8-Sep-09 10:00 AM  EDT  

September 2009 Featured Article: Unica 

One of the most exciting things about analytics is being able to manipulate data and address specific business issues with the functionality available at your fingertips (theoretically) yet often times this isn’t the case in practice. For a variety of reasons, practitioners spend a great deal of time engaged in the manual and technical tasks of set-up and validation to answer a historical reporting question asked for by the business days ago. How is that us analysts can break through both of those barrier?

1. Driving down the latency from a business person’s request to a valid data-driven response?

2. Offering business people an increasing choice of intelligence, including both historical reports and analysis (what happened?) as well as predictive analysis (what do we believe is going to happen and which visitors do we believe are going to exhibit a certain behavior in the future?)

More and more often companies are asking web analysts to quantify the value of visitor. As practitioners in the data mining world are well aware, this is a complex analysis to perform and not something typically found falling out of a “report”. What they also know, however, is that an accurate LTV in conjunction with other predictive attributes often becomes the basis for implementing highly targeted “Longitudinal” (staged messaging over a period time) messaging strategies. The great news in today’s market is that the tools to perform this critical analysis are finding their way into the toolbox of the web analyst and stand to forever enhance both the their standing and value to senior executives and stakeholders.

What are those tools and how can web analysts begin figuring out how to use them?

The primary tools are not just software products, though software plays a vital role.

First of all, an analyst needs an analytical data mart of web behavioral information at the individual visitor level. This type of analysis is not done in the aggregate.

Secondly, the analyst needs to understand the nature of customer’s engagement with their company and fully understand the data itself. In order to use data to model behavior, the analyst’s role is to see through the noise and run valid experiments using raw data, derived variables, roll-ups, etc.

Third, the analyst needs at least 2 critical software capabilities:

1. They need an analytical solution that gives them the flexibility to manipulate the data and visualize it as you would in traditional reporting, but it must be highly dynamic- this is a real-time iterative business problem, not a batch reporting one.

2. They need an easy to use statistical mining application that can consume the working file of analytical data, determine the appropriate statistical algorithm and variables for answering the question at hand, and the automated ability to build/train/deploy a completed model. The last critical piece of necessary functionality/capability is part technical, part business process. The analyst needs to champion a closed-loop process for enhancing their web data mart with their predictions and tracing actual behavior (regardless of channel) to determine the success of their predictive experiment.

One of the most exciting aspects of Unica’s entrée to the web analytics marketplace is our deep experience in helping companies implement this sort of analytical rigor and the very unique set of web analytical and predictive solutions that can help a web analyst who is today doing reporting become a web analyst who is briefing executives on how customers and visitors are going to behave in the future and how much they are worth to the company.

About the Authors

Lee Isensee, Marketing Architect, has been in online delivery, ecommerce, and analytics for 15 years and is highly regarded for his work in web analytics and attrition measurement, conversion optimization as well as enterprise data warehouse integration.

Rick Fuller, Account Manager, has been in the marketing automation and web analytics space for a decade. Rick's background and experience includes BI/OLAP, online personalization, web analytics and traditional direct marketing technology solutions.

 

For additional information on this Newsletter article, please contact:

Lindsay De Santis
(781) 876-6249

Source: Lee Isensee & Rick Fuller, Unica

Related Documents:

  unica_cycle.gif    8.491 KB (8491 bytes) 

Content Tags: newsletter • newsletter

Tags: analytics web

 

Other Recent Articles:

Return to Web Analytics Articles Search