Adobe Parses Target Audiences to Promote Brands

Parsing the segments of target audiences has become a time sink for retails and other data analysts looking to compare and contrast those segments to glean insights, spot overlaps and pinpoint meaningful differences among audience segments for particular brands.

Adobe Systems (NASDAQ: ADBE) analytics unit said this week is leveraging machine learning to automate that time-consuming task by adding a segment comparison tool to its Analysis Workspace platform. The company said the framework automates the process of discovering the subtle differences between a brand’s target audience segments, including buyers and fence-sitters, while uncovering overlaps. It also compares individual data points to discover differences overlooked by overwhelmed data scientists.

The audience segment comparison tool is the latest in a growing list of automation tools that leverage machine learning as a way to cope with huge datasets.

Audience segmenting is considered a critical strategy for marketers, allowing them to use different marketing tactics for specific groups and demographics. With the rise of data analytics, segmentation has extended beyond age, gender and geography. “With the massive amount of customer and behavior data at our disposal, it’s even more important to identify the key characteristics of audience segments that are most significant to a brand—to better understand the behavior that drives more positive interaction, sharing and conversions among different groups of customers,” noted Trevor Paulsen of Adobe Analytics.

The segment comparison capability released in June is the first in a series of audience analysis and discovery tools within its Segment IQ platform rolled out at a company event earlier this year. As with other automation tools leveraging machine learning, the audience parser is intended to winnow the “cascades of data” to find overlaps and differences in target audiences, then compare those differences to improve targeted marketing campaigns.

Adobe is promoting the tools as simplifying audience analysis by “examining any two analytics segments across all of your dimensions and metrics to automatically discover their most statistically significant differences.”

A demonstration available here divides audiences between those who did and did not purchase a product. The analytics tool seeks to determine the key factors that led to a purchase. “Purchasers” are compared with “everyone else” to determine the distinguishing characteristics between the two audience segments. Purchasers also could be compared to frequent visitor’s to a retailer’s website, for example.

The tool runs a “high optimized data science process” to compare every available metric, dimension (for example, which search engine was used to find a retailer’s site) and audience segment. The output is a visualization of online purchasers compared to frequent website visitors and the resulting overlap between the two audience segments.

Adobe said the key product of its segment comparison tool is “the most statistically significant metrics that distinguish purchasers from frequent visitors.” According to the demonstration, the main difference was that purchasers used mobile devices to buy products.

The company also asserts that segment overlaps discovered by its audience analysis tool can be used to “find the key correlations and patterns within your customer base.” It also stresses that the time needed to conduct a segment analysis can be reduced from days to a several minutes.

Recent items:

How a Top Ad Firm Stopped Fearing Big Data and Learned to Love Analytics

New Relic Unleashes the NERD to Analyze Raw Performance Data