How to Prioritize CRO Tests: ICE, PIE, and Impact Frameworks

The Prioritization Problem

A mature CRO research process generates more test ideas than you can run simultaneously. With limited developer time and traffic, choosing which tests to run first is itself a strategic decision — running the wrong tests first wastes months of testing cycles.

Prioritization frameworks convert subjective judgment ("I think this matters") into structured, defensible scoring that can be reviewed and debated by the team.

Our CRO team uses a hybrid framework that draws from the models below.

The ICE Framework

Created by Sean Ellis (Growth Hacking pioneer). Score each test idea on three criteria, 1–10:

• I — Impact: If this test wins, how much will it improve the conversion metric? (1 = minor lift, 10 = transformational)
• C — Confidence: How confident are you this will win? Based on quality of evidence supporting the hypothesis (1 = gut feel, 10 = strong multi-source evidence)
• E — Ease: How easy is this to implement? (1 = months of dev work, 10 = 30-minute CSS change)

Score = Average of I + C + E

Higher ICE score = run this first.

ICE strength: Simple and fast. Works well for early-stage programs where you need to build momentum.

ICE weakness: "Ease" can over-weight easy but low-impact tests. Teams game the score.

The PIE Framework

Developed by Chris Goward at WiderFunnel. Similar to ICE but with different criteria:

• P — Potential: How much room for improvement exists? (A page converting at 0.5% has more potential than one at 8%)
• I — Importance: How much traffic does this page receive? A 1% lift on a high-traffic page beats a 3% lift on a low-traffic page in absolute revenue
• E — Ease: Implementation complexity, same as ICE

Score = Average of P + I + E

PIE strength: Explicitly weights traffic volume, which keeps focus on pages that matter commercially.

PIE weakness: Doesn't explicitly score evidence quality (confidence). Two tests with equal PIE scores but vastly different evidence quality should not be treated equally.

Custom Impact Framework (Recommended)

For mature CRO programs, a custom framework combining the best of both adds nuance:

Weighted total determines priority rank. The "strategic learning value" factor captures tests that might not have the highest immediate impact but will answer questions that inform many future decisions.

Practical Prioritization Tips

• Run a priority score review meeting monthly: New research may change scores; don't let the backlog stagnate
• Don't run more than 3–4 tests simultaneously: Traffic splits between too many tests make all tests slower to reach significance
• Always include one "fast win" in each sprint: Easy-to-implement tests with good evidence maintain team momentum even during longer tests
• Track predicted vs. actual scores: Over time, if certain hypothesis sources (heatmaps, surveys) consistently produce winning tests, weight them higher in confidence scores

Connect your testing results back to GA4 conversion data and present ROI from your CRO program quarterly. Businesses that measure CRO ROI consistently increase their testing budgets over time.

Our CRO team builds prioritized test backlogs as part of every engagement. Contact us to develop a CRO testing roadmap for your business.