SPSS
From SmartEnoughWiki
Contents |
[edit] Company: SPSS
SPSS is a well known provider of analytic tools and claims the most predictive analytic customies.They pride themselves on a quick ROI – various independent studies have shown very high ROI scores – and they show up as a leader in most of the analytic tool surveys done.
SPSS regards its core competency as helping customers do more with less - increase revenue (through acquire, grow and retain) and reduce risk and fraud. To do this SPSS’ customers need to understand their customers better, predict their behavior and preferences and then use these results to take more appropriate actions. Some SPSS customers do use the products for supplier and product analysis, but customers are the focus of most. SPSS uses the Capture, Predict, Act, Capture cycle to describe how it does this. Capture focused on data collection and analysis; predict on hypothesis testing, data mining and predictive modeling; act on integration with operational processes.
In particular they are trying to help their customers to get a true 360 view and this means using attitudinal data, interaction data, descriptive data and behavioral data from all touchpoints. They have all the usual predictive analytics – regression models, neural nets, clustering, segmentation, forecasting etc – and deliver them through role based environments. Their latest product, Clementine v12, has lots of automation to make suggestions, find useful characteristics or visualization techniques. There is still plenty of opportunity for a good modeler to make a difference but they are trying to eliminate the repetitive, foundational activities and automate some best practices like using multiple models.
The end result of using SPSS’ tools is to embed analytics in operational systems - for offers in call centers, cross-sell and upsell on websites, direct marketing, assessment at point of risk in claims and so on. Deployment options include database updates, generated code, real time rules, real time scoring of customers and more.They also allow the combination of batch scoring with real-time so that the results of an interaction can be combined with already known information to improve treatment at the point of contact.
SPSS like to adopt a phased approach to deliver success – starting focused in one, high ROI area, and growing with an ROI at each stage. That said, their focus is now helping their customers become Predictive Enterprises by embedding analytics into key business process and using analytics to drive core business decisions. To make analytic decision making a way of life – tom Davenport has called this becoming an analytic competitor but I like the SPSS phrase “predictive enterprise”. Like some other analytic companies they are starting to help their customers connect and combine analytics – don’t sell to people who are high risk for instance.
[edit] Product: Dimensions
Capture tools (survey, questionnaires, online forms) that support translation, multi-channel deployment etc. Also Text analysis tools – sentiment, data mining, Natural Language Processing (both their own and some from Language Weaver) – that are very integrated with data mining and with surveys
[edit] Product: Stats
SPSS stats is primarily used by knowledge workers to gain insight – direct integration with operational systems is not common
[edit] Product: Clementine
Data mining, text analysis tools and text mining with deployment using some of their own rules, integration with Business Rules Management Systems also something recognized as important. In general deployment is important to SPSS both for real-time and mixed batch/real-time scoring. Clementine is more heavily focused on operational systems and deep embedding is an emerging trend growing fast especially in newer customers. SPSS sees lots of text analytics interest also, typically driven by the integration of structured and unstructured analytics.