Channel: Information and Analytic Strategy - David Loshin RSS Feed for Information and Analytic Strategy - David Loshin

 

Analyze Customer Profiles and Behavior for Better Decision Making

Originally published May 24, 2012

If the intent of business intelligence and analytics is optimizing business opportunities, then most business scenarios involving revenue generation and managing the customer experience must reflect what can be learned from interactions with customers. There are at least three aspects to this consideration: who the customers are, how they behave, and how you can influence changes to customer behavior that benefit both you and the customer.

Customer Knowledge and Customer Profiles

The first aspect is “customer knowledge,” implying awareness of key customer characteristics that are relevant to your organization’s business processes. Your analytical framework will most likely want to capture these characteristics in a “customer profile” to help in both analyzing the different archetypes of customers you have and how each of the archetypes interact with the business.

The phrase “customer profile” is used in similar contexts with different meanings. In one sense, a profile provides a general overview of the customer with details about inherent characteristics (such as “name,” or “birth date”), descriptive demographic characteristics (such as where the individual lives, whether the individual is married), preferences (such as the customer’s favorite sports team), as well as analytical characteristics such as purchasing patterns or credit worthiness.

A slightly adjusted view of a customer profile is mapped to your business and the way your business interacts with customers. In this view, each customer entity is grouped into classes of customer profiles. The value of each customer type is calculated in terms of specific variables relevant to creating value, such as the number of products purchased, or the frequency of store visits, or the variety of products bought.

Customer Behavior

Customer behavior is also a somewhat fuzzily defined term. For our purposes, let’s suggest that “customer behavior” models are intended to capture information about the actions a customer performs under specific circumstances.

As an example, let’s say that a retail company emailed a special coupon for an in-store purchase of a particular item. There are a number of specific circumstances associated with this scenario: the presentation of the offer, the method of presentation, the time at which the offer was presented, the time the customer took an action, the time frame associated with the offer and a particular retail location. Given this scenario, the retailer can track customer actions in relation to the circumstances – e.g., the customer ignored the offer, or took advantage of it at some specific time and location.

Developing Behavior Models

In essence, the initial objective of capturing and analyzing customer behavior is to develop models reflecting customer decision-making processes. In turn, these models are expected to help predict behaviors associated with the different customer archetypes. To continue the example, the company can link a tracking mechanism to the email campaign, such as a bar code to be scanned at the point of sale. After the conclusion of the campaign, customer response statistics can be collected. That data set can then be subjected to dimensional analysis based on the customer profile characteristics. This will allow some segmentation to suggest any correlation between selected variables and purchasing the product, such as showing predispositions like:
  • Customers between the ages of 30-40

  • Customers living within 2.5 miles of the retail location

  • Customers with an income between $80,000 and $100,000 per year

  • Customers who vacation in Florida during December
Remember, though, that correlation does not necessarily imply causation. Identifying potentially dependent variables may suggest a predisposition, but establishing the predictive nature of this suggestion requires additional research. The bottom line is that there is a need feeding the insight back into the process, and you should ask questions such as:
  • Can the correlation be validated using additional campaigns?

  • Can the causal nature of the correlation be verified? Does this require some alternate investigation such as surveys or focus groups?

  • Can the variable dependences be refined to better target the customer profiles?

Influencing Change

If the first objective is to understand who the customers are and what they do, the next objective is to apply what you believe about customer archetypes and behaviors to influence changes in relation to customer behavior. There may be many approaches to behavior change, and that effort can be directed inside the organization or outside to the customer community.

In one approach, the results of customer profile analysis might highlight a gap in a business process in which the intended target customer audience is not being reached. In that case, the process can be adjusted to broaden the reach to the targeted customer types. An example of this is determining that a particular age segment is not responding as expected, and altering the marketing and advertising plan to insert television advertisements on channels or networks with the desired age-group audience.

In another approach, you might see that one specific customer segment is predisposed to taking a certain action. In this case, you might determine if there are other customers who might be easily moved from one segment to the desired segment through a sequence of engagement. An example is an airline seeking to engage more business travelers by offering “elite status” incentives for purchasing air travel. At some point, travelers with improved elite statuses will have ratcheted into the desired customer segment.

There can be some variety in the approaches to influencing change, some of which are hybrids of changing both internal processes and external characteristics. The common thread, though, is that analyzing customer profiles and behavior will prove to be a meaningless task unless actions are taken in reaction to what is learned.

SOURCE: Analyze Customer Profiles and Behavior for Better Decision Making

Recent articles by David Loshin



 

Comments

Want to post a comment? Login or become a member today!

Be the first to comment!