Premium

Integrate Recommendations with email

Adobe Target supports send-time personalization of recommendations in email.

Three methods for integrating Target Recommendations with your Email Service Provider (ESP) are available. Your ESP’s capabilities determine which method to use. Your account manager or consultant can help you choose the option that works best for you.

Using method 1 or method 2 requires your ESP to make calls to an external API on a per-customer/per-email basis and wait for content to be returned. These methods are not supported by all ESPs; contact your ESP to determine if it is compatible with this integration pattern.

Using method 3 requires your ESP to join a list of recommendations by product ID or category ID to your list of emails. This method can be based on an attribute such as the customer’s last viewed product, last purchased product, or most viewed category. However, your ESP must have access to this data in its customer profile in order to perform the join. Contact your ESP to determine if it has access to this data and is compatible with this integration pattern.

Open-time personalization of recommendations is not supported by Adobe Target.

Method 1: Use the delivery API (Preferred)

The delivery API is a POST request that works with build-time email. This option is the preferred method for build-time email.

Most email clients do not allow POST requests. Therefore, this API is not recommended for open-time use cases. Some email clients, such as Gmail or Outlook, can cache the content or block the image and require the recipient to pro-actively allow the image to render.

You cannot return default content using the delivery API.

The following code is a sample API delivery request:

curl -X POST \
  'https://clientcode.tt.omtrdc.net/rest/v1/mbox/?client=clientcode' \
  -H 'authorization: Bearer 3423614b-4843-4664-83c4-c6c3f6c8869b' \
  -H 'cache-control: no-cache' \
  -H 'content-type: application/json' \
  -d '{
  "mbox" : "email-mbox",
  "tntId" : "111499796294071-449025.28_44",
  "requestLocation" : {
    "host" : "prod"
  },
   "profileParameters" : {
   },
  "mboxParameters" : {
    "at_property": "b468a242-64a4-32a0-ca0c-890bddd78789",
    "entity.id": "article-123",
    "entity.event.detailsOnly" : "true"
  }
  "contentAsJson":  true
}'

Where clientcode is your Target client code.

See Delivery API documentation for more information.

Method 2: Use a rawbox email template

A rawbox is similar to an mbox request, but for non-Web environments, such as email service providers (ESPs). Because you don’t have the Adobe Experience Platform Web SDK or at.js to use in rawbox requests, you must create your requests manually. The examples below explain how to work with rawbox requests in email.

This approach allows you to track performance of recommendations in emails, test them in the normal way with a recommendation, and continue tracking on the site.

Set up a Recommendations activity in Target, using the Form-Based Experience Composer option. For the location, select the name of the mbox you’ve chosen to use in the rawbox request coming from the ESP. Select a design with the look and feel you want for your email. At email build time, the ESP makes a call to the Target servers for each rawbox in each email being generated. Your ESP must have a way to include the returned HTML in the email when it is sent.

The email system you use must be able to handle the following scenarios:

A valid response is received, but no recommendations are present

• In this case, the response is whatever is set as the mboxDefault parameter value. See explanation below on this parameter.
• The email provider should have a default HTML block of recommendations to use in this case.

The Target server times out and returns without data

• In this case, the Target server returns the following content:

  //ERROR: application server timeout

• The email application should search for that text and must be able to handle the error. The email provider has multiple options for handling this case:

  • Try another server call immediately (recommended, perhaps with an attempt counter).
  • Toss out that particular email and continue to the next one.
  • Queue that particular email and rerun failed emails as a batch at the end of the initial run.

Sample request URL

https://client_code.tt.omtrdc.net/m2/client_code/ubox/raw?mbox=mbox_name&mboxSession=1396032094853-955654&mboxPC=1396032094853-955654&mboxXDomain=disabled&entity.event.detailsOnly=true&mboxDefault=nocontent&mboxNoRedirect=1&entity.id=2A229&entity.categoryId=5674

Required parameters:

Optional parameters

Potential Target server responses

Method 3: Use the Recommendations Download API

Set up a recommendation as usual, but choose download only in the presentation section instead of a template and mbox combination. Then in the ESP, tell the ESP what recommendation ID you created. The ESP accesses the recommendation data via API. This data shows which items should be recommended for a particular category or key item, such as items in an abandoned cart. The ESP stores this data, connects it with their own look and feel, displays information about each item, and delivers that in the emails.

With this option, the recommendations server cannot directly track the performance of a recommendation or split traffic across multiple algorithm/template combinations. Also, the recommendations are not tied to a visitor profile.

For more information about the download API, see Legacy APIs > Download.