Recommendation Campaigns for Email

You can add personalized recommendations to your email campaigns using Dynamic Yield.

You define the widget layout, the algorithm to use when determining which items to display, as well as other settings such as the number of slots the widget should contain in Dynamic Yield. Users who have been identified can be targeted with personalized algorithms such as user affinity. If a user has not been identified, and a personalized algorithm was selected, the fallback algorithm will be used. 

By enabling an integration with one of the supported ESPs in the Integrations page, Dynamic Yield will provide you with an embed code that is already customized and ready to copy/paste into these ESPs. If you are using a different ESP, you copy the Other embed code option and customize it as required by your ESP. 

When you place the embed code in an email, the email calls Dynamic Yield to render the recommendation in real-time when the email is opened. This works only as long as the email campaign still exists and is active. If the user opens the email again, they will see the same content in their recommendation for 4 days. User interaction with the widget such as CTR is tracked and reports can be viewed in Dynamic Yield.

Prerequisites for Email Recommendation 

• Enable the integration with your ESP. If you are using one of the supported ESPs, enable the integration in the Settings ›  Integration page.
• If you want to use a personalized algorithm such as user affinity, identify users using the following Omnichannel Events: Signup, Newsletter Subscription, Login, and the Identify API.
• If you are not using a standard template, create a custom one or edit one of the existing templates. For details, see Creating Custom Email Recommendation Templates below. Note that the email recommendation template can include variables that can be configured in the widget (e.g. title, colors, etc.).
• Limitation: Email recommendation widgets defined in mobile apps (iOS & Android), currently cannot present direct and assisted revenue. The reports display email opens, clicks and CTR.

Creating Email Recommendation Widgets

1. Go to Email › Recommendations and click Add New.
   
2. Define a name and add notes and labels for internal tracking (optional).
3. Click Select a Template and select an email recommendation template.
4. Enter values for the variables defined in the email recommendations template (e.g. title, color, etc.).
   
5. Choose the algorithm to determine what items are displayed in your recommendation. You can modify the algorithm results with the exclude recently purchased products and shuffle results checkboxes. For details, see Email Recommendation Algorithms below. Users who have been identified can be targeted with personalized algorithms such as user affinity. If a user has not been identified, and a personalized algorithm was selected, the fallback algorithm will be used. 
6. You can add a custom filter rule to pin specific products to slots, or apply include or exclude conditions to make sure specific products are, or are not, included in your recommendation widgets. For more details, see Recommendation Rules.
7. Use the URL parameters to track clicks within your analytics software (optional)
8. Click Save. You can copy one of the embed codes that is displayed now, or get the embed code at any point in the future by clicking the button next to any email recommendation campaign.

Using Email Recommendations in an Email Campaign

1. Go to Email › Recommendations.
2. Click the embed code button next to the email recommendation campaign you want to embed.
3. Select the tab with your Email marketing provider (ESP) and copy the code, or copy the code from the Other tab if you have not enabled an ESP from the Integrations page.
   
   

   If you are using the Other option, replace any of the following texts that exist in the code:
   REPLACE_WITH_EMAIL – The email merge field that represents the email address or external identifier, depending on how you identified your users. It ensures that personalized algorithm or filtering rules targeting specific audiences consider the right person. If you use external identifiers, add the code dy_cuidtype=external immediately after this value. For example: dy_cuid=user234xy&dy_cuidtype=external
   REPLACE_WITH_CAMPAIGN_VERSION – Email merge field that presents the unique Id of the email campaign version (every new send represents a new version). In order to minimize the time it takes to return email recommendations, Dynamic Yield caches recommendation results requested for the same user and campaign. Changing campaign version ensures that cached results are not used and that we re-calculate the recommendations results. Notice that some email clients also cache images, changing the campaign version will not eliminate this.
   REPLACE_WITH_LOCALE – The localization string for users supporting multiple languages. The language code should match what you specified in your product feed. For example:
   "lng:en_EN:name":"White Pants"
   For details, see Multi-Language Support. 

4. 'Test your complete widget by setting up a test email campaign in your ESP and including the embed code in the html of the email. Send the email internally to one or more colleagues and open the email to validate. This step is critical to validate that everything has been set up correctly, that the links and redirects to the live website behave as expected and that the images render properly.

Creating Custom Email Recommendation Templates

You create a custom email recommendation template by creating a template for the items, as well as a template for the overall recommendation widget.

1. Go to Assets › Templates and click Add New.
2. Enter a Template name, and select Email Recommendation Item as the template type. This template will define the look and feel of each individual item in the recommendation widget.
3. Use HTML, CSS or JavaScript to define the recommendation items. Use variables to define what will be pulled into the items from your product feed such as product URL, product image, product name, etc.
4. Save the template.
5. Select Add New again and create a template of type Email Recommendation Widget. This template will define the look and feel of the overall recommendation widget and should include email recommendation items.
6. Specify the email recommendation item template you just created.
7. Define the widget using HTML and in-line CSS.
   
   Use the variable ${item} to indicate the location of the recommendation items in the widget. Double click the variable to specify the type as Email Recommendation Item and to specify the minimum and maximum number of items as well as a tooltip (optional). The actual number of items that will be included in the widget is determined later for each implementation of the template.
   
8. You can add additional variables to your widget depending on your use case.
9. Save the Template. You can now use this template when creating Email Recommendation Widgets.

Email Recommendation Algorithms

You can select any one of the following algorithms to determine which products are displayed in your email recommendation. The algorithms fill all slots in your template with the top results of the algorithm from your data feed.

In some cases, if you create a very narrow rule that filters out most of your feed, it is possible that there will not be enough results to fill the available slots. In this case, the remaining slots will render as white images. Clicking on the white image will redirect the recipient to ‘yoursite.com/404’ page. Your main domain is taken from the account settings, so make sure you enter your site’s main domain in your when you created your site.

Algorithm	Description
Popularity	Scores items based on the weighted sum of all product interactions – such as purchase, add to cart and product view – favoring recent interactions over historical ones. You can expose the user to more products by intelligently reordering the algorithms top results by checking Shuffle Results. This works by requesting more products from the main algorithms than is required. For example, if a widget has 4 slots, this will request the 16 most popular products. All of the results get a weight based on their popularity, and then the a lottery system that takes the weights into account randomly selects 4 items. The results should reflect that items that are almost as popular as the top 4 from the 16 returned results now have an almost equal chance of being displayed, but items that are nowhere near as popular now have a low chance. Without this option, the top four results would always be displayed.
Purchased Together	Recommends products that have been purchased together with the products you enter here. You can specify up to 10 items from your feed. It scores items based on the number of occurrences they have been purchased together in the same transaction but demotes products that are typically purchased with many other items. Consequently, products that are strongly linked to one another rather than having an arbitrary connection to a very popular product.
Similar Products To	Recommends items that are similar to the products you enter here, factoring in item popularity. You can specify up to 10 items from your feed. The algorithm uses categories and keywords provided from the product feed to score similarities between items, assigning higher scores to rare keywords shared between a pair of products and lower scores to more common terms.
Viewed Together	Recommends items that have been viewed in the same session as the items you specify here. You can specify up to 10 items from your feed. It scores items based on the number of occurrences they have been viewed together in the same session but demotes products that are typically viewed with many other items, implying a weaker connection.
User Affinity	Personalized for each individual user and scores items based on derived user preference and product popularity. The algorithm bases its recommendation on the visitor’s interactions with the products (views, add to carts and purchases) and their weighted scores. Then it analyzes the attributes of the products (brand, color, style, category, etc.) and calculate the user’s affinity profile. The strategy works in real time and can detect preference changes over time. When there is not enough data to produce a personalized recommendation, the algorithm will offer a set of fallbacks strategies scaling from the most personalized to the most popular.
Collaborative Filtering	Items viewed, purchased, or added to cart by users who are similar to the current user. This algorithm is only supported for small feeds (less than 50,000 products). This feature is available to select customers upon request to your Customer Success Manager.
Deep Learning	Predicts the next item that a user is most likely to interact with, based on similar engagement patterns of your users. This algorithm is supported by any feed type. This algorithm is not available for media sites. It is available to select customers using ecommerce sites upon request to your Customer Success Manager to be part of our gradual rollout.
Viewed with Recently Viewed	Items which are usually viewed in the same session with the last items viewed by the current user.
Purchased with Recent Purchases	Items which are usually purchased together with items the current user recently purchased.
Purchased with Last Purchase	Items which are usually purchased together with items in the last purchase by the current user.

 

Fallbacks

In case the strategy returns fewer items than number of slots, the system will first ‘loosen’ any constraints imposed by the basic filters. If there are still less results than the required items, the recommendation engine will run a respective fallback algorithm:

• User Affinity → Popularity
• Collaborative Filtering → User Affinity → Popularity
• Similarity → Popularity
• Purchased Together → Similarity → Popularity (except in cart pages, where similarity is skipped)
• Viewed Together → Similarity → Popularity
• Viewed with Recently Viewed → Viewed Together → Similarity → Popularity
• Purchased with Last Purchase → Purchased Together → Similarity → Popularity
• Purchased with Recently Purchased → Purchased Together → Similarity → Popularity
• Deep Learning → User Affinity → Popularity

Notes:

1. The following strategies have no fallback: Recently Viewed, Last Purchase, and Recently Purchased.
2. Fallbacks do not override a product properties filter. For example, if you excluded products ‘of a different gender’ property than the currently viewed product – we will respect the constraint even when there are not enough products to display in all available slots. However, if there are no items based on the filter, we will remove the last filter and ignore it to prevent the widget from returning empty. 

Email Recommendation Reports

You can view a detailed report based on the timeframe defined in the date picker by clicking Report at the top of the email recommendation list.

 

You can use the date picker to modify the time range of the report.

The graph on the top of the report displays aggregated metrics for all of your recommendation campaigns. 

If you select one or more of the campaigns below, the graph is replaced with individual lines for each selected campaign. 

• Direct Revenue: Revenue generated from products bought after clicking a recommended item and buying that item (or items from its group ID) within a 30 day window.
• Assisted revenue: Revenue generated from products bought after clicking a recommended item and buying any product within the session.
  For more information, see Direct vs. Assisted Revenue. 
• Clicks: Clicks on a recommended product in an email.
• CTR: The number of clicks divided by the number of opens.

The list on the bottom of the report displays detailed information about each of your email recommendation campaigns. 

You can use the column selection button  to display any of the following metrics: 

• RPM: Assisted revenue per thousand email opens.
• Clicks: The number of clicks on a recommended product in the email. 
• Opens: The number of times the emails were opened.
• Direct Revenue: Revenue generated from products bought after clicking a recommended product and buying that product (or products from its group ID) within a 30 day window.
• Assisted revenue: Revenue generated from products bought after clicking a recommended product and buying any product within the session.
• CTR: The number of clicks divided by the number of opens.

 

Improving Recommendation Image Quality

Dynamic Yield allows you to increase the resolution of the images included in your recommendations. This may have a minor effect on the loading time when the email is opened.

Image Resolution Setting

In the <img src=""> attribute of the email recommendations HTML snippet, you can use the dy_zf parameter to request images with a higher resolution.

The default value is 1. To request images with higher resolution, change the value to 2 (dy_zf=2). 

Since this effectively retrieves larger images, the recommendation images within the email would have appeared larger. To prevent this, we add the width attribute to the <img> tag in the HTML snippet, ensuring that the email recommendation design is not affected.

The value for the width attribute in the HTML snippet provided to you is taken from the email recommendation item template used for this widget.

Code example:

<img src="https://em.dynamicyield.com/emop/999999/5964/85616/16223/3/0?dy_ts=1569238360296&
dy_zf=2&dy_cuid=REPLACE_WITH_EMAIL&dy_version=REPLACE_WITH_CAMPAIGN_VERSION
&utm_source=Dynamic%20Yield%20recommendations&utm_medium=email
&utm_term=10-percent-return" width="195">  

Responsive Image Resolution 

You can implement responsive images by providing more than one version of the images and allowing the browser determine which image is the most appropriate for their resolution. 

Specify the images in the srcset attribute as seen in the following example:

// default img source
<img srcset="https://em.dynamicyield.com/emop/999999/5964/85616/16223/3/0?dy_ts=1569238360296
&dy_cuid=REPLACE_WITH_EMAIL&dy_version=REPLACE_WITH_CAMPAIGN_VERSION
&utm_source=Dynamic%20Yield%20recommendations&utm_medium=email&utm_term=10-percent-return, 

// img source for high resolution screens
https://em.dynamicyield.com/emop/999999/5964/85616/16223/3/0?dy_ts=1569238360296&dy_zf=2
&dy_cuid=REPLACE_WITH_EMAIL&dy_version=REPLACE_WITH_CAMPAIGN_VERSION
&utm_source=Dynamic%20Yield%20recommendations&utm_medium=email&utm_term=10-percent-return 2x" 

// fallback img tag where srcset is not supported
src="https://em.dynamicyield.com/emop/999999/5964/85616/16223/3/0?dy_ts=1569238360296&dy_zf=2
&dy_cuid=REPLACE_WITH_EMAIL&dy_version=REPLACE_WITH_CAMPAIGN_VERSION
&utm_source=Dynamic%20Yield%20recommendations&utm_medium=email&utm_term=10-percent-return"
 width="195" />

For email clients that don’t support the srcset attribute, the fallback is to dy_zf=2 in the example above.

Text Quality in Images

If your text is less than 12 pixels, we recommend using a font-weight of 500 or more in your CSS to prevent some letters from appearing too thin. 

 

Frequently Asked Questions

What happens when a recipient returns to an email in which one or more of the items are no longer available?
Since we cache the images of recommendations, when a recipient returns to an email they will still see the image. However, when they click on the item, it will redirect him to the ‘yoursite.com/404’ page. We grab your main domain from the site settings.

What happens if the recipient or email client block external images?
For security reasons, some email clients do in fact block external images until the recipient opts to display external images; this will prevent the recommendations from displaying until the recipient either changes the settings or clicks on “display external images”.

Why am I not seeing any images in some or all of my email recommendations?
Make sure that your SKU's do not contain the string "//", this is not supported and will cause an error when trying to render the email recommendation. 

Are there any out of the box templates for email recommendation campaigns?
There's only one template that is created out of the box in your site's template - it is called Row.