New A/B Results page Explore tab
The Explore tab lets you explore your experiment data and save the explorations you create. Segment the results by user attribute or behavior, filter to a date range, and narrow the view to specific metrics or variations. View the data as a table or a graph. When an exploration highlights a data point you want to focus on, save it to the explore tab. The exploration is then available the next time you or your colleagues open the results page.
Open the Explore tab
You can open the explore tab in the following ways:
- Click Explore on the experiment results page.
- In the Summary tab toolbar, click Change Date Range, Segment by Cohorts, Group by Property, Filter out data, or click Explore results data drop-down list.
Explore your data
The explore tab is built on Optimizely Analytics explorations, so you analyze experiment results with the same segmentation, filtering, and visualization tools. See Explorations.
The toolbar offers three types of segmentation. Each one opens the same exploration view.
Filter out data
Filter out data lets you segment based on one or more user attributes. You can combine different attribute-value pairs to create the segment that you want. For example, to analyze mobile users in the US, set device to mobile and country to US. Click Run to show results for that segment.
Build each condition with an operator such as is, is not, contains, or in, and combine multiple conditions with and and or groups to narrow the data further. For the full set of operators, see Filtering.
Group by Property
Group by Property breaks the results down by every value of a single attribute. Unlike Filter out data, where you select a value for each attribute, Group by Property needs only the attribute. Optimizely finds every value and breaks the results into segments you can compare. For example, group by country to compare performance across all countries.
For more on attribute-based segmentation, see Segmentation.
Segment by Cohorts
Segment by Cohorts segments by behavior rather than by attribute. Define a user behavior (a single action or a sequence of actions), and Optimizely filters the results to only the users who performed it. This is valuable when you want to analyze users who completed a specific funnel, for example.
Create a cohort that defines what users did, for example, visitors who added an item to the cart, then click Run. The results show only the users who match the cohort. Rename the cohort to describe it, for example, Users who added to cart.
A cohort is a behavioral group defined by conditions: the events users completed, time windows (such as within seven days of an action), and aggregations, combined with AND and OR logic. Build a cohort inline for a single exploration, or reuse a saved cohort. For details, see Cohorts.
You can apply Filter out data, Group by Property, and Segment by Cohorts separately or combine them. Applying them together gives you the flexibility to build the exact analysis you want. For example, break the results into segments by a user attribute with Group by Property. Then exclude specific values of that attribute with Filter out data and the is not operator.
Change the date range
Click Change Date Range to limit the results to a specific period.
Focus and format the exploration
After you segment the data, you can refine and format the exploration:
- Filter to specific variations – Show only the variations you want to compare. For example, show the original and one variation, and hide the others.
- Focus on specific metrics – Remove the metrics you do not need, so the view shows only the metrics you want.
- Choose a representation – The exploration shows a table by default. Click Chart to switch to a visualization, such as a line chart (metric over time or improvement over time) or a single value. For the full set of chart types and display options, see Visualization.
Save explorations
Save an exploration so it is available to everyone who views this experiment's results page. When an exploration shows what you want, click Add to Explore to add it to the explore tab.
When you finish, click Save so the explorations are available the next time you or a colleague opens the explore tab.
Add text and images
Besides explorations, add text and images to build a readout of the experiment. Add text to give context about the analysis or the explorations you built. Add an image to illustrate the variations or other important details, so you build a compelling narrative around the experiment.
Comments
You can add comments on the explore tab to capture context and collaborate with your team:
- Add a comment – Click Comments, enter your notes, then click Send.
- Edit a comment – Click More options > Edit Comment. Make your changes, and click Confirm.
- Delete a comment – Click More options > Delete Comment, then click Confirm.
