Evergage Unveils New Machine-Learning Innovations to Improve the Impact of and Ability to Analyze Personalization Efforts

Nov 01, 2018

Evergage, The 1-to-1 Platform company, announced two new machine-learning innovations to help companies drive greater results with 1-to-1 personalization. With Evergage Decisions and the Evergage Data Science Workbench, companies can tap into their rich customer data, maintained in the Evergage platform, to maximize the performance of their personalization campaigns.

Evergage Decisions

For companies with multiple content assets – such as promotions, messages, and images – it can be challenging to match the ideal experience to a specific visitor in real time. For example, a financial services firm might have a defined area on its homepage to highlight promotions – such as for credit cards, mortgages, auto loans, and 401k plans – and wants to display the optimal one to each visitor, considering the likelihood of engagement and the value to the business.

With the new Evergage Decisions algorithms, B2C and B2B companies can apply industry-leading artificial intelligence (AI) to automatically deliver the most relevant content – or complete experience – to each website visitor, application user, and email recipient.

Complementing Evergage’s already-powerful, machine-learning-driven, 1-to-1 personalization and recommendation capabilitiesContextual Bandit, the first algorithm launched as part of Evergage Decisions, goes further – delivering the most relevant offer or experience with the highest potential value to the company. This is computed by taking into account both the probability of engagement at the user level and the revenue opportunity or synthetic value (for non e-commerce use cases) at the business level 

Factoring in deep behavioral data, including a visitor’s digital engagement and the context of each session, along with other situational and attribute criteria (e.g., referral source, browser, device type, lifetime value, geolocation, etc.), Contextual Bandit: 

  • Estimates the probability of each person interacting with each available offer or experience on a given channel (website, web app, mobile app, email) in real time.
  • Uses advanced machine learning to predict the content for each visitor with the highest-value return – weighing the probability of someone accepting a particular offer or promotion, with the business value of that offer to the company. And unlike A/B testing methodologies, which can only determine the best choice for all visitors, Contextual Bandit delivers automatic personalization to determine the best experience for each individual visitor.
  • Frees up marketers to focus on creating powerful messaging and offers,rather than spending lots of time defining rules about which experience to show which audience every time.

Data Science Workbench

Evergage is also introducing another breakthrough for its customers: the Evergage Data Science Workbench, giving companies’ data scientists a way to access and work with the rich data stored on Evergage servers. Such data – maintained in the unified profile of each and every customer, visitor and account a company has – includes in-depth behavioral data, contextual information, and explicit survey responses, combined with first--and third-party information from other systems (e.g., CRMs, email service providers, data warehouses).

With the Data Science Workbench, this trove of information – previously available only to Evergage’s own data scientists – is now available to Evergage customers’ data scientists as well, in an environment that supports activities such as data transformation, numerical simulation, statistical modeling and data visualization.

The Data Science Workbench is a new component of the recently announced Evergage Gears framework – which enables companies to extend the core capabilities of Evergage’s platform and drive even greater value from their Evergage investments.

While Evergage natively provides companies with a way to deliver 1-to-1 personalized experiences based on deep behavioral analytics data, companies now also have the ability to access this raw data for their own analytical purposes. Their data scientists can use the Data Science Workbench to:

  • Initiate pulls of live data and cache it into the workbench.
  • Investigate the data using familiar tools, including Spark, Python and R.
  • Create visualizations, execute data transformations, build models and more.
  • Enhance the unified profiles in Evergage with the output of such models.