CRO Glossary

A/B testing is a controlled experiment that compares two versions of a webpage or element to determine which one performs better on a target metric.

Above the fold refers to the portion of a webpage visible to a user without scrolling, considered prime real estate for key messages and calls to action.

Audience segmentation is the practice of dividing website visitors into groups based on shared characteristics — such as behavior, demographics, or traffic source — to deliver targeted experiences and analyze experiment results.

Average order value is the mean revenue generated per completed transaction, calculated by dividing total revenue by the number of orders in a given period.

Bayesian A/B testing is an experimental framework that quantifies the probability one variant beats another using prior beliefs updated with observed data, rather than accepting or rejecting a null hypothesis.

Bounce rate is the percentage of sessions in which a visitor leaves a page without taking any action or navigating to another page on the site.

A call to action is an element — typically a button, link, or form — that prompts a visitor to take a specific desired step, such as signing up, buying, or requesting a demo.

Click-through rate (CTR) is the percentage of people who click on a link, button, or ad after seeing it.

Client-side testing is an A/B testing approach where a JavaScript snippet loads in the browser and modifies the page after it has been delivered from the server.

Cohort analysis is the study of a group of users who share a common characteristic or experience within a defined time period, used to track behavior and retention over time.

A confidence interval is a range of values that is likely to contain the true effect of a change in an A/B test, giving you a measure of uncertainty around your result.

An adaptive testing algorithm that selects the best variant for each individual user based on their context — like device, location, or behavior — rather than allocating traffic uniformly.

A conversion event is a specific user action — tracked via a JavaScript event tag — that signals a visitor has completed a desired goal, such as a form submission or purchase.

A conversion funnel is the series of steps a visitor takes from first arriving on your site to completing a desired action, like signing up or making a purchase.

Conversion rate is the percentage of visitors who complete a desired action on a page, such as signing up, making a purchase, or submitting a form.

Conversion rate optimization (CRO) is the systematic process of increasing the percentage of website visitors who complete a desired action, such as signing up, purchasing, or booking a demo.

A variance reduction technique that uses pre-experiment data to make A/B tests more sensitive, reducing the sample size needed to detect a given effect.

Customer acquisition cost (CAC) is the total cost of acquiring a new customer, including marketing, sales, and advertising expenses divided by the number of new customers gained.

Customer lifetime value is the total net revenue a business expects to earn from a customer over the entire duration of the relationship.

Dynamic landing pages automatically change their content — headlines, images, CTAs, or layout — based on visitor attributes like traffic source, location, or behavior.

An edge route is a URL path handled at the CDN edge — the server closest to the visitor — enabling near-instant responses for tasks like variant assignment, redirects, and personalization without a round trip to the origin server.

A feature flag (also called a feature toggle) is a mechanism that lets teams enable or disable features in production without deploying new code, commonly used for gradual rollouts and experimentation.

The flicker effect is the brief flash of the original page content that users see before client-side A/B testing JavaScript applies a variant, causing a visible jump in the page layout.

Friction refers to any element of a user experience that creates resistance, confusion, or effort, reducing the likelihood of completing a desired action.

A guardrail metric is a secondary measurement in an A/B test that must not degrade below a defined threshold, even when the primary metric improves.

A heat map is a visual representation of user behavior on a webpage, showing where visitors click, move their mouse, and how far they scroll.

A holdout group (or control group) is a segment of users deliberately excluded from an experiment or feature rollout, used to measure the true impact of changes over time.

An interaction effect occurs when two simultaneous A/B tests influence each other's results, causing one or both experiments to produce misleading conclusions.

A key performance indicator (KPI) is a measurable value that tracks how effectively a team or campaign is achieving its core business objectives.

A landing page is a standalone web page designed for a specific marketing campaign or traffic source, with a single focused call to action.

Lift is the percentage improvement in a metric (like conversion rate) that a test variant achieves compared to the control, measuring the real impact of a change.

The minimum detectable effect is the smallest relative change in a metric that an A/B test is designed to reliably detect given a fixed sample size, confidence level, and statistical power.

Multi-armed bandit testing is an adaptive experimentation method that dynamically allocates traffic to better-performing variants during the test, reducing waste and reaching results faster than traditional A/B tests.

Multivariate testing is an experimentation method that tests multiple page elements simultaneously to find the best-performing combination.

The novelty effect is a temporary lift in engagement or conversions caused by the unfamiliarity of a new variant, which fades as users acclimate to the change.

A p-value is the probability that the observed difference between test variants is due to random chance rather than a real effect. Lower p-values indicate stronger evidence.

The peeking problem is the inflation of false positive rates that occurs when experimenters check results during a running test and stop early upon seeing significance.

Web personalization is the practice of dynamically tailoring page content, layout, or offers to individual visitors based on their behavior, attributes, or context.

The tendency for users to click, read, or convert more based on where an element appears on a page than its actual content or quality.

A progressive rollout is a deployment strategy where a new feature is gradually released to increasing percentages of users, reducing the risk of widespread issues.

A rage click is a rapid succession of clicks on the same element by a frustrated user, typically indicating a broken interaction, confusing UI, or unmet expectation.

A referrer is the source that sent a visitor to your website — such as a search engine, social media platform, another website, or a direct URL entry.

Regression to the mean is the statistical tendency for extreme measurements to move closer to the average on subsequent observations, which can mislead CRO teams into attributing natural variation to a test change.

Revenue per visitor (RPV) is the average revenue generated per website visitor, combining conversion rate and average order value into a single metric.

Sample size is the number of visitors or users needed in an A/B test to detect a meaningful difference between variants with statistical confidence.

Scroll depth is a behavioral metric that measures how far down a page a visitor scrolls, typically expressed as a percentage of total page height.

An experimental method that allows continuous monitoring of results and valid early stopping, without inflating false positive rates the way traditional peeking does.

Server-side testing is an A/B testing approach where variant logic executes on the server before the page is delivered to the browser, eliminating the visual flicker that browser-based tools can cause.

A session is a single visit to a website, starting when a user arrives and ending after a period of inactivity or when they leave. Sessions are the standard unit of measurement in web analytics and experimentation.

Social proof is a psychological phenomenon where people look to the behavior and opinions of others to guide their own decisions, applied in CRO through reviews, testimonials, and usage statistics.

Statistical power is the probability that a test correctly detects a real effect when one exists, typically set at 80% for A/B testing.

Statistical significance is a measure of confidence that the result of an A/B test reflects a real difference rather than random chance.

A Type I error is a false positive in hypothesis testing — concluding that a variant outperforms control when no real difference exists.

A Type II error is a false negative in hypothesis testing — failing to detect a real effect because the test lacked sufficient power or sample size.

A unique visitor is a single person who visits your website during a given time period, counted once regardless of how many times they return.

UTM parameters are tags added to URLs that track where traffic comes from, which campaign drove it, and what link was clicked — enabling precise marketing attribution.

A variant is an alternative version of a webpage or element being tested against the original (control) in an A/B or multivariate experiment.

Win probability is the estimated likelihood that a test variant will outperform the control, expressed as a percentage and updated continuously as data comes in.