Adobe Analytics Earns a Perfect Score for Predictive Analytics

Our vision at Adobe is to help make the enterprise smarter. To this end Adobe has invested and continues to invest heavily in providing world-class predictive analytics capabilities designed for the non-quant, empowering marketers and analysts with powerful data mining and machine learning capabilities.

Forrester Research explains that a Web analytics technology with a perfect score must provide, by default, “ predictive analytics within defined product campaign use cases and strategies.”

And here’s where I get to crow a little. The latest Forrester Wave: Web Analytics Q2, 2014 report provided Adobe with a perfect score in the area of predictive analytics. Forrester Research reports, “Adobe has set high expectations with its vision of making sophisticated analytics available and valuable to all marketers.” This vision is something we are accomplishing in a big way.

A Smarter Enterprise through Predictive Marketing

While executives push their businesses to reach higher levels of sophistication and analytical maturity, few organizations have been able to move beyond descriptive and simple diagnostic analytics into the world of predictive and prescriptive analytics. Adobe Analytics is ultra-focused on augmenting the skill set of the digital analyst to do the work of a specialized quant, or data scientist.

Adobe Analytics offers a robust portfolio of advanced statistics, data mining, predictive modeling, and machine learning algorithms designed to answer the primary questions and challenges of marketing. Rather than spending weeks to months manually prepping data for a predictive model that ultimately is delivered to the business in the form of a PowerPoint presentation, Adobe Analytics provides the insights in minutes in visually rich experiences. It then ties the output to clear audience activation and critical marketing execution paths across the Adobe Marketing Cloud.

Adobe’s current advanced analytics offering includes the following (but are not limited to):

• Library of statistical formulas/functions for summary statistics and inference
• Anomaly detection and time-series forecasting models
• Data exploration capabilities (i.e., correlation matrices, multivariate regression models, and what-if scenario simulation)
• Audience discovery across 10, 20, 50, or more features (i.e., clustering)
• Audience likelihood scoring (logistic regression, etc.)
• Audience classification for target rules and activation (i.e., decision trees)

These predictive analytics capabilities can be applied to an ever-expanding number of workflow scenarios. This means having the ability to produce customer-value based models used in determining which customers to engage for a particular campaign and simulations of campaign “lift” by targeting highest-probability customers.

Prediction and classification model outputs can be saved immediately for broader analysis, ongoing reporting, and cohort analysis. Additionally, Adobe Analytics is integrated with R and supports model export via the Predictive Model Markup Language (PMML) so you can activate audiences through any channel, both online and offline.

Combined with our other investments in better data, more intuitive experiences, and stronger integrations, Adobe Analytics provides the most comprehensive Web analytics solution in the market. And that is something to crow about.