LifeData Analytics

There is a wealth of insightful data generated from real life actions and activities that take place outside the interaction between an enterprise and its customer base. Versium specializes in this data and building analytics on top of it.

What makes Versium’s Analytics Platform unique is its ability to merge huge volumes of real life data with enterprise data to make the predictive output far more intelligent.

LifeData Analytics

 

How Real Life Predictive Analytics Works:

Step 1: Real Life Insights and computer model “training”

Real Life Insight
Enterprise partner supplies a historical data set, which includes categorized data for training the analytics engine. This information and the various segments it includes is then enhanced with real life data providing a deeper and more personal level of customer insight.

Proprietary enterprise data can be of any categorical type or segment level including data such as:

    • customer value levels
    • transaction activity and levels
    • response rates and metrics
    • interests, and preferences
    • spend levels
    • consumptions levels
    • timeliness of paying bills
    • website visit data

Pattern and Signal Identification
The output from step one is statistical summary of real life attributes and combinations of attributes (signals) that correlate to the different segments. At the same time the predictive model is now trained and ready for an un-categorized set of data (i.e. prospect list etc) to be scored.

Step 2: Predictive Analytics, Scoring and Intelligence

Enterprise partner inputs un-categorized data such as a prospect list or any data set where analytics is desired to be applied. This data set is first enhanced with real life data and insights. The analytics engine then provides predictive scoring of where each record in the un-categorized list will fall within the various predefined segment levels and categories established from the initial training set.

Case examples

Digital Offer Optimization

Desired Analytic Output: Predicted best offer to be presented to consumers from 250 offer types based upon previous actions of consumer with like real life attributes.

Data training set: Historical consumer transaction activity.

Real Life Insights: Statistical data output of what consumer with particular attributes and signal types prefer based upon computer modeling of real life data vs preferences identified through traditional recommendation engines that are based solely on transaction data.

Result: 70% increase in user engagement with analytics driven suggestions

Prospect Scoring

Desired Analytic Output: Score enterprise prospect list based on likelihood to convert and customer value level anticipated so that conversion marketing activities can be optimized.

Data Training Set: Historical prospect list with conversion statistics and final customer value level.

Real Life Insight: Real Lifes attributes and signals associated with the different prospect conversion levels and customer value levels.

Result: Value identification of prospects and optimized marketing resource allocation for conversion relative to predicted customer value. Increased ROI on marketing spend.

Fundraising

Desired Analytic Output: Score likelihood of donation and predicted donation level.

Data Training Set: List of historical donors and donation level

Real Life Insight: Attribute identification of individuals with a propensity to donate at different levels.

Result: Optimized fundraising resource allocation driven by predictive output of donation likelihood and donation level.

API Interface
Versium’s Predictive Analytics Platforms is industry agnostic. It can provide scoring results for any type of data that has been trained on a historical categorized set. ( i.e. energy consumption can be predicted). In cases where proprietary enterprise data contains sensitive or private information, the analytics results can occur behind an encryption interface so no sensitive or private information is exposed.

 

LifeData Analytics