Multi-touch campaigns enable you to test changes across multiple touchpoints (such as multi-page or funnel testing). Unlike regular A/B tests, in multi-touch campaigns, each variation can include changes on different pages on your site. This enables you to:
- Test multiple changes on the site against a control group. For example, test the effectiveness of adding a welcome banner to the homepage and an overlay welcoming new users against not adding either.
- Test across different Dynamic Yield campaign types. For example, test whether a prominent overlay is better than a subtle notification for capturing emails.
Note: Multi-touch campaigns are designed only for A/B test traffic allocation and cannot use the Dynamic Allocation mode.
Creating multi-touch campaigns
- Go to Web Personalization › New Campaign and select Multi-Touch.
- Enter a name, and add notes and labels (optional).
- Specify target users in the targeting section, your primary metric, experience settings, and advanced settings.
- Create a variation by defining one or more touchpoints. Each touchpoint is defined by:
- Settings: The settings that apply to the touchpoint type, such as location on the screen. These settings might differ from standard campaigns. For example, the list of triggers for Messaging touchpoints contains fewer options than standard message campaigns.
- Template: The available templates are determined by touchpoint type.
- Design: Adjust the template look and feel, text, and other variables.
- Add touchpoints and variations. Each variation can include up to 20 touchpoints.
- Configure the traffic allocation of each variation and the control group.
- Save & Publish to go live.
Multi-touch campaign reports
Multi-touch campaigns can be analyzed like any other experience by clicking the report icon in the campaign list. The variation performance table can be broken down by audience, and you can view how the test affected additional metrics. Learn more about analyzing reports in Experience Reports.
Multi-touch campaigns include an additional section called Touchpoint Performance. Each variation has a table displaying the performance of each touchpoint. This area is designed to compare the results for each variation and is not ideal for comparing individual touchpoints due to the expected discrepancies between the number of exposed users. For details about why this happens, see this article's Frequently Asked Questions section.
Multi-touch advanced settings
- Fire a Google Analytics event (legacy): Only visible if the Google Analytics integration is enabled. You can disable this option to prevent reporting data about a specific recommendation to Google Analytics.
- Track campaign performance in your analytics platform: Fires events for any analytics extensions and custom code you have implemented. If you disable this option, it won't fire events to any of your analytics implementations.
- Serve on Every SPA Event: Relevant for single-page applications that don’t generate a browser refresh with every screen change. If you need the recommendation to relaunch or reevaluate its targeting conditions upon every screen change in your SPA, fire a track_pageview event and enable this option. For more information on working with SPAs, see Dynamic Yield for Single-Page Applications.
Frequently Asked Questions
What happens if one of the touchpoints replaces the same element that another (regular) dynamic content campaign replaces? What is the priority between multi-touch campaigns and other campaign types?
Multi-touch campaigns have a higher priority than other campaign types, such as dynamic content or recommendations. If both are defined on the same element, the touchpoint content is served to the user.
If a touchpoint is defined as an overlay that conflicts with an overlay campaign with the same trigger, will the touchpoint be served and the overlay campaign not be served?
If there is a conflict between a multi-touch campaign and a notifications, custom code, or visual edit campaign, both campaigns are served.
In the reports, why does the number of users for each touchpoint not add up to the number of users in the variation?
The number of users in the variation table means users who were assigned to the variation, and in the touchpoints table, it's users who were actually served with a touchpoint.
A user is assigned a variation when they meet the "Who" targeting conditions of the campaign AND the touchpoint targeting of at least one of the touchpoints. The user is then assigned a variation according to the variation's allocation. If not all the variations have touchpoints on the same pages (say, variation A has a touchpoint on the homepage and variation B doesn't), a user can enter a page that is targeted by one of the touchpoints and get assigned a variation that doesn't have any touchpoints on that page. This leads to counting the user in the variation, but not for any of its touchpoints.
What attribution window options are supported for multi-touch campaigns?
Currently, only "variation is served" is supported, which starts as soon a user visits a page that at least one of the campaign touchpoints is targeted to, while meeting the targeting condition of the campaign.