We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Use Advanced Analytics to Dig Deeper into Word-of-Mouth Behavior

Originally published September 28, 2011

Many retail, consumer packaged goods (CPG) and media organizations across the globe are creating digital platforms to increase engagement with their consumers. One of the keys for driving engagement is to track word-of-mouth behavior (WOM) of identified registered users of digital platforms. This article presents a framework for quantifying and digging deeper into drivers of word-of-mouth behavior using advanced analytics. Specifically, I focus on three high-impact scenarios (Figure 1):

  1. Understanding drivers of word-of-mouth behavior using multivariate/regression analysis

  2. Understanding discriminants of viral and non-viral behavior using Multivariate Analysis of Variance (MANOVA)/ discriminant analysis/ Analysis of Variance (ANOVA)

  3. Network link analysis to understand hot spots within a social network


Figure 1: Three Drivers of Word-of-Mouth Behavior

(mouseover image to enlarge)

This article  will also focus on manifestations of word-of-mouth behavior to amplify and make messages viral on three primary channels – Facebook, email and Twitter (Figure 2).

Figure 2: Three Primary Channels in Word-of-Mouth Behavior


There are three important drivers that have created a sudden surge of interest in word of mouth behavior:

  1. Source of digital traffic. More people visit digital platforms as a result of friend referrals on social media (Facebook, Twitter) than by their own searches. This means it is important to quantify the current state of word of mouth behavior of the digital platform and understand what drives that behavior.

  2. Trust. According to The American Marketing Association, about 90% of buyers trust peer reviews and only 15% of buyers say they trust traditional advertising. Peer recommendations are very important in gaining trust.

  3. Stickiness driving purchase action. Once a platform becomes a trusted or “sticky” source for friends and family, it is more likely to result in the ultimate intent of the brand manager, which is a purchase by the digital consumer.

Let’s first look at a few real life examples of acts that constitute word of mouth behavior on a digital platform:

  • In the media/publishing industry, an article written on demystifying today’s world economy in a digital publication website can go viral through Facebook

  • In the CPG industry, discount coupons for a brand could be transmitted among a close network of friends and family

  • In the financial industry, an interesting review of a financial product by a key opinion leader can be Tweeted

There are four engaged actions a digital consumer is likely to take:

  • Facebook “likes.” A digital consumer may press Facebook’s “Like” button

  • Facebook wall post. A consumer may post a comment or referral on his/her Facebook wall.

  • Email activation. A consumer may type in 10 email addresses of friends to refer a coupon.

  • Tweet.  A digital consumer may lick on “Tweet” icon to broadcast about an interesting recipe

Table 1 represents a list of questions that word-of-mouth  analytics strive to answer based on the four engaged actions listed above:

 No.  Word-of-Mouth Scenario  Analytical Construct  Business Implication
 1. What are the key drivers of word-of-mouth behavior? Regression Align navigation experience to trigger WOM behavior
 2. What factors discriminate viral content from non-viral content? Discriminant Analysis/ANOVA Align content on digital platforms
 3. Who are the key mavens in the social network who are capable of "tipping" an article and making it go viral? Network Link Analysis Narrow Campaigns

Table 1:  Three Questions Word-of-Mouth Analysis Answers

Analytical Scenario 1:  What drives word-of-mouth behavior?

There are many factors that could potentially drive word-of-mouth behavior among a community of registered users on a digital platform (Figure 3). It could be the frequency of logins. It could be the sequence in which they consume content. It could be the age group in which the registered user belongs. For example,  it’s quite conceivable that people in the 18-25 bracket value opinions by a particular registered user who has significant influence in the online community, and the sentiment index of that comment could trigger other registrants to broadcast that comment on Facebook, Twitter or email. All of these factors can be introduced into a multivariate process like regression, and the relative weight of each of these influencers can be determined.

Figure 3: Driver Analysis Using Regression


Analytical Scenario 2: What differentiates viral activations from non-viral activations?

Not all activations are “viral.” For example, a microsite to launch a new product can have hundreds of registered users referring friends and family members, but no proportionate increase in the engagement metrics for the product. It means even though a link has been passed along within the community, the recipients of the link have not executed the desired action. It is important to weed out the viral activations from the non-viral activations to see what factors are specific to each. Techniques like discriminant analysis / ANOVA and MANOVA can help the analyst isolate the key variables that differentiate viral activations from non-viral activations (Figure 4).

Figure 4: Discriminants of Viral vs. Non-Viral Activations

Network Link Analysis

Analytical Scenario 3: Who are the top “thought-leaders” in the social network formed by referral behavior?

If a registered user uses email to pass content for consumption among friends and family, a social graph analysis of linkages can determine the influence of key people on creating influence surrounding the article (Figures 5, 6).

Figure 5: Example of Top "Thought Leaders" Network

 From  To  Content Type
 # of Activations
 Scott  Radha  Digital coupons
 Radha  Sanjoy  Product review
 Radha  Sanjoy  Digital coupons
 Sanjoy  Krishnan  Product review
 Krishnan  Scott  Digital coupons
 Melody  Joe Henry
 Product review
 Krishnan  Joe Henry
 Digital coupons

Figure 6: Activations Resulting in Network

To conclude, we have seen why quantifying word-of-mouth behavior is important as an engagement vehicle and how the implementation of three specific analytical scenarios can help businesses understand drivers of “stickiness” of their digital platforms. A firm understanding of the drivers of word-of-mouth behavior sets the stage for increasing the revenue contribution from the digital platforms.

  • Derick JoseDerick Jose

    Derick Jose is the vice president of Advanced Analytics/Research within MindTree's Data & Analytic Solutions (DAS) Group, one of the world’s largest information management practices, which offers customers a one-stop-shop to capture, analyze, enhance, and view their business information. The DAS practice combines MindTree’s proven analytics, business intelligence, information management and research services for customers in the consumer packaged goods (CPG), retail, financial services, insurance, travel and media markets. Derick has 20 years of experience spanning consulting, advanced analytics and business intelligence solutions. He has worked extensively in the CPG, banking, telecom and retail industries. Derick can be contacted at Derick_Jose@mindtree.com.

    Editor's Note: More articles and resources are available in Derick's BeyeNETWORK Expert Channel. Be sure to visit today!

Recent articles by Derick Jose



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

Be the first to comment!