Email with last visited brand

A good way to personalize communication is to offer customers products from a brand they recently viewed. This is an especially good solution for companies offering multiple brands.

You can build this kind personalization in a very easy way and make it possible to send an email campaign with personalized offers from specific brands to people who have viewed this brand recently.

Example of use - Retail Industry

Challenge

One of our clients wanted to send to their customers personalized emails based on the last brand that the client viewed.

This company is an example of a multi-brand company in which you can select products based on brands. In the URL information of every product, they always have information about its brand.

Solution

Based on that we created an automation which lets us monitor the last-viewed product and the category it belongs to. Thanks to this, we were able to send personalized emails with brand-related AI recommendations. Every customer got a different email recommendation based on their previous behavior.

Prerequisites

General

• Tracking code on your website.
• Email account configured.
• AI integration.
• Product feed implemented.

First concept - manual preparation without jinjava**

Create an aggregate

1. Go to Analytics > Aggregates > Create aggregate.
2. As the aggregate type, select Profile.
3. Enter the name of the aggregate.
4. Optionally, to let other users know about the objective of the analysis, write a short description.
5. Click Analyze profiles by and select Last.
6. Choose event page.visit, as a operator choose URL and add a regular expression which indicates that the URL has to contain one of the following brands.
7. Save your aggregate.

Prepare an AI recommendations

1. Go to Communication > Recommendations > Add recommendation.
2. Enter the name of the recommendation (it is only visible on the list of recommendations).
3. In the Type & Items feed section and click Define.
4. From the Items feed dropdown list, select a product feed that has a trained model.
5. Select the recommendation type (you can choose the type of campaign on your own, based on your business needs).
6. Click Save.
7. Prepare the AI recommendations separately for every brand.

Create email campaign

1. Go to Communication > Email campaign > Create new.
2. Create email campaigns for every brand separately, adding the AI recommendation campaign prepared for each brand.
3. Save your campaigns.

Prepare the automation

1. In Synerise, go to Automation > Workflows > New workflow.
2. As the trigger node, select Profile Event. In the node configuration choose the event page.visit which tells us if the client visits a specific website with a specific url, use regular expression – this indicates that the customer visited the products from one of the following brands.
3. Add optional Delay.
4. Add Profile Filter which checks every agreement which customers have (cookie agreements, email agreements and so on).
5. Split Path to profile filters, which checks what the last aggregate of the customer was. Every filter indicates a different brand and sends a different campaign. For example, if the last brand a customer visited was ABC, he will get an email with products from this specific brand. We separated every segment in this place in the automation, which lets us check to which brand the customer is matched and send him personalized mailings.
6. Add node with Email as a result of workflow. In the node configuration choose your personalized email campaigns.
7. Add End node.

Second concept - preparation with jinjava

We helped our client to prepare this process in a smarter way. We changed the part with all profile filters to single profile filters and only one event.

1. Create the aggregate the same as in the first concept.

2. Create the AI campaigns for each brand separately.

3. Create ONE email template with the special jinjava code.
   
   Based on the jinjava code we can create a smarter process for the whole campaign.
   
   We took the aggregate described above and checked the code to see the last brand the customer visited (what phrase customer had in the url).
   
   So if, for example, our customer viewed brand ABC, we added the code the title: ABC, an image and a link to the product. And the same applied to all other brands. So we prepared all such information in one place, in one code.
   
   So if in the aggregate we had the ABC brand – in the email template the ABC brand was added to a specific product along with a link to it.
   
   Also, we add a lot of IF’s in our jinjava code. This indicates that IF ABC brand was the last one the customer visited, then the ID of the AI campaign for the ABC brand will be used in that case.
   
   Using this method, we can change the content of the mailing dynamically depending on what we have in the aggregate. Information about the specific brand will go to the output in the form of dynamic code, which we have to paste in the email campaign in a specific place. Thanks to this it will change dynamically based on the results from the aggregate for a specific customer.
   
   Remember to add the AI campaign to the content of your email campaign based on the campaigns created earlier.
   
   

4. Create an automation with additional jinjava code.

5. In Synerise, go to Automation > Workflows > New workflow.

6. As the trigger node, select Profile Event. In the node configuration choose the event page.visit which tells us if the client visits a specific website with a specific url, use regular expression – this indicates that the customer visited the products from one of the following brands.

7. Add optional Delay.

8. Add Profile Filter which checks every agreement which customers have (cookie agreements, email agreements and so on).

9. Split Path to profile filters, which checks what the last aggregate of the customer was. Every filter indicates a different brand and sends a different campaign. For example, if the last brand a customer visited was ABC, he will get an email with products from this specific brand. We separated every segment in this place in the automation, which lets us check to which brand the customer is matched and send him personalized mailings.

10. Add node Send Email as a result of workflow. Add the jinjava code to the email template instead of adding it in the form of many brand filters, which makes the whole automation very large and more complicated.

11. Add End node.

It is easier to create a second version with jinjava to avoid mistakes and make it faster and easier to build and to manage.

Read more

• Aggregates
• AI campaigns
• Automation