Technology Provider: Optimove
User Company: Adore Me
Industry: Consumer Retail
Application: Marketing / Advertising
Problem
Adore Me required a more personalized and targeted method for customer communications and marketing. The company required better attribution on marketing spend, and more personalization and automation across multi-channel campaigns.
Actions Taken
Adore Me used Optimove’s segment modeling technology to find an initial set of “personas” (customer segments with distinct qualities or behaviors) to be targeted for specific offers or incentives. Over their first year of collaboration, Adore Me was able to regularly target over 60 unique personas.
Adore Me’s work with Optimove allowed them to expand beyond email marketing (previously, their only mode of customer contact) into push notifications, Facebook Custom Audiences, Google Display Network, and more. Optimove also aided in validating marketing tests for impact on conversion rates and financial impact, states Josselin Petit-Hoang, Adore Me’s Head of CRM.
Results
Adore Me reported a 15% increase in monthly revenue generated by Optimove-driven campaigns, as compared with control groups, in addition to a 22% increase in average order amount. The company credits Optimove with fully automating 85% of customer campaigns.
Transferable Lessons
Customer segmentation and targeting is difficult work, and nearly all companies can make more informed decisions about their targeting and their split-tested offers and communication.
Pattern recognition is one of the most outstanding capabilities of machine learning. In addition to finding patterns in streaming metrics (business intelligence) or stock market data (finance), patterns can be recognized in CRM data that allows for potentially useful “clustering” that might have evaded human perception (particularly at scale).
Year-over-year, marketing continues to shift towards improved personalization (so-called “one-to-one marketing”) and improved attribution. With more marketing moving into digital channels, quantifiable data will likely only speed up this trend of personalization. Particularly for companies with a substantial volume of lead and customer data, split-testing will also become much more mainstream in industries where it has not been the norm.
