Tag: marketing science

Loyalty Programs and CRM — Insights from Marketing Science 2009 Conference

Posted on

At the INFORMS Marketing Science 2009 Conference, I presented my current research project on the effects of loyalty program expiration policy change.  For those who missed the conference, here’s a summary of what I presented. I have also included the summary of the other two research projects that were presented during the same session.

Shortening Loyalty Program Expiration Period May Not Be Bad

One of the headaches companies running loyalty programs have is the liability associated with unredeemed points in program members’ accounts. One way of reducing such liabilities is to shorten the expiration period associated with program points, as major US airlines did in 2007. However, it is possible for such a policy change to alienate existing customers. To see whether this is the case, my co-author and I analyzed data from a convenience store chain that has switched from a no-expiration policy to a monthly expiration policy. To our surprise, we found that, participation in the program has actually increased significantly since the program change. Overall in-store sales have also increased, while fuel sales remained unchanged. What we plan to do next is to see how individual consumers have adjusted their purchase behavior in reaction to the policy change. We suspect that the addition of an expiration policy imposes what we call an “expiration pressure” on consumers, as consumers are pressured into making more purchases to reach a reward threshold before the points expire. However, different consumers (e.g., those with different patronage levels) will experience the pressure differently. I’ll report more findings when we are further along with the project. For now, you can download the presentation slides. We welcome any feedback or comments you may have.

Two other researchers also presented their projects on loyalty program and customer relationship management in the same session. As these are also relevant to loyalty managers, I am summarizing them below:

Loyalty Program Increases Share of Wallet by 10%

Martin Boehm from IE Business School presented his co-authored research project on the effect of loyalty program membership on consumers’ share of wallet at an European supermarket chain. By looking at a consumer panel’s behavior before and after loyalty program enrollment, they show that the loyalty program increased share of wallet by 10%.  This lift is negatively correlated with a consumer’s original share of wallet before the program enrollment. In other words, those with a high share of wallet exprienced minimal lift, whereas those with a low share of wallet experienced the highest lift. These results echo my earlier research findings showing a similar pattern. However, their research better controls for self-selection bias (i.e., better customers are more likely to enroll in loyalty programs) and therefore provides an even stronger argument for loyalty program impact.

Email and Mail Are More Effective Customer Contact Channels…

At least in the context of auto dealership’s services. Using data from an auto dealership, Andrea Godfrey at University of California at Riverside and her co-authors compared the effectiveness of phone, mail, and email customer contact in increasing sales (in this case, service revenue).  They found that mails and emails were similarly effective in increasing sales, while phone contact was the least effective. The effects of mails and emails were both curvilinear, meaning that the effects of those contacts reach maximum after a few times and then drop after the threshold. Not surprisingly, the exact effectiveness of each contact channel on individual consumers also depends on the consumers’ channel preference.  I hope these findings will help eliminate a few annoying dinner disruptions and result in less waste of papers. But I am not so sure I like the prospect of receiving more emails either. Hmm…

Questions or comments?  If you have any questions regarding any of these research projects that I have summarized here, please feel free to let me know, and I’d be happy to answer your question or forward your question to the right author.

What Do Marketing Scientists Have To Say About Social Media?

Posted on

This year’s INFORMS Marketing Science Conference paid lots of attention to social media and social networking. Quite a few sessions were devoted to the topic. Here I picked a few presentations with especial relevance to business practice and summarized them below.

1. Freemium model depends on users’ group membership and content contribution.

Presented by Gal Oestreicher-Singer from Tel Aviv University, this research analyzed a large dataset of 150,000 random existing users and a separate group of new subscribers of Last.fm. The authors wanted to find out under what conditions users are more likely to pay for premium services in a freemium model. Results suggest two main factors: (1) group membership and leadership: those functioning as leaders of Last.fm user groups are more likely to pay for a premium account. (2) Content contribution, in this case, posting of journal entries. Those who have written a journal entry on the site are more likely to pay up.  The effects of these factors, the authors found, exceeded the traditionally accepted influence of content consumption and local network size.  The authors also argue that the network position of the user within the community can also affect likelihood to pay. But this area is still being explored.

2. How To receive more incoming social links for your online shop?

In the social networking space for e-commerce, such as eBay, individual merchants can connect with other merchants by building links between their shops. Such incoming links can build traffic and increase business. So how can one increase the likelihood of being linked to?  Andrew Stephen, who recently joined the faculty at INSEAD, reported findings from his dissertation research that answer this question.  His research shows that reciprocity and assortment diversity are the two drivers of incoming links. In other words, merchants are more likely to have outgoing links to those who have incoming links to them, and they are more likely to link to those who have a large assortment of merchandise.

3. Celebrity, offensiveness, and honest labeling of content drive YouTube video success

I was especially excited about this presentation because it partially answers a burning question that has been on my mind for a while: “What makes some YouTube videos more successful than others?” The two presenters, Caroline Wiertz and Thorsten Hennig-Thurau from Cass Business School, City University, London, together with their co-author Michael Paul compared 100 top hit videos with 100 matching “flop” videos that were uploaded on the same day. They found three factors that have the biggest impact on the success of a video: offensiveness of the video content (i.e., controversy factor); featuring or association with a celebrity; and honest labeling of a video content’s via title, tags, and thumbnail. In a separate study, the three researchers tracked the number of views for 360 videos newly uploaded to YouTube over the course of a month. Their results suggest that the drivers of video views are different at different diffusion stages. In the beginning (the 1st day after posting), past channel success drives the number of views for the video. During the following days, however, existing # of views, volume and valence of user comments, and video description (via title, tags, and thumbnail) take over in predicting video popularity. In the final stage of their one-month tracking period, existing # of views and honest labeling become the most important driving factor. Their research findings offer clues to companies as to what to manage and pay attention to when doing viral marketing through YouTube.

4. Negative word-of-mouth effect accentuated by reviewer similarity and brand reputation

Ever wondered why bigger brands suffer huge backlashes from negative word-of-mouth? Debanjan Mitra from University of Florida addressed this question. By looking at book reviews from Amazon.com and Barnes & Noble websites and correlating them with sales rankings, he and his co-authors found that, besides the established finding of negative word-of-mouth having a larger effect on consumers than positive word-of-mouth, the similarity between reviewers and the audience further accentuate the effect. In other words, we are more affected by the negative opinions of people who are just like us. Furthermore, they found that the situation is even worse for higher-reputation brands (e.g., authors with higher ranking books).  Apparently, the higher reputation your brand has, the more you have to lose and the more devastating negative word-of-mouth can be. In the authors’ presentation title, they labeled this effect as “the weakness of strong ties”.

What do you think?  Do you agree with these marketing scientists findings and insights?