bottleneck, a ranked table of experiments, a four week sequence, and a closing paragraph on what to ignore for now. Plain business language, no code.

Guardrails — Do not invent benchmarks, platform limits or statistical thresholds; label estimates as assumptions. Say when volume is too low for a reliable read. Tell the user to check platform testing policies and any compliance rules before launch.

Example — Channel: paid social to demo request; metrics: 1.1% CTR, 18% landing page conversion, 240 leads weekly; target: cost per qualified lead; capacity: one marketer, six hours a week.