# Presenting Information as a Table or as Cards A/B Test

Page: https://alidemirbas.com.tr/en/lab/ab-test-library/tests/presenting-information-as-a-table-or-as-cards-ab-test

## Hypothesis

If visitors choose by comparing options attribute by attribute, a table with aligned rows may help more of them finish the choice. If they scan options one at a time, cards that group each option's details may come out ahead on selection completion rate.

## When to run this test

Run this comparison when visitors on a selection page seem to struggle to compare options, such as heatmaps with movement back and forth between options, long pauses in session recordings, or support tickets asking how the options differ. If one format fits the way visitors decide, more of them complete a selection in that arm, with fewer exits. If the trouble comes from too many options, missing information or unclear labels, fix the content first; this test only compares two ways of presenting the same information.

## Primary metric

**Selection completion rate**: Compare eligible assigned visitors who complete the predefined selection after seeing the information in a table or cards.

## Other metrics

- **Time to select**: Measure time among visitors who start a selection and report non-completions separately.
- **Page exit rate**: Compare exits among eligible assigned visitors within the same follow-up window.

## What to check before and during the test

- **Format**: Compare a data table with separate option cards.
- **Layout**: Define one layout for each format before the test.
- **Information**: Keep attributes, values, labels and option order the same.
- **Differences**: Make sure differences between options remain clear in both formats.
- **Mobile**: Keep content readable, including any necessary horizontal scrolling.

## What to watch when reading the result

- Selection completion rate makes the call. Time to select only covers visitors who start a selection, so it can mislead on its own; a shorter time that comes with a higher page exit rate is not a win.
- In cards, state the differences between options in text instead of implying them through order alone. Use semantic table markup only for data that really is tabular; do not mark up a layout table that way, and do not leave a real data table unmarked. Otherwise the two arms do not give screen-reader users the same information.
- Only the content type and option set you tested are covered, so validate again before applying the result to other comparisons. Break results out by device, because a table may need horizontal scrolling on a narrow screen.
- The gap between table and cards may be small. Enter the selection completion rate you expect and the smallest gap you care about in the [sample size calculator](https://alidemirbas.com.tr/en/calculators/sample-size-calculator) to get the visitors needed per variant; if your traffic cannot reach that, do not report either format as the clear choice.

## Frequently Asked Questions

### How can a card layout obscure differences between options?

Grouping information by option can make comparison harder. Keep differences explicitly stated in both formats rather than implying them through order. Use completion of the defined selection as the primary outcome.

### Does faster selection mean cards are better?

Measure time among people who start choosing and report noncompleters separately. Compare page exits within the same follow-up window too.

### How should the table work on mobile?

Keep it readable, allowing necessary horizontal navigation rather than shrinking text excessively. Do not mark layout-only structures as semantic data tables. Avoid generalizing one content type to every comparison.