400%+ Increase in E-commerce Revenue for Fast Fashion Retailer
Increase in E-commerce Revenue
Increase in Email Opens
Fast Fashion E-commerce Retailer
Vidora engaged with one of the premier fast fashion retailers who has a significant digital presence. As their customers increasingly migrated online from more traditional retail shopping tendencies, it became imperative for the fast fashion retailer to have a compelling digital presence which took full advantage of the digital medium to customize consumer experiences. Machine learning and business decisioning powered by machine learning became a key component of their strategy.
In this case study, we highlight how Vidora helped our partner maximize revenue and conversions by transforming consumer data into better business decisions. The fast fashion retailer was able to translate large-scale consumer data across millions of online users to business decisions around how to engage users with targeted marketing campaigns. In the process, their E-commerce business increased email revenue from these campaigns by over 400%.
Our partner’s marketing department often relies on promotional emails to encourage transactions and drive additional revenue. These emails are typically sent on a weekly cadence to millions of active online customers. The key goal for this retailer is to promote new products to customers who are most likely to be interested in that category of products. For example, if they introduced a new line of activewear, their goal would be to only find users that would be interested in purchasing items from this new line. After identifying these users, they would target these users with campaigns.
The challenge with promotional emails is the risk of emailing users about products from categories which aren’t relevant to their interests. Sending the wrong promotion to the wrong user will directly impact top-line revenue. In addition, sending irrelevant messages to users risks those users turning off that particular marketing channel. Our partner wanted to leverage their rich 1st-party data assets to find the right users to target across dozens of weekly campaigns and maximize revenue from promotional marketing campaigns.
Data Sources Across Millions of Active Monthly Users
Cortex allowed the fast fashion retailer to leverage multiple data sources. Note that this fast fashion retailer had significant scale including millions of monthly active users and billions of 1st party data points collected on a monthly basis. Data sources used included:
- Onsite and application behaviors (clicks, URLs visited, etc)
- Purchase activity
- Email and marketing engagement
- App signups
- Clothing catalog information
Configuring Cortex and Building Machine Learning Pipelines
Cortex offers a range of different model types to build depending on what types of decisions the business is trying to automate and optimize. For the use case of targeting promotional emails, the fast fashion retailer built future events pipelines to predict which users would transact with which content category within the next few days. Dozens of future events pipelines were built corresponding to the various weekly promotional campaigns. Pipeline creation was easy and accomplished with a few clicks within Cortex. The fast fashion retailer was able to process these millions of users and billions of behavior events easily and on an ongoing basis.
Once the 10+ pipelines were built to predict the likelihood of purchase for each category, the Cortex Next-Best-Action module was used to select the top performing category for each user. The next-best-action for each user was the category which maximized conversions. For instance, if a user had a 80% chance of purchasing Shoes and 56% chance of purchasing shirts, Cortex assigned that user to the Shoes category.
The fast fashion retailer compared the results of using Cortex to their previous approach of targeting users in order to measure the increase in revenues and open rates. The key results included a 400%+ increase in revenues using Cortex and a 60% increase in open rates. This translated to millions in incremental sales for the fast fashion retailer.
The right business decisioning technology, powered by machine learning, can result in a dramatic increase in top-line revenue. This fast fashion retailer was able to leverage their 1st party data assets, in conjunction with Cortex, to power next-best-action campaigns across millions of users and in the process realize large increases in revenues.