Visual Statistics and Clustering for Unsupervised Segmentation with SAS Customer Intelligence 360
It's no longer a debate that analytically driven decisions are better in deriving measurable impact. The hype behind AI is primarily focused on augmenting decisions, process, natural language processing, and computer vision. The result is a set of trends made up of these features:
- Algorithms producing better analytics and accuracy
- Automation of machine learning aligned with greater productivity
- Embedded analytics, making AI more impactful and consumable
- Human-like interfaces, creating approachability
What happens when AI becomes useful for your brand? It can effectively be renamed from “artificial intelligence” to “analytical integration” while becoming part of any external customer experience that your brand facilitates. But the lack of factors such as these present barriers to AI adoption:
- Talent
- Stakeholder buy-in
- End-to-end solutions
- Data strategy
Transforming hype into reality for AI must focus on data, discovery, and deployment. Taking action enabled by AI-enhanced decisions completes the enviable last mile of embedding analytics into personalization strategies. Given that customer behavior when working with segmentation varies over time, there isn’t one algorithm that rules them all.
Figure 1: Segmentation and algorithms
SAS delivers behavioral customer segmentation for structured, semi-structured and unstructured data enabling users to leverage unsupervised, semi- and supervised algorithms include k-means, k-modes, k-prototypes, hierarchical clustering, Gaussian mixture models, deep clustering, decision trees, neural networks, gradient boosting, and more. Techniques for defining actionable segments are rules-based, conditional, and algorithmic in an user interface supporting no-, low- and high-code users.
Let’s take a closer look at the influence of machine learning in the campaign management process by transitioning to a technology demonstration using SAS 360 Discover and SAS Visual Statistics. Our mission will be to derive actionable target audience segments within SAS 360 Engage: Direct from the insights of a k-means clustering analysis.
Clustering is a method of unsupervised segmentation that puts observations into groups that are suggested by the data. The observations in each cluster tend to be similar in some measurable way, and observations in different clusters tend to be dissimilar. Observations (or customers) are assigned to exactly one cluster. From the clustering analysis, users can generate a cluster ID variable to tag customers for use in campaign management processes.
Learn more about how the SAS platform can be applied for marketing data management here.
