Display Funnels on a GDS dashboard

1. Connect the dashboard to the data source where the steps data is stored.

If you are using Google Analytics, you need to set up e-commerce or events with Google Tag Manager, or edit your website code by adding JavaScript. You can also use page URLs as steps, and you don’t need to implement any additional tagging. If you are storing step and conversions data on a warehouse like Google BigQuery or any other, you can use the available connectors on GDS.

2. Click on Add a chart at the top menu and select the horizontal bar chart to build the funnel view.

You can also use vertical bar charts but usually, the funnels are represented with horizontal ones.

3. Drag and drop on the Dimensions box, the field that contains the funnel steps names.

The fields would be Event Actions for Events, Shopping Stages for E-commerce, or Page for URL steps.

4. Drag and drop on the Metrics box, the field that represents the volume of actions for each funnel step.

For events this should be Unique Events for e-commerce, or pages you can use Sessions.

5. Add a filter to show only the relevant values for the funnel steps.

This could be certain URL pages, specific events, shopping stages, or checkout stages.

6. Create a new data source by clicking on Resources > Manage Blended Data at the top menu, to calculate the conversion rate for each step.

Each data set will represent one step of the funnel, setting the same field from point 4 as Dimension, and including a filter for that step.

7. Click on Add a Chart > Scorecard and create a calculated field for the conversion rate, for each step of the process, dividing the volume of each step over the first step.

This scorecard should be next to each step of the bar chart you created before.

8. Use the style tab on the menu on the right to adjust your chart design, to match with the rest of your dashboard.

You have different ways to customize your chart. You can uncheck the Show Axes option, set as transparent the grid color, and change the bar colors or show data labels.