# Highlighting One Option With Background Color A/B Test

Page: https://alidemirbas.com.tr/en/lab/ab-test-library/tests/highlighting-one-option-with-background-color-ab-test

## Hypothesis

If one predefined option gets a contrasting background, visitors may scan the choices faster and decide sooner. That could shorten selection completion time, though the gain only counts if the share of visitors who never make a selection does not rise.

## When to run this test

Plan this test when visitors seem to hesitate on a selection screen, such as a long gap between the start and selection events, session recordings of the cursor moving back and forth between options, or support tickets asking which option fits. A helpful highlight means visitors choose in less time, selection completion rate does not fall and later changes do not climb. If the hesitation comes from unclear option descriptions, fix the wording first. Pick the option to highlight because it fits what visitors actually need, and do not add a "most popular" label without data to support it.

## Primary metric

**Selection completion time**: Use the same start and selection events in both arms, and interpret completion time with the share who do not complete.

## Other metrics

- **Post-selection change rate**: Track later changes among users who selected an option over the same follow-up period; not every change means regret.
- **Visibility of other options**: Use the same event definition to check whether non-highlighted options remain visible.
- **Selection completion rate**: Compare completed selections among all eligible assigned visitors.
- **Post-selection satisfaction**: If surveyed, use the same question, timing and sampling rule in both arms; omit this measure if there is no survey.

## What to check before and during the test

- **Change**: Change only the background color of one predefined option.
- **Selection mix**: Read shifts in selection alongside overall completion.
- **Later changes**: Track corrections and cancellations over the same period.
- **Alternatives**: Keep other options' text, order, size and access unchanged.
- **Experience**: Review first-time and experienced users separately.

## What to watch when reading the result

- Always read selection completion time alongside the share who never select; faster completers can hide more people giving up. If choices shift toward the highlighted option, weigh that against overall completion and the post-selection change rate, without treating every change as regret.
- The highlight color must keep text contrast at an accessible level. If visibility of other options drops, or they become hard to read or click, do not treat a time gain as a reason to ship.
- What you learn is specific to the highlighted option and the selection screen you tested; highlighting a different option is a separate test. Review first-time and experienced users separately, and track corrections and cancellations over the same follow-up window.
- Time measures can be pulled by a few very long sessions, so look at the spread instead of a single average. The [A/B test significance calculator](https://alidemirbas.com.tr/en/calculators/ab-test) shows whether the gap in selection completion rate is beyond chance; it cannot test the time gap itself, so read that with the same caution.

## Frequently Asked Questions

### Why can selection time be misleading?

Faster completion among finishers can conceal more abandonment. Compare time with nonselection and completion rates rather than considering only completers. Use selection completion time as the primary outcome, read together with the nonselection rate.

### Is movement toward the highlighted option a good sign?

Not alone. Assess the shift alongside total completion. Track later changes, corrections and cancellations without calling every change regret. Avoid steering people away from their needs or making unsupported most-popular claims.

### What should the background color preserve?

Maintain required contrast and keep other options readable and usable. Measure exposure and review new and experienced users separately.