> For the complete documentation index, see [llms.txt](https://docs.absolutely.io/absolutely/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.absolutely.io/absolutely/analytics.md).

# Analytics

Read your results, understand the numbers, and decide what to do with them.

Every test has a results page that compares its groups. This section explains how to read it and what to do with what you find.

| Page                                                                                                        | What it answers                                        |
| ----------------------------------------------------------------------------------------------------------- | ------------------------------------------------------ |
| [Read your test results](/absolutely/analytics/read-your-test-results.md)                                   | What's on the results page                             |
| [How visitors and orders are counted](/absolutely/analytics/how-visitors-and-orders-are-counted.md)         | Who's in a test, and which group gets an order         |
| [Metric definitions](/absolutely/analytics/metric-definitions.md)                                           | What every number means                                |
| [Confidence and statistical significance](/absolutely/analytics/confidence-and-statistical-significance.md) | When a result can be trusted                           |
| [Decide: roll out, keep control or retest](/absolutely/analytics/decide-roll-out-keep-control-or-retest.md) | How to turn results into a decision                    |
| [Track profit: add product costs](/absolutely/analytics/track-profit-add-product-costs.md)                  | How to make results show profit                        |
| [Filter and segment results](/absolutely/analytics/filter-and-segment-results.md)                           | How to look at one device, country or source           |
| [Custom events](/absolutely/analytics/custom-events.md)                                                     | How to measure clicks, page visits and your own events |
| [Export results and orders](/absolutely/analytics/export-results-and-orders.md)                             | How to get the data out                                |
| [Analytics FAQs](/absolutely/analytics/analytics-faqs.md)                                                   | Short answers                                          |

## The short version

1. Let a test run for at least one full week.
2. Wait for at least 100 results per group and 90% confidence or more.
3. Judge price and shipping tests by profit per visitor, and content tests by revenue per visitor.
4. Check the winner on mobile and desktop before you roll it out.
