Experiment hypothesis prioritization

실험 가설 우선순위

Rank the next tests by likely impact, confidence, and effort.

···
html
<div class="priority-demo"><div class="priority-title">NEXT EXPERIMENT <span id="evidence-label">EVIDENCE: LOW</span></div><div class="priority-row" data-key="shipping"><b>SHIPPING CLARITY</b><i></i><strong></strong></div><div class="priority-row" data-key="form"><b>SHORTER FORM</b><i></i><strong></strong></div><div class="priority-row" data-key="copy"><b>NEW CTA COPY</b><i></i><strong></strong></div></div>
css
.priority-demo{width:min(87%,485px);font:700 clamp(10px,3.2vmin,15px)/1.15 sans-serif}.priority-title{display:flex;justify-content:space-between;margin-bottom:clamp(8px,2vmin,17px)}.priority-title span{color:var(--accent)}.priority-row{display:grid;grid-template-columns:minmax(90px,1fr) 33% 25px;gap:7px;align-items:center;border:1px solid var(--line);padding:clamp(5px,1.5vmin,11px);margin:5px 0;transition:transform .5s}.priority-row i{display:block;height:clamp(8px,2vmin,16px);background:var(--accent);transition:width .5s}.priority-row strong{color:var(--accent)}.priority-row.top{border-color:var(--accent);background:var(--surface)}
js
const rows=[...document.querySelectorAll('.priority-row')],label=document.querySelector('#evidence-label');let high=false;function draw(){const scores=high?{shipping:8,form:6,copy:3}:{shipping:4,form:6,copy:3};rows.forEach(row=>{const score=scores[row.dataset.key];row.style.order=String(10-score);row.querySelector('i').style.width=(score*12)+'%';row.querySelector('strong').textContent=String(score);row.classList.toggle('top',score===Math.max(...Object.values(scores)))});document.querySelector('.priority-demo').style.display='flex';document.querySelector('.priority-demo').style.flexDirection='column';label.textContent=high?'EVIDENCE: HIGH':'EVIDENCE: LOW';high=!high}draw();setInterval(draw,2700);

Hypothesis prioritization decides what to test first when ideas outnumber capacity. Compare expected impact, confidence in evidence, and implementation effort using a shared rule; the score supports judgment.

The demo reranks three hypotheses when evidence confidence changes. After selecting one, specify the observation and success metric.

Precise-looking scores still depend on uncertain estimates. Record the original rank and what the test taught you so the criteria can improve.

When to use

Use it when test ideas exceed engineering capacity or available traffic.

Open as page ↗