Research & Analysis / Data Interpretation
Compare your metrics to industry standards, past performance, or control groups with automated significance testing.
Why This Prompt Exists
“Is 5% conversion rate good?” Without a benchmark, you have no idea — and most people guess.
You get:
- celebrating performance that’s actually below average
- panicking about metrics that are fine given your context
- decisions based on “feels good” instead of relative comparison
- missing real problems because you have no baseline
- stakeholders asking “compared to what?” and you having no answer
But benchmarks provide context:
- industry standard: how do similar companies perform?
- past performance: are we improving or declining?
- control group: did our intervention actually do anything?
- target/goal: are we meeting expectations?
- competitor: are we winning or losing?
Without comparison, numbers are meaningless.
This prompt compares your metrics to relevant benchmarks with statistical rigor.
The Prompt
Assume the role of a performance analyst who benchmarks metrics. Your task is to compare your data to relevant benchmarks and interpret the difference. Generate: 1. COMPARISON SUMMARY - Your metric: [value] - Benchmark: [value and source] - Raw difference: [absolute and percentage] 2. STATISTICAL ASSESSMENT - Is the difference statistically significant? (if data available) - Margin of error / confidence interval for your metric - Is the difference practically meaningful? 3. CONTEXT FACTORS - Why you might outperform benchmark (advantages) - Why you might underperform (disadvantages) - Differences in measurement or definition 4. TREND ANALYSIS (if time-series data) - Are you closing the gap or widening it? - Rate of change vs. benchmark rate 5. RECOMMENDATION - Celebrate (you're genuinely ahead) - Investigate (gap exists, cause unknown) - Take action (gap is meaningful and negative) - Adjust benchmark (not the right comparison) INPUTS: Your metric: [E.G., "Email open rate = 22%"] Benchmark source and value: [E.G., "Industry average for SaaS = 25% (Mailchimp 2024 report)"] Sample size / confidence (for your metric): [E.G., "N=50,000 emails, margin of error ±0.5%"] Context / industry: [E.G., "B2B SaaS, enterprise customers"] Time period: [E.G., "Q2 2026"] RULES: - Distinguish between statistical and practical significance - Flag when benchmarks are from different populations (not apples-to-apples) - Note that average is not always the right target (sometimes you want to be above or below) - Consider whether the benchmark is aspirational or realistic
How To Use It
- Run this before any performance review — always bring a benchmark.
- For each KPI in your dashboard, define what “good” means (benchmark + target).
- Use industry benchmarks from sources like Gartner, Forrester, or industry reports.
- Compare to your own past performance as a baseline (year-over-year, quarter-over-quarter).
- When benchmarks aren’t available, use a control group from your own data.
Example Input
Your metric:
“E-commerce conversion rate = 3.2%”
Benchmark source and value:
“Industry average for apparel e-commerce = 2.5% (Statista 2025)”
Sample size / confidence:
“N=250,000 sessions last month”
Context / industry:
“DTC apparel, average order value $65”
Time period:
“May 2026”
Why It Works
Most performance reporting says “we’re at X%” without answering “is that good or bad?”
This framework improves outcomes by forcing:
- comparison summary (your number vs. benchmark, clearly stated)
- statistical assessment (is the difference real or noise?)
- context factors (why the comparison might be unfair — or generous)
- trend analysis (are you improving against benchmark?)
- clear recommendation (celebrate, investigate, or act)
Great benchmarking doesn’t just compare numbers — it tells you what to do next.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI coding tools, debugging frameworks, and practical strategies for developers and engineers.