How Adobe Target works

Learn how Adobe Target works, including details on the JavaScript libraries (Adobe Experience Platform Web SDK and at.js). This article also covers the various activity types that you can create, Target usage-counting strategies, the Target Edge Network, SEO, and bot detection.

Key points include:

Adobe Target JavaScript libraries libraries

Target integrates with websites using the Experience Platform Web SDK or at.js:

Reference the Experience Platform Web SDK or at.js on every page of your site. For example, add one of these libraries to your global header. Alternatively, use tags in Adobe Experience Platform to implement Target.

The following resources contain detailed information to help you implement the Experience Platform Web SDK or at.js:

Each time a visitor requests a page optimized for Target, a real-time request is sent to the targeting system to determine the content to serve. This request is made and fulfilled every time a page loads, governed by marketer-controlled activities and experiences. Content is targeted to individual site visitors, maximizing response rates, acquisition rates, and revenue. Personalized content helps ensure that visitors respond, interact, or make purchases.

In Target, each element on the page is part of a single experience, which can include multiple elements.

The displayed content depends on the type of activity that you create:

A/B Test

In a basic A/B test, content is randomly chosen from the assigned experiences. You can set traffic allocation percentages for each experience. Initially, traffic may be unevenly distributed due to random splitting, but it equalizes as traffic increases. For example, with two experiences, the starting experience is chosen randomly. Low traffic can skew visitor percentages toward one experience, but this situation balances out with more traffic.

Specify percentage targets for each experience. A random number is generated to select the experience to display. While the resulting percentages might not exactly match the targets, higher traffic leads to a closer split to the target goals.

See Create an A/B Test for more information.

Auto-Allocate

Auto-Allocate identifies the winning experience among two or more options. It then automatically reallocates more traffic to the winner, increasing conversions as the test continues to run and learn.

See Auto-Allocate for more information.

Auto-Target (AT)

Auto-Target leverages advanced machine learning to choose from multiple high-performing, marketer-defined experiences. Auto-Target delivers the most tailored experience to each visitor based on individual customer profiles and the behavior of previous visitors with similar profiles. Use Auto-Target to personalize content and drive conversions.

See Auto-Target for more information.

Automated Personalization (AP)

Automated Personalization (AP) combines offers or messages and uses advanced machine learning to match different variations to each visitor. AP personalizes content based on individual customer profiles to drive lift.

See Automated Personalization for more information.

Experience Targeting (XT)

Experience Targeting (XT) delivers content to specific audiences based on marketer-defined rules and criteria. Including geotargeting, XT is valuable for defining rules that target specific experiences or content to particular audiences. Multiple rules can be set in an activity to deliver different content variations to different audiences. When visitors view your site, XT evaluates them to determine if they meet the criteria. If they qualify, they enter the activity and see the experience designed for them. You can create experiences for multiple audiences within a single activity.

See Experience Targeting for more information.

Multivariate Test (MVT)

Multivariate Testing (MVT) compares combinations of offers in page elements to determine which combination performs best for a specific audience. MVT helps identify which element most impacts the activity’s success.

See Multivariate Test for more information.

Recommendations

Recommendations activities automatically display products or content that might interest customers based on their previous activity or other algorithms. Recommendations help direct customers to relevant items that they might not otherwise discover.

See Recommendations for more information.

How Target counts server-call usage usage

Target counts only server calls that provide value to customers. The following table shows how Target counts endpoints, single mbox, batch mbox calls, execute, prefetch, and notification calls.

The following information helps you understand the counting strategy used for Target server calls, as shown in the table below:

The edge network concept_0AE2ED8E9DE64288A8B30FCBF1040934

An ‘Edge’ is a geographically distributed serving architecture that ensures optimal response times for visitors requesting content, regardless of their location.

To improve response times, Target Edges host only activity logic, cached profiles, and offer information.

Activity and content databases, Analytics data, APIs, and marketer user interfaces are housed in Adobe Central Clusters. Updates are sent to the Target Edges, which are automatically synced with the Central Clusters to continually update cached activity data. All 1:1 modeling is also stored on each edge, allowing complex requests to be processed locally.

Each Edge Cluster contains all necessary information to respond to visitor content requests and track analytics data. Visitor requests are routed to the nearest Edge Cluster.

For more information, see the Adobe Target Security Overview white paper.

Target is hosted on Adobe-owned and Adobe-leased data centers worldwide.

Central Cluster locations house both data collection and data processing centers. Edge Cluster locations contain only data collection centers. Each report suite is assigned to a specific data processing center.

Customer site activity data is collected by the nearest of seven Edge Clusters. This data is then directed to a pre-determined Central Cluster destination (Oregon, Dublin, or Singapore) for processing. Visitor profile data is stored on the Edge Cluster closest to the site visitor. Edge Cluster locations include the Central Cluster locations, as well as Virginia, Mumbai, Sydney, and Tokyo.

Instead of processing all targeting requests from a single location, requests are handled by the Edge Cluster nearest to the visitor. This approach mitigates the impact of network and Internet travel time.

Target Central Clusters, hosted on Amazon Web Services (AWS), include:

Target Edge Clusters, hosted on AWS, include:

The Target Recommendations service is hosted in an Adobe data center in Oregon.

You can allowlist Target Edge Clusters, if desired. For more information, see allowlist Target edge nodes.

Protected user experience concept_40A5E781D90A41E4955F80EA9E5F8F96

Adobe ensures the availability and performance of its targeting infrastructure is as reliable as possible. However, communication breakdowns between a visitor’s browser and Adobe servers can interrupt content delivery.

To safeguard against service interruptions and connectivity issues, all locations are set up to include default content (defined by the client). This default content is displayed if the visitor’s browser cannot connect to Target.

No changes are made to the page if the visitor’s browser cannot connect within a defined timeout period (default: 15 seconds). If this timeout threshold is reached, default location content is displayed.

Adobe protects the user experience by optimizing and safeguarding performance.

Search Engine Optimization (SEO) friendly testing concept_C0C865663CAB4251B66A1F250FD25E6A

Adobe Target aligns with search engine guidelines for testing. Google encourages user testing and states that A/B and Multivariate Testing do not harm organic search engine rankings if certain guidelines are followed.

Adobe Target aligns with search engine guidelines for testing.

For more information, see the following Google resources:

Guidelines were presented in a Google Webmaster Central Blog post. Although the post dates back to 2012, it remains Google’s most recent statement on the matter, and the guidelines are still relevant.

Google states that following these guidelines “should result in your tests having little or no impact on your site in search results.”

In addition to these guidelines, Google also provides one more guideline in the documentation to their Content Experiments tool:

Google states as an example that “if a site’s original page is loaded with keywords that don’t relate to the combinations being shown to users, we may remove that site from our index.”

Adobe feels that it would be difficult to unintentionally change the meaning of the original content within test variations. However, Adobe recommends being aware of the keyword themes on a page and maintaining those themes. Changes to page content, especially adding or deleting relevant keywords, can result in ranking changes for the URL in organic search. Adobe recommends that you engage with your SEO partner as part of your testing protocol.

Bots bots

Target uses the DeviceAtlas metric “isRobot” to detect known bots based on the User Agent String passed in the Request Header.

Traffic identified as bot-generated is still served content. Bots are treated like regular users to ensure that Target aligns with SEO guidelines. However, bot traffic can skew A/B tests or personalization algorithms if treated like normal users. Therefore, known bot traffic in your Target activity is treated differently. Removing bot traffic provides a more accurate measurement of user activity.

For known bot traffic, Target does not:

For known bot traffic, when using Analytics for Target (A4T), Target does not:

For known bot traffic when using client_side logging, Target does not return: