Most analytics dashboards offer dozens of metrics. Most of them are noise. Effective CRO requires tracking a small set of metrics that are directly linked to revenue — and ignoring the rest. Here's how to think about what to measure.
The Core Conversion Metric
Every CRO program starts here:
Conversion Rate = (Conversions ÷ Unique Visitors) × 100
This is your north star. Everything else either explains why it is what it is, or helps you predict where it could go. Track it by page, by traffic source, and by device — a 4% desktop conversion rate and a 1.2% mobile rate tell very different stories.
Funnel Metrics: Where Visitors Drop Off
A single conversion rate number hides the shape of your funnel. Break it down by stage:
| Stage | Metric | What It Reveals |
|---|---|---|
| Landing page | Scroll depth, time on page | Whether messaging is engaging visitors |
| CTA click | Click-through rate | Whether the offer is compelling |
| Form start | Form start rate | Whether the friction point is the form |
| Form completion | Form completion rate | Whether the form is too long or complex |
| Thank-you / order | Final conversion rate | End-to-end success |
A high CTA click rate but low form completion rate tells you the offer is working — but the form is killing you. That's where to focus.
Click-Through Rate (CTR)
CTR measures what percentage of visitors who see a given element click it — a CTA button, a product card, a navigation link.
CTR is most useful when:
- Comparing variants in an A/B test
- Diagnosing whether a specific element is underperforming
- Evaluating ad creative and email subject lines
On its own, CTR is a leading indicator, not a success metric. A button with a 20% CTR but a 0.5% purchase rate is not performing well — the downstream steps are broken.
Bounce Rate
Bounce rate is the percentage of sessions where a visitor lands on a page and leaves without taking any action. High bounce rates on landing pages often indicate:
- Messaging mismatch between ad/source and page content
- Slow page load times
- Content that doesn't immediately communicate value
A 70% bounce rate on a blog post is unremarkable. A 70% bounce rate on your pricing page is a serious problem.
Revenue Per Visitor (RPV)
For ecommerce and SaaS, revenue per visitor is often more useful than raw conversion rate. It accounts for order value and purchase frequency — not just whether someone converted.
RPV = Total Revenue ÷ Total Visitors
Two variants can have the same conversion rate but very different RPV if one consistently leads to higher-value purchases. Always include RPV as a secondary metric in ecommerce tests.
Metrics to Watch Carefully (But Not Obsess Over)
Time on Page
High time on page can mean your content is compelling — or that visitors are confused and struggling to find what they need. Interpret it alongside scroll depth and exit rate before drawing conclusions.
Pages Per Session
Useful for content sites. Less meaningful on single-purpose landing pages where the goal is conversion, not exploration.
Pageviews
A vanity metric for CRO purposes. More pageviews with the same conversion rate means more traffic — not better optimization.
Setting Up a CRO Measurement Framework
A practical CRO measurement setup has three layers:
- Primary metric — The conversion event that defines success (sign-ups, purchases, demo bookings)
- Secondary metrics — Supporting indicators that explain the primary (CTR, form completion rate, RPV)
- Guardrail metrics — Metrics you monitor to ensure tests don't have unintended negative effects (cart abandonment rate, support ticket volume)
Define these before you run any test. Teams that define metrics after seeing results end up cherry-picking their way to false confidence.
How AI Changes the Metrics Picture
Traditional A/B testing optimizes for a single metric against a fixed traffic split. AI-driven platforms like Surface AI optimize across multiple metrics simultaneously — balancing conversion rate with revenue per visitor and session quality in real time, without requiring manual configuration for each experiment.