Growth hackers live and die by velocity — more experiments per sprint, faster hypotheses, tighter feedback loops. The bottleneck is rarely the idea; it's the writing that surrounds it: hypothesis docs, A/B briefs, funnel analyses, conversion copy, stakeholder summaries. AI is the unfair advantage for exactly that work — the parts that eat hours but don't require your intuition. These 10 prompts give it back.
Experiment Design & Hypothesis Writing
The quality of your experiments is determined before you run them. Sloppy hypotheses, undefined success metrics, and no prioritization framework waste cycles that could be shipping signal. These four prompts tighten the front end of your experiment pipeline — so every test you launch is scoped, measurable, and worth running.
1. Growth experiment hypothesis
The Prompt
"I want to run a growth experiment on [channel/feature]. Write a structured hypothesis using the format: We believe [change] will cause [outcome] because [rationale]. Include a measurable success metric and a suggested test duration."
Why it works: Most growth experiments fail the hypothesis test before they even launch — they describe a change, not a prediction. The structured format (We believe / will cause / because) forces you to state your assumption and your rationale explicitly, which is exactly what you need to learn from the result regardless of outcome. Adding a measurable success metric and test duration makes it actionable, not just theoretical.
2. A/B test brief
The Prompt
"Write a concise A/B test brief for [experiment idea]. Include: what we're testing, the control vs. variant, the primary metric, sample size guidance, and what a winning result looks like."
Why it works: A/B test briefs written quickly and informally are the ones that get argued about after the fact. Defining the control vs. variant in writing before launch eliminates post-hoc scope creep, and specifying "what a winning result looks like" forces you to set the success bar before the data comes in — not after you've seen it. Sample size guidance keeps you from calling a winner on 40 conversions.
3. Experiment retrospective
The Prompt
"We ran an experiment: [describe test and results]. Write a 200-word retrospective covering what we tested, what the results showed, what we learned, and what we'd do next."
Why it works: Experiment retrospectives are the most skipped document in growth and the most compounding one. A team that documents every test — wins and losses — builds a library of what works on its specific funnel. The 200-word constraint forces brevity without losing the four sections that actually matter: what we tested, what we saw, what it means, and what comes next.
4. Prioritization framework
The Prompt
"I have these growth experiment ideas: [list them]. Score each one using an ICE framework (Impact, Confidence, Ease, 1–10 each) and rank them. Explain your reasoning for the top 3."
Why it works: ICE scoring is only useful if it's applied consistently across the entire backlog — and doing that manually for a list of 15 ideas takes 30 minutes you don't have. This prompt applies the framework across all ideas at once and explains the reasoning for the top results, which means you can defend the prioritization to your team without a whiteboard session. The score is a conversation starter, not a verdict.
Want all your growth prompts in one place?
The Ultimate ChatGPT Prompt Pack includes 100+ ready-to-use prompts for growth hackers, marketers, and operators — $19, instant download.
Get 100+ ready-to-use prompts → Prompt Pack ($19)Funnel Analysis & Conversion Copy
Conversion is where growth work gets expensive if you're moving slowly. Drop-off hypotheses take time to write clearly; headline variants take time to generate in volume; activation emails take time to draft without sounding like a marketing template. These three prompts compress that work so you can test faster.
5. Funnel drop-off analysis
The Prompt
"My conversion funnel shows: [traffic → signups → activation → paid]. The biggest drop is between [step A] and [step B]. Write 3 hypotheses for why users are dropping off and one experiment idea for each."
Why it works: Funnel drop-off is easy to spot in the data and hard to diagnose without context. This prompt turns the bottleneck into a structured diagnostic: three distinct hypotheses tied to the specific drop point, each paired with a testable experiment. The output is a ready-made backlog item, not a vague observation — which means you go from "we're losing people at activation" to "here are three things we can test this sprint."
6. Landing page headline variants
The Prompt
"I'm testing a landing page for [product/feature]. Write 5 headline variants targeting [ICP]. Each should be under 10 words and speak to a different pain point or desired outcome."
Why it works: Headline testing is one of the highest-ROI experiments on a landing page and one of the slowest to generate manually. Writing 5 short, differentiated variants for the same product takes surprisingly long when you're also trying to cover distinct pain points and outcomes. This prompt generates a testable headline set instantly — each one under 10 words, each one hitting a different angle. Pick the two you'd run and move.
7. Onboarding email — activation nudge
The Prompt
"Write a short activation email (under 100 words) for a user who signed up for [product] but hasn't completed [key action] within 48 hours. Use curiosity + urgency, no pressure."
Why it works: Activation nudge emails fail in two ways: they're too long to read, or they feel like a follow-up from a sales rep. The "under 100 words" constraint forces the copy to stay punchy, and "curiosity + urgency, no pressure" is the tonal brief that keeps it from reading like a chaser. The result is an email that feels like a helpful nudge, not a check-in. Swap in your product name and action, and it's ready to test.
Channel Experiments & Reporting
Growth work doesn't end when the experiment does — you have to communicate the results upward and run outreach in parallel. Cold DMs, stakeholder summaries, and weekly standups are all writing tasks that don't require your judgment, just your time. These three prompts handle them.
8. Cold DM / outreach message
The Prompt
"Write a 3-sentence cold DM for [channel: LinkedIn / Twitter / Reddit] targeting [ICP]. Goal: get a reply, not a sale. Lead with a specific insight about their situation, not a product pitch."
Why it works: Cold DMs fail because they open with what you want, not what's relevant to the recipient. The "lead with a specific insight about their situation" instruction forces the output away from generic openers and toward something that demonstrates awareness. Three sentences is the format that works on DM channels — long enough to be specific, short enough to actually get read. The constraint to get a reply, not a sale, keeps the tone right.
9. Growth experiment summary for stakeholders
The Prompt
"We ran [experiment] over [timeframe]. Results: [data]. Write a 150-word summary for leadership that covers what we tested, what the numbers show, and our recommended next step."
Why it works: Leadership summaries of experiments fail when they lead with methodology instead of outcomes. The 150-word constraint keeps the summary from becoming a report, and the three-part structure — what we tested, what the numbers show, recommended next step — is exactly what a leadership audience needs to make a decision or a follow-up ask. It also gets you out of the habit of presenting data without a recommendation.
10. Weekly growth standup update
The Prompt
"Write a concise weekly growth update covering: experiments running this week, one win, one thing that didn't work, and the top priority for next week. Keep it under 200 words."
Why it works: Weekly growth updates written from scratch always run long because there's always more context to add. The four-section structure — running experiments, one win, one failure, top priority — forces the update to stay scannable and honest. Calling out one thing that didn't work builds credibility with your team and keeps the channel from becoming a highlights reel. Under 200 words means it gets read.
Growth is a game of iteration rate. The teams that run more experiments, learn faster, and communicate results clearly win — not because they're smarter, but because they move faster. These prompts remove the writing bottleneck from your experiment pipeline so more of your time goes to the judgment calls that actually require your brain. Start with whichever section is your current drag: hypothesis writing, funnel copy, or stakeholder reporting. One prompt, one sprint. Ship it.
Ultimate ChatGPT Prompt Pack
$19
100+ prompts for growth hackers, marketers, and builders. Copy, paste, get results.
No-Code Automation Starter Kit
$27
Automate the repetitive parts of your growth workflow. No coding required.