By using website you agree to our use of cookies as described in our cookie policy. Learn More

TDWI Upside - Where Data Means Business

Q&A: What Is the Value of Behavioral Analytics? (Part 1 of 2)

Analyst and expert David Loshin explains the theory and uses of advanced behavioral analytics.

As interactive businesses multiply and the explosive growth of online, mobile, and the Internet of Things continues, business applications must handle millions more interactions and transactions. Business success, meanwhile, continues to depend on driving customers toward profitable transactions.

In this two-part article based on a recent TDWI Webinar, industry analyst and data expert David Loshin describes the importance of behavioral analytics and discusses how aspects of behavioral analytics are suited to different business use cases.

Loshin, president of Knowledge Integrity, Inc, is a recognized thought leader and expert consultant in the areas of data quality, master data management, and business intelligence and a regular contributor to Upside.

Why has behavioral analytics become so important?

Over the past 10 years or so, the world has transformed into an information economy. All of the activities taking place within business processes are creating a new kind of data. We're not just talking about human-oriented applications producing this data, but also machine-oriented applications as well as hybrids -- humans and machines acting together.

You need to integrate this data -- coming from the millions of touch points that the human or machine actors interact with during each process -- into models that both support and drive the business processes of the information economy.

[Editor's note: The discussion with David Loshin continues here.]

Applications gather data to refine what we know about a customer so that we can do some profiling and customer analytics. Profiling explores only the characteristics of the customers themselves, not what they do or when or why. That's when behavioral analytics comes in. It's an analysis method used to explore what customers do, what kind of environment they do it in, and how and when they do it -- and especially, what triggers the desired results or the decision to act along various paths.

Our job is to streamline actions to the desired path. That path might be checking out with a full cart, making an in-app purchase in a mobile game, registering for an event, downloading a white paper, or registering for a conference. There are all kinds of use cases and scenarios in which we want to look at profitable actions or desired results and then examine the path that can get the actor to take that profitable action.

In short, behavioral analytics is important because it helps enterprises encourage those profitable actions and facilitates the massively scalable customer service needed today.

Let's take a step back and ask, what exactly is behavioral analytics?

According to Wikipedia -- and I'm reading here -- behavioral analytics is:

"a subset of business analytics that focuses on how and why users of e-commerce platforms, online games, and [W]eb applications behave. [It allows] one to take seemingly unrelated data points in order to extrapolate, predict and determine errors and future trends. It [...] connects individual data points to tell us not only what is happening, but also how and why it is happening."

Let me point out that this definition specifically focuses on one classification of application. In fact, behavioral analytics can be applied to any situation in which you have a broad constituency of individuals or entities and you want to monitor the behaviors taking place within the system in order to get to the optimal result. It needn't be limited to e-commerce or online situations.

Behavioral analytics is composed of a combination of classification and segmentation that allows application designers to engineer efficient pathways to encourage profitable outcomes.

The idea is that we want to look at not just who is interacting with our environment and the characteristics of the parties that are interacting, but also look at what they are doing, when they are doing it, and how it's being done. If we do that, then we can look at the sequences of events that lead to the scenario we want to achieve, and then revise the application to encourage those paths.

We want to reduce opportunities for people to go off those paths and allow them to reach the desired results faster. This benefits the organization, but it also benefits users -- their experience is enhanced if they can get to the end result faster.

Can we take this discussion of behavioral analytics and apply it to e-commerce specifically?

We can lay out a typical approach like this: the user visits a website, searches for a specific term, navigates to a specific product page, clicks through some recommendations, maybe adds an item to the cart -- and then maybe leaves the site without buying.

If we want to discover why, we should look at the transaction histories of all the buyers and cluster them together by characteristics. We then examine the customers' transactions, not only those where the desired final transaction took place -- but also where the desired transaction did not take place. Our goal is to see why customers left the site without completing the purchase. What happened along the way?

Once we have this data, we can take some actions within our application so that the customer who previously might have abandoned the cart wouldn't do so in the future. We might modify the search to present different paths -- maybe to items that are frequently purchased. Maybe we modify the search to present items whose sales actually completed, as opposed to ones that didn't.

You can see that we essentially use the results of our behavioral analytics to refine our processes and tweak how our application behaves, all in the interest of trying to influence how visitors behave.

What are some examples of using behavioral analytics in other industries?

Health care offers many examples. In the interests of proactive health care, we might examine the characteristics of patients who are filling their prescriptions on time versus those who aren't. We might look at doctors prescribing different medications over time versus those prescribing the same medications -- what are the optimal health benefits to patients?

We can also use behavioral analytics to look at machine data -- for example, examining manufacturing lines across a number of different factories. Looking at sensor readings from the factory floors, we can ask: what steps led to a failure? Can we build a predictive model anticipating future failure based on that?

Across all of these use cases, some of the commonalities are: many different entities operating simultaneously, sequences of actions that lead to some desired result, and an opportunity to find places to improve the process and the user experience. Additional fields where advanced behavioral analytics is being used include threat detection, fraud analysis, content management, and performance optimization.

In all of these cases, the intent is to figure out where your processes are going right and where they are going wrong. When they go wrong, how do you fix them? When they go right, how do you get that to happen again?

[Editor's note: The discussion with David Loshin continues here.]

About the Author

Linda L. Briggs is a contributing editor to Upside. She has covered the intersection of business and technology for over 20 years, including focuses on education, data and analytics, and small business. You can contact her at

TDWI Membership

Accelerate Your Projects,
and Your Career

TDWI Members have access to exclusive research reports, publications, communities and training.

Individual, Student, and Team memberships available.