As a quick reminder, here is the onboarding video about Email Recommendations:
Email recommendations are dynamic images in your email showing a thumbnail image and other product information, such as product name and price. These images are rendered in real-time when the recipient opens the message, to ensure that your customers always see the most relevant and up-to-date product recommendations.
Adding recommendation to your emails
In the VCE (block-based editor), there are two ways to add product recommendations to your campaigns.
1. Adding a recommendation block to a template
You can add a product recommendation block to your email template and then use this template to create your email campaign.
To add a recommendation Content Block to an email template, proceed as follows:
- Go to Campaigns, Email Campaigns.
- Click Create Campaign (remember to choose Block-Based Editor if you are asked to choose the editor).
- On the Your Templates page, click one of the two options to Create Template, or click the Edit icon to edit an existing campaign.
- Select the Blocks tab and then click Create New Block.
- Click Insert Recommendation to add a product recommendation image to the block.
- Customize the recommendation image, and when you are finished, click Use this. The image now appears in the live preview.
- To have multiple recommendations in your email, just repeat the previous step multiple times.
- Save the Content Block.
When you now create an email from this template, the recommendation block will be available for selection.
If you are tempted to copy the HTML tag that the Email Widget Editor creates, make sure to change the unique
e-editable ID for each tag, otherwise each recommendation image will show the same product.
Using a recommendation block in an email
When you add a recommendation block to an email:
- You do not need to worry about the content of the recommendation block. Predict will supply as many images as you have defined when the email is opened.
- You cannot change the number of images in a recommendation block. If you want a different number of images, you must create a block in the template for that, or use an image block (see below).
- You can still edit the style of the image using the Email Widget Editor.
2. Converting an image into a recommendation in the email
If your template does not have a recommendation block, or you want a different number of images, you can convert any image into recommendation with a few clicks. This way, you can also add recommendations directly to your campaign while you are editing them. Proceed as follows:
- Go to the Content Creation step of the editor.
- Click the Blocks tab and add a block containing an image to the email, or select an existing image in the live preview.
- Hover your mouse over the image to display the block toolbar.
- Click the Email Widget Editor icon.
- Customize the recommendation image, and when you are finished, click Use this.
Please keep in mind that there is no way to add more recommendations to a block than the number of images already there. So if you want to have 3 product recommendations in one block, you have to use a content block with 3 images, and so on.
Use the preview to check that the dimensions of the recommendation fit the dimensions of the image you are replacing.
Only images having the
e-editable attribute can be turned into product recommendations.
Customizing product recommendations
Email Widget Editor
When you add a product recommendation to your template or to your email campaign, the Email Widget Editor will pop up, which allows you to configure the recommendation image. It displays a live preview on the right.
Here, you can select the product catalog fields you want to display in the recommendation and customize their appearance.
In the Recommendations menu, you can change the recommendation logic. We currently offer the following types of recommendations in emails:
- Mail Personal – Delivers personalized product recommendations to email recipients based on their recent browsing activity and purchase history.
- Abandoned Cart – Delivers the last products abandoned in the cart by the email recipient as well as product recommendations related to these.
- Post purchase – Delivers complementary product recommendations to last purchased items.
- Mail Category – Delivers personalized product recommendations from specific categories.
- Recently Viewed – Delivers the exact items last viewed by the visitor and may also use Mail Personal product offerings to provide the required number of products.
- Recently Purchased – Delivers the exact items last purchased by the visitor, and may also use Mail Personal product offerings to make up the required number of products.
If the purchase was made more than 363 days ago, the purchased items will not be displayed. In this case, only the Mail Personal product offerings will be shown.
Any kind of email campaign will benefit from intelligent product recommendations, including welcome emails, ad hoc newsletters, re-engagement campaigns and birthday congratulations. For more information on how to use these, see Email Recommendations - BestPractices and Examples.
- Availability zone – Delivers certain items that are related to a certain zone, and has localization with their specific languages, currencies.
- Link to – Set a link to the selected Product Catalog field from the drop-down list. The localized link field will be referenced in the widget rendering template when displaying a recommender widget for visitors who are looking at, for example the Spanish version of the website. For more information, see field localization.
In the Style section, you can make some basic changes to the way the recommendations look in terms of image dimension, horizontal or vertical layout, alignment and background color.
In the Formatting section, you need to select either one of the currency format presets or apply custom price formatting. In the latter case, you are free to edit the currency symbol and the numeric display of the price. The language field allows you to select the decimal separator by entering the language code.
For detailed information, see table of valid locale IDs.
Currency format: The price pattern you provide tells our recommender engine the following:
- What currency symbol you want to use.
- Where you want to display it (before or after the amount).
- Whether you want to use digit grouping and decimals in the price.
Group the digits in the price: Always use the comma (,) as the thousands separator. Use the full stop (.) as the decimal separator in the price pattern (e.g. #,###.## or ####.##).
In the Language field, you have to enter a Java Locale ID (and not an ISO country or region code), otherwise the country/region-specific separators will not be applied. In the prices shown in the actual product recommendations, the separator symbols specific to the language you set in the Language field will be displayed.
Define the Price format according to your needs:
- If the price pattern ends in .##, no trailing 0 will be displayed (e.g. € 1,234.5) in the recommendations.
- If you want to include the trailing 0s, format the price to end in .00.
- If you want to see a leading 0 in the recommendations, format the price to ###,##0.00 (e.g. € 0.99).
The block to which I want to add the recommendation also has some text under the image. Can I populate this using the recommendation editor?
No, because the product recommendation is itself an image containing several fields rendered into the image itself. For recommendations, you should use blocks that only contain images or images and other optional fields, which you can delete. See the next FAQ.
I have a block that only contains images, but I cannot convert it into a product recommendation. Why is that?
The images may not be editable. Please check if they have the
e-editable attribute in their tag.
The product thumbnail looks blurry. Is it possible to get higher resolution images?
Always ensure that you upload high-resolution images larger than the thumbnail size, so that they can be scaled down with no loss of quality.
Which time zone is used for the Recommendations?
The time zone for the predict last session date is in UTC. The last session date is calculated according to the order of the product viewed.
Which products are in the Recommendations if the contact has not purchased or browsed the items yet?
If a contact has no history the recommendations are populated randomly from a fallback list containing the best trending products.