Personalize onsite experiences for unknown visitors using partner-aided visitor recognition
Learn how to use partner-aided recognition to deliver personalized experiences to your web property visitors. Use this tutorial to understand the implementation sequence of various elements in Experience Platform and other Experience Cloud solutions to display a personalized experience to authenticated and unauthenticated visitors.
Why consider this use case why-this-use-case
Fragmentation of digital experiences as consumers interact with brands in myriad ways is very real and is becoming increasingly harder to solve for. The best customer engagement strategies for cohesive experiences, targeted recommendations, and tailor-made interactions are all constrained by user recognition.
This is where partner-aided real-time recognition can make a meaningful difference. ÃÛ¶¹ÊÓƵ allows identity partners to plug into our sophisticated client-side data collection and market-leading experience optimization offerings, to effectively raise the bar on experience delivery from the first visit onwards, without prior history or authentication.
This is especially valuable to verticals that have low authentication rates, like Consumer Packaged Goods, on-line retail, and more.
Industry example industry-example
As an example, consider a home improvement brand that has low authentication rates. This brand would like to deliver personalized experiences to first time visitors, without prior history or authentication, and without the fading reliance on third-party cookies.
This brand chooses to leverage partner recognition technology to probabilistically recognize the visitor and serve a more personalized experience. This helps advance consideration, as the visitor moves down the marketing funnel. For instance, the brand might use partner-provided demographic signals for on-site content that appeals to people who have moved recently and offer a discount on popular DIY products.
Prerequisites and planning prerequisites-and-planning
When planning to use partner-provided attributes to deliver personalized experiences to your authenticated and unauthenticated visitors, consider the following prerequisites in your planning process:
- What inputs are expected by your partner’s recognition technology so they can layer on additional attributes?
- To what extent are you comfortable delivering personalization in different channels and for different use cases based on probabilistically derived datasets, versus deterministically confirmed attributes?
- How should the experience for a pre-authenticated but recognized visitor change when they authenticate?
UI functionality, Platform components, and Experience Cloud products that you will use ui-functionality-and-elements
To successfully implement this use case, you must use multiple areas of Real-Time Customer Data Platform and other Experience Cloud solutions. Make sure that you have the necessary attribute-based access control permissions for all these areas, or ask your system administrator to grant you the necessary permissions.
-
Data Collection
-
Data Management in Real-Time CDP
-
Web property personalization
- Edge segmentation
- Edge Personalization destinations
- ÃÛ¶¹ÊÓƵ Target (or a personalization platform of your choice. This use case tutorial highlights ÃÛ¶¹ÊÓƵ Target as personalization engine)
Video walkthrough video-walkthrough
View the video tutorial below for a walkthrough of how personalize onsite experiences for unknown visitors:
How to achieve the use case: high-level overview achieve-the-use-case-high-level
- As a customer, you license from the data partner the ability to fetch insights in real-time on otherwise anonymous website visitors.
- As a customer, you deploy client-side libraries on your properties to call partner APIs and you configure Web SDK or Mobile SDK to send partner-provided signals to Real-Time CDP.
- When browsing your website or app, the visitor is probabilistically recognized by the partner, who returns attributes along with an ID.
- Real-Time CDP runs edge segmentation to evaluate incoming event hits and persists results against the ECID identifier.
- ÃÛ¶¹ÊÓƵ Target uses edge segmentation output to render the experience back to the visitor for in-session personalization.
- The event is persisted in its entirety for downstream workflows like analysis and retargeting.
How to achieve the use case: Step-by-step instructions step-by-step-instructions
Read through the sections below which include links to further documentation, to complete each of the steps in the high-level overview above.
Data management - Create an identity namespace, schema, and dataset to manage data attributes data-management
In preparation for achieving the use case to personalize unauthenticated visitors’ experience, you must first set up the data management structure in Real-Time CDP to receive the incoming real-time event and audience qualification data.
Create Partner ID identity namespace
First, you need to create a partner ID identity namespace. Navigate to Customer > Identities in the left rail and then select Create identity namespace in the upper right corner of the screen.
Read more about how to create a partner ID identity namespace.
Create a schema
Next, create an Experience Event schema to hold the time-series data that you will later be collecting from your web properties, and make sure to use XDM ExperienceEvent as the base class for the schema. Read about how to create a schema using the Experience Platform UI.
As you create your schema and add field groups to to your schema, consider adding the Visit Web Page and Identity Map field groups. This is in addition to other field groups which are applicable to your digital property and data collection practices.
Additionally, you can create or re-use an existing field group and add it to your schema, to capture partner-provided insights about the visitor. Read how to create a field group and how to add fields to the field group. For instance, if you are expecting to personalize against partner-provided insights like age range, employment status, monthly spending power, or buying behaviors, have your field group include appropriate fields.
Assuming that the data partner provides a stable identifier for the visitor and you would like to bring that into Real-Time CDP, be sure to have an appropriately named field for the identifier in your custom field group. You should also mark the field as an identity in the identity namespace you created earlier. Remember also to enable the schema to be included in Profile.
Create a dataset
Next, you must create a dataset to hold the time-series data that you collect from your web property visitors and that you will use for onsite personalization.
Read the tutorial on how to create a dataset and remember to select the option to create the dataset from a schema. Create the dataset based on the schema that you created in the previous step.
Similar to the step when creating a schema, you need to enable the dataset to be included in the Real-Time Customer Profile. For more information about enabling the dataset for use in Real-Time Customer Profile, read the create schema tutorial.
Implement event data collection on your web property implement-data-collection
After setting up your data management configuration, you now need to implement real-time event data collection on your web property. You need to instrument your property with the ÃÛ¶¹ÊÓƵ data collection library - Web SDK - to collect real-time event calls and send them back to Real-Time CDP. This item consists of a few separate tasks across a few data collection components.
First, use the application switcher in the upper right corner of the screen to navigate to the Data Collection section.
The Data Collection section of the UI looks similar to the image below.
Create datastream
As a first step in the data collection section, create a new datastream. The datastream is the foundation of how data is collected and correctly routed to the right ÃÛ¶¹ÊÓƵ app, in this case Experience Platform.
As you create the datastream, in the Event schema field, select the schema that you created previously.
Select the event dataset that you created earlier from the dropdown, check the boxes next to Edge Segmentation and Personalization Destinations, and select Save.
Note that you do not have to select a profile dataset in this scenario since you are bringing in event-based time-series data.
Create tag property
Think of a property as a container that you fill with extensions, rules, data elements, and libraries as you deploy tags to your site.
Navigate to Tags and select New property.
Fill in the required fields and select Save.
Get complete information about how to create a tag property.
Next, you must install various extensions within the property. Select your tag property and navigate to the Extensions section.
Notice that the Core extension is already installed. You must install two further extensions, as detailed in the next sections.
Install Web SDK extension
Note that this tutorial indicates how you can instrument your website with Web SDK. You can also use on your app to personalize the experience to your app visitors.
In the screen to configure Web SDK, navigate down to the Datastreams section and provide information on the Experience Platform sandbox that you are using. Select the appropriate sandbox and the datastream created in the previous steps from the next dropdown. You can choose the same sandbox and datastream values for all other environments. Leave the other settings unchanged and select Save.
Get complete information on how to install Web SDK.
Install ID Service extension
Use the Experience Cloud ID Service extension to create a unique device-based first-party identity for visitors across all Experience Cloud solutions. Search for ID Service in the extension catalog, and install it. Keep all the default settings when installing the extension.
Set up environments
Next, head on over to the Environments section from the left-hand navigation. In this step, you must connect your website to the ÃÛ¶¹ÊÓƵ Edge Network to retrieve and deliver visitor information in real-time.
Select the box icon on the right for the development environment, and copy the standard version of the JavaScript code snippet that appears in a modal window.
You must add this code snippet to the very top of your website. As a result, your website will make a call to the ÃÛ¶¹ÊÓƵ Edge Network to retrieve JavaScript logic that will be loaded and executed on the page. This allows for functionality like visitor ID generation, data collection, and real-time experience personalization to work.
Set up data elements
Data elements are the building blocks for your data dictionary (or data map). A single data element is a variable whose value can be mapped to query strings, URLs, cookie values, JavaScript variables, and so on. Read more about data elements.
For the purpose of this use case, you must set up two data elements.
First, set up a partnerData
element. Navigate to the Data Elements section and select Create New Data Element.
Name the data element partnerData
, leave the extension value as Core, and set Data Element Type as JavaScript Variable. Enter partnerData
in the field titled JavaScript variable name and select Save.
To set up the second data element, name the new variable pageVisit
, set the Extension to ÃÛ¶¹ÊÓƵ Experience Platform and choose XDM Object as the data type.
From the schema, select the third-party attributes that correspond to the values that you are expecting from the data partner. Then, select the radio button titled Provide entire object. Select the icon that looks like a database and choose the partnerData
data element that you created previously.
Set up rules
In the Rules section, you can configure your website to send a personalization request to ÃÛ¶¹ÊÓƵ with the attributes loaded into the data elements that you just created. Read more about creating rules.
Select Create new Rule. Name this rule Personalize and select the + sign next to Events. Select Page Bottom as the event and save.
Select the + sign next to Actions. Update the extension to ÃÛ¶¹ÊÓƵ Experience Platform Web SDK and set Action Type to Send event.
From the Type dropdown selector on the right, select web.webpagedetails.pageViews
as the event type.
Select the database icon next to XDM data and select the pageVisit
data element.
Scroll down the list of Action configurations and be sure to check the box titled Render visual personalization decisions. This is important to allow experiences delivered via ÃÛ¶¹ÊÓƵ Target or other similar products to be rendered visually on the web page. Select Keep Changes, and then Save the rule.
Set up publishing workflow
To deploy this configuration to the webpage, the next step is to build a library which includes the resources that you just created. Read more about setting up a publishing flow.
Select Publishing Flow and then Add Library.
Select Add all Changed Resources, give the library a name, set the environment to Development and select Save & Build to Development.
Test your website
At this point, your website should be fully instrumented with Web SDK. To test that data collection is working as expected, you can navigate to your website and use your browser’s developer tools to inspect network traffic.
Input interact
in the search box, refresh the page, and you should see network calls from your website to the ÃÛ¶¹ÊÓƵ Edge network populating.
Personalization personalization
You are now ready to create and activate audiences for personalization.
Create audience and set up edge segmentation
In the Platform UI, navigate to Customer > Audiences and create an audience to capture your website visitors.
You must set up your audience with edge segmentation so the audience membership of your visitors is evaluated in real-time, as they visit your web property.
Make sure to also set up an active-on-edge merge policy for the edge audiences.
Integrate with ÃÛ¶¹ÊÓƵ Target or other custom personalization destination
You are now ready to integrate with a personalization engine to display personalized content to your website or app visitors. ÃÛ¶¹ÊÓƵ recommends using the ÃÛ¶¹ÊÓƵ Target destination for this purpose.
Limitations and troubleshooting limitations-and-troubleshooting
Note the following limitations as you explore the use case described on this page:
- If you use Partner IDs, be aware that these IDs are not used when building your identity graph.
Other use cases achieved through partner data support other-use-cases
Explore further use cases enabled through partner data support in Real-Time CDP:
- Supplement first-party profiles with attributes from trusted data partners to improve your data foundation and gain new insights into your customer base and gain better audience optimization.
- Use third-party data support in Real-Time CDP to expand your profile base with prospect profiles from data partners and engage with them to acquire or reach new customers.
- Expanded activation of prospect profiles and prospect audiences to select destinations.