TextOre Blog

Agile Analytics

August 1, 2013

In the age of Big Data, analysts have to move quickly and adjust their course when necessary. We may call it 'agile analytics'.

Online content creation is a continuous flow, ever increasing in speed and volume. Today, anybody with Internet access can be a content producer in just a few clicks.

Keeping track of these movements and their implications to your business, brand, and products requires you to act quickly and decisively. 


In the past few years, agile development has dominated the software industry. Agile development is based on iterations and incremental development, adaptive planning, and time-boxing. Agile methods are designed to make it possible for teams to respond rapidly and flexibly to changes in the environment. Agile development methods allow developers to respond to changes in requirements, technologies, markets, and usage patterns.

Similarly, information analysts must rapidly respond to ever-changing conditions, trends, and demands. While it’s important to have long-term strategies and high-level goals, success is unlikely should you fail to adapt to changing circumstances. In many cases, the goal is a moving target.

From the outset, projects in analytics are defined by topics, information types, and key assumptions. From the main topics you may derive search terms, while sources of content are dependent upon the type of information you plan to mine.

Throughout the course of a project, both topics and information types must be evaluated continuously and critically. A first-iteration analysis will often turn up not just answers to initial questions, but also generate new questions. Analytical methods, tools, and processes must be able to react to such twists and turns. Experimenting with new search terms, comparing results, and reevaluating sources are processes that must be performed in real time.


Time-boxing, like agile software development, is a valuable method in analytics. Constantly overwhelmed by new data, the main challenge is often to get through relevant content before new content appears. By time-boxing analytics deliverables, one can avoid an ever-increasing backlog of unmined data.

Effective analytics software facilitates a great amount of automation and scheduling, which makes time-boxing easier. Automated content retrieval and scheduled results generation make it possible  to filter through results, pull out relevant content, and report findings in a timely manner.


Seeing is believing.  Giving decision makers partial or preliminary results early in the process permits you to receive valuable feedback. When explaining unfamiliar issues, pictures go further than words.  It doesn’t take a professional information analyst to tell if something looks good or if something is missing from an analysis.

If decision makers get an early look at preliminary results or draft reports, they also become “anchored” in their expectations for final deliverables. This reduces the risk of analysts getting major additions to their workloads in the last stages of an analytics project.


Although an agile analytics process should accommodate changes in scope, methods, terms, sources, and reporting, the overall business goals behind the project must not be forgotten. After all, an information analyst does not exist in a vacuum, and must relate to the needs of the company’s decision makers.

New insights, however valuable, must be relevant to the business. Moving the scope away from key business needs will not lead to success.

Rather, new findings may be presented as additional opportunities and used to spin off new analytics projects with relevant scopes and business-related goals. While responding to major existing business issues, analysis can expand your business by examining new areas, creating new ideas, and raising new questions.



Do you want to write a comment?