A/B Test Library

An experiment idea that varies only the cart CTA button color against the surrounding palette. See the setup, the tested element and the click-rate metric.

A cart experiment stripping repeated and non-functional calls to action down to a single one. Review both arms, the reasoning and the tracked metric.

An experiment idea adding a bar showing how much is left until free shipping. Examine the control arm, the added module and the basket value metric.

Compares an open coupon box in the cart with one behind a link. Read the two arms, the discount-hunting risk described and the revenue metric chosen.

An experiment idea placing forced account creation against an added guest checkout option. Look through the arm definitions, the friction case and KPI.

A structural experiment weighing a single checkout page against a stepped flow. See how each arm is framed and what completion metric is used here.

An experiment idea adding a countdown to a listing page campaign banner. Read both arms, the urgency reasoning and the resetting-timer credibility risk.

An open-ended experiment idea on how category order shapes browsing direction and discovery. See the tested variable and the conversion metric chosen.

A pagination experiment comparing continuous scrolling with button-triggered loading. Look at each arm, the orientation trade-offs and the metric used.

An experiment idea comparing two and three column mobile grids. See both arms and the trade-off between showing more products and keeping image detail.

Shoppers can swipe or hover to a second image on the card in this experiment. Read both arms, the note on detail page traffic and the conversion metric.

An experiment idea adding a quick add button to product cards on the listing page. Review both arms, the shortened purchase path and the tracked metric.

A new user opens a dashboard with no data yet. Compare a bare empty table with a card pointing to a first action, tracked as first-action completion.

A fixed widget order against an order driven by recent usage. The page sets out the predictability trade-off, the arms and the widget engagement metric.

A little-known feature introduced with one dismissible tip card, versus no introduction. Includes the interruption risk and the feature trial metric.

Labels above the input against labels aligned to its left. Read why scan path, vertical space and long or translated labels pull the decision both ways.

The floating label that starts inside the field and rises on typing, compared with a label that never disappears. Setup, arms and completion metric.

Fields laid out in two columns against the same fields stacked in one. Explains the zigzag scan problem, perceived length and how it gets measured.

Reordering every field from easy to hard, against leaving the current neutral order. Includes the counter-argument about pushing hard fields to the end.

Showing the whole form at once, or revealing fields as each one is filled. Read the expectation-setting argument against the perceived-length one.

Placing the cursor in the first field the moment the page loads, or leaving it alone. The page weighs the desktop gain against the forced mobile keyboard.

The primary call to action after the explanation section, against the same action in the first screen. Read the early-versus-convinced tension in full.

A landing page test idea comparing a headline that names the problem with one that states the solution. Review both arms, the setup and the KPI.

Compare a broad benefit line with a specific, measurable one. The page lays out both variants, the trade-off behind each and the KPI being tracked.

This test idea adds a second call to action at the foot of a long page. See the single-CTA control, the counter-arguments and the metric followed.

Swap the static hero image for a muted autoplaying video. The page shows both arms, the cost autoplay carries and the KPI the experiment tracks.

A trust test idea that adds a bar of recognizable client logos. Examine the control without logos, when logos can misfire and the metric tracked.

Enlarge one priority service icon while all the others stay equal in size. Compare both arms and see the risk of overshadowing the rest of the menu.

Forced signup at launch against guest mode first. Examine both arms, what each costs in first impression and the signup rate used as the KPI.

Push permission at app launch or after the first value moment. Compare both arms and see why a refused prompt is hard to ask for a second time.

Compare a plain loading screen with the same screen carrying one campaign message, keeping the actual loading time unchanged.

Give returning users a continue where you left off home screen. Examine both arms and why revenue per visitor is the metric being tracked here.

Add a recently viewed products row to the app home screen. Compare both arms and see why showing an already purchased item is flagged as a risk.

An open test concept for pricing-plan count. Define two specific plan sets first, then compare revenue per visitor and plan-selection distribution.

A test example putting a row-by-row plan comparison table against separate plan cards, with the reasoning for each format and the metric being tracked.

Experiment idea that moves the price above or below the benefit section, showing what each order does to expectation setting and to perceived value.

A/B test idea that swaps the headline price between a full annual figure and its monthly equivalent, with notes on how clearly the total must stay visible.

Experiment example adding a per-unit figure beside the package total, covering how it eases comparison and how it can crowd the price area or oversell.

A test idea that adds a no-auto-renewal line beside the price, with the subscription-fatigue reasoning behind it and the comparison burden it may create.

Experiment example swapping the main product image for a genuine, permitted customer photo, weighing realism and trust against perceived brand quality.

Compare one product image with four predefined product angles while keeping image quality consistent, using conversion and returns as outcomes.

Experiment idea adding a short line about the model's height and the size worn, and how that detail touches both size choice and the return rate.

Move the star rating and review count from directly below the price to just below the product description and compare downstream conversion.

Experiment example weighing a clean catalog against keeping the page live with a back-in-stock signup, measured by how many visitors register for alerts.

A test idea promoting the monthly installment figure over the full amount on a product page, with the transparency and cancellation risks it carries.

An experiment example comparing two trial CTA labels, a short neutral one against wording that names the trial length. Both arms and the signup KPI.

A SaaS experiment example: one self-serve CTA against a second demo request option for enterprise buyers. See both arms, the reasoning and the tracked KPI.

A pricing experiment example comparing a monthly default against an annual default plan selection. Read the commitment trade-off, arms and KPI.

An experiment example on lead form length: the current field set against a shorter one. Read the volume-versus-quality trade-off, arms and KPI.

A CRO experiment on when the upgrade prompt appears, before the user sees value or at the moment the limit is hit. Both arms and the upgrade KPI are shown.

A CRO experiment example comparing a badge that overflows the top of the plan card against one that stays inside it. See both arms, the setup and the KPI.

A CRO experiment on how filters are applied: results updating immediately against an apply button that submits selections together. Both arms and the KPI.

Compare an always-visible search field in the header with search placed inside the menu, tracking search use and downstream conversion.

Compare showing previous searches when the search box opens with an empty search box, tracking search use, conversion and privacy concerns.

An experiment example on one of the highest-exit screens: a blank no-results page against spelling suggestions, popular categories and similar products.

Compare relevance and best-selling as the default search sort while keeping the result set and filters unchanged, using search conversion as the primary KPI.

An experiment example comparing a standard menu against one that stays fixed while scrolling, weighing faster navigation against lost screen space.

A thank-you page experiment example adding a related product suggestion after checkout. See both arms, the reasoning behind the timing and the KPI.

This page details an account creation invite shown on the thank-you page after guest checkout: both arms, the friction argument and the measured KPI.

A referral experiment example showing an invite-a-friend offer right after purchase. See both arms, the timing rationale, the risks and the KPI.

Swapping a plain order confirmation for a personalized thank-you block, written up as an experiment example with both arms and the survey-based KPI.

A microcopy experiment example comparing an action label such as Send with a benefit label such as Get my quote. See both arms and the trade-offs.

A button style experiment example weighing a filled button against an outlined one. See both arms, the hierarchy reasoning and the completion KPI.

Should the secondary action be a text link or a button? This experiment example lays out both arms, the hierarchy trade-off and the completion KPI.

One shared button style across a page, versus section-by-section variation. See both arms of the experiment, the reasoning and the tracked KPI.

Compare Buy now with See pricing in the same CTA position and style, measuring progression and final conversion rather than clicks alone.

Pairing icons with written labels, set up as an experiment example. See the icon-only arm, the labelled arm, the clutter note and the tracked KPI.