.png)
In 2019, Rusta and Sift Lab Initiated a Collaboration. Rusta had large amounts of transaction data that they felt were not being fully utilized. They used to send out the same campaigns to all customers or based target groups on manual selections. Campaigns were followed up on overall key figures without seeing if it was the marketed products that were sold.
"We had long felt that we could use our data more effectively, but our intentions often led to costly and time-consuming IT projects. With Sift Lab, we quickly gained new insights without IT resources. They provided us a simple and quick way to turn data into insights and commercial value."
Alexander Stenvall
Head of Digital and Customer Insights, Rusta
Rusta hires Sift Lab to investigate the potential in their data
Rusta wanted to see if Sift Lab could help them use their data more effectively and set up a project in three steps:
1. Sift Lab's platform automatically synced Rusta's data via their CRM system Voyado.
2. Various machine learning models analyzed these data in the platform.
3. The results of the analysis were presented.
The analysis clearly showed how Rusta had customers with different purchasing behaviors and preferences: some customers bought certain product categories and not others, some stores sold different types of products, different types of customers had different lifetime values and likelihood of making a next purchase, and so on.
Insights that could be transformed into commercial value, quickly
The knowledge that Rusta had customers with different behaviors was not new, but Sift Lab's platform made it possible to categorize each customer based on their behavior. Sift Lab's integration with Rusta's CRM system Voyado also made it easy to see if the analysis could also create commercial values.
Rusta set up a number of tests where they let the platform select target groups for different campaigns. The target groups were automatically exported to Voyado where Rusta ensured that the email was sent out as usual. The tests showed a 13–62% better conversion compared to manually selected target groups.
And how long did the project take? Within a week, data had been synced, analyzed, and presented so that Rusta could prepare the campaigns they wanted to test. All without involving IT resources.
From Project to Process
The results convinced Rusta of the value of working with machine learning and segmentation to increase conversion on campaigns.
