Abandoned Carts: How Orcanta Increased Email Marketing Revenue by 126%

Goal

Orcanta’s marketers were looking to achieve the following two business goals:

• Increase revenue from online sales using direct marketing. The company wanted this channel to generate at least 5% of the total online revenue instead of 1.5%. The company’s marketing team decided to bring users back to the website by sending them cart abandonment emails. In order to do this, the team needed to create a segment of users with abandoned shopping carts and contact such users within an hour after they leave the website, with an incentive to complete the purchase.
• Reduce email marketing costs. Before, the company only used the Eperflex email retargeting platform, paying 0.3 euro per click in emails. This is how it works: first, Orcanta sends online user behavior data to Eperflex. Next the data is associated, by the cookies, to email addresses from the platform’s database. In doing so, the company was able to identify the user who initiated the session in 15% of website visits. In order to reduce costs and identify more users, Orcanta’s CRM team decided to send out regular emails using the Actito platform.

Challenge

Orcanta’s marketers use Actito to send cart abandonment emails to users whose email addresses they already know. By adding tracking parameters to all links in emails, the marketing team can associate each session across multiple devices with a user’s ID from the company’s CRM system. At present, Orcanta can identify about 45% of user sessions on their website.

At the same time, Orcanta’s marketers were looking for a solution that would allow them to automatically send CSV files with relevant data about whom to send emails, and what products to include in each email, to the Actito platform. The CSV file should be updated each hour and have the following structure:

The problem is that Google Analytics only allows combining dimensions and metrics if they share the same scope: Hit, Session, User or Product-level. This made it impossible for the company to combine the parameters that have different levels of scope, while Orcanta needed to send all the data to Actito in a single file. For more information about this limitation, see one of our blogposts.

Another problem was that the data becomes available in Google Analytics within 24 hours after a hit is sent from the site, and the company needed to have the data updated each hour, in order to send more timely emails and improve conversion rates.

Solution

In order to combine dimensions of different scopes in one report, and in order to achieve the company’s goals, Orcanta analysts decided to:

1. Use OWOX BI Pipeline to send data about user interactions with the product pages to Google BigQuery, directly from the website.
2. Send the data about email addresses and User IDs to Google BigQuery from the company’s CRM system.
3. Use an SQL query to merge the data about users who abandoned carts, and product parameters in the carts, in a single table.
4. Create a CSV file based on the data from the table, and send the file to the Actito’s FTP server.

The resulting data flow chart looks as follows:

Now, let’s take a closer look at what the company did and what tools they used to increase revenues.

Step 1. Collect website user behavior data in Google BigQuery

Orcanta’s analysts chose to use the Google BigQuery data warehouse for data collection and consolidation. The main reason for choosing Google BigQuery was that, it allows creating complex custom reports without facing any limitations.

With the help of OWOX BI Pipeline the data about each user interaction with the website is sent to Google BigQuery in near real-time. The OWOX BI team helped Orcanta make sure that the data is correct and supplement the data with product image URLs. The values of the URLs are sent to Google BigQuery as custom dimensions and matched with the product IDs in a dedicated dictionary.

In addition, OWOX BI Pipeline enabled the company to more accurately identify users across sessions, thanks to data reprocessing. This means that User ID values can be assigned retroactively. For example, if a User ID was assigned in November to a website visitor who had visited the website earlier in October, the ID value will also be assigned to all previous sessions of this visitor.

Step 2. Send the data about emails and User IDs to Google BigQuery

Email addresses and User IDs of authenticated users are stored in Orcanta’s CRM system. In order to use this data in the Actito platform, the company imports it to Google BigQuery using an in-house solution developed by Ciklum.

Step 3. Merge the data about users and products

The OWOX BI team helped Orcanta create the SQL query that is used to bring together the data about users who abandoned carts and the parameters of products in abandoned carts, all in a single table.

Step 4. Create a file with parameters needed for email marketing

The OWOX BI analysts developed a script that runs the query every hour and updates the data in the table. Orcanta’s analysts use this data to create a CSV file and send it to the Actito’s FTP server. The data from the CSV file is used by the service to automatically send out cart abandonment emails with information about the products and an incentive to complete the purchase.

Results

Thanks to the cooperation with Actito and OWOX BI, Orcanta managed to:

• Reduce email marketing costs.
• Identify 180% more users who initiated the session. Now the company’s marketers only use Eperflex if they weren’t able to identify a user with the company’s internal solution.
• Achieve the 7.5% email-to-sales conversion rate by sending out cart abandonment emails within an hour and and a half after a user leaves the website.
• As a result, emails contributed to 4.5% of the company’s total online sales revenue.

How do you re-engage website visitors? Share your ideas and questions in the comments below.