Earn Back the Love of Demoted Customers

Posted on

If your business offers a tiered loyalty program with an annual cycle, beginning of the year may be a “heartbreaking” time for customers who failed to earn enough and was demoted from their previous premium tier status. Some argue that this is one of the downsides of a hierarchical loyalty program. What can you expect from these customers? What can you possibly do to earn back the love (or at least not lose the respect) of these demoted customers? I looked into research in this area to find some answers.

Negative Impact from Demotion

Just like anyone who have flown first class may feel quite miserable going back to the economy cabin, losing one’s premier tier status in a loyalty program is likely to have some damaging impact. This negative impact can have a more “rational” source and a more emotional source. On the rational side, getting demoted means the loss of important benefits such as higher point earning ratio and free perks. So objectively speaking, consumers lose something concrete when they get demoted. But that objective loss is only part of the story. The most important impact comes from the emotional effect of demotion. It is common for demoted customers to feel frustrated, disappointed, uncomfortable, or even angry.

Whether the demoted customer is driven by rational loss or by an emotional response, the ultimate outcome is higher dissatisfaction, lower loyalty toward the company, and a higher likelihood to switch brands. Not surprising? Here comes the kicker. Not only is demotion bad, the damage from demotion (e.g., silver to base tier) is much stronger than the positive effect from an equivalent promotion in the opposite direction (e.g., base tier to silver). In one study, researchers found that even after a demoted customer is restored to the lost premier tier at a later point, the overall loyalty level is still lower than before the customer was demoted. This is quite a serious issue. Unless most of your customers are on an upward trajectory in terms of how much they spend with your business, your gain from rewarding customers with premium tiers may never quite compensate for the loss you will suffer when consumers lose their premium tiers.

Who Suffer the Most?

The negative impact from demotion is not the same across customers though. Knowing which customers tend to respond more strongly can help you take proactive measures to avoid really negative consequences, e.g., keeping some customers in the old tier even if they have not quite spent that much last year. Below are some factors suggested by previous research, many of which should be fairly intuitive. Continue reading “Earn Back the Love of Demoted Customers”

10 Resources for Learning Marketing Analytics

Posted on

It’s hard to believe a whole month of 2019 is already gone. A new year is a great time to learn new things. In that spirit, I would like to share with you several resources I love for learning marketing analytics. As many of you are aware, marketing analytics has become an important skill set to have in today’s workplace. This list will help put you on the right path to becoming more proficient in marketing analytics.

Online Courses

1. Marketing Analytics Course by Columbia University on EdX. As part of the ColumbiaX’s MicroMaster’s program in business analytics, this 12-week online course offers a great introduction to marketing analytics. It covers key topics such as customer segmentation, customer choice modeling, optimal pricing, and conjoint analysis for new product design. Each week you watch a set of pre-recorded videos and complete an assignment on a given topic. You can also ask instructors questions on the discussion board. I like that the course teaches the basic concepts of marketing analytics as well as their implementation using R. You can take the course for free, or you can pay to get a course completion certificate at the end. It appears to be offered several times a year. The most recent one started on Jan. 27, 2019, not too late to sign on if you are interested.

2. Marketing Analytics courses from DataCamp. The DataCamp website is devoted to data science training. It offers 200+ data analytic courses and uses a subscription-based model. Currently I can find four specific marketing analytic courses: Marketing Analytics in R: Statistical Modeling; Marketing Analytics in R: Choice Modeling; Cluster Analysis in R; and Sentiment Analysis in R. There are also many other general data analytics courses dealing with either tools (R, Python) or methods (e.g., machine learning, data visualization, statistics, etc.). Each of these courses is not very long, at a few hours. But across all courses, there are a lot of things to learn.

3. Social Media Data Analytics. For those interested in analyzing textual data rather than numbers, there are a number of online courses available on this topic. I chose this Coursera course for its wider coverage and depth of materials. Offered by my alma mater, Rutgers University, this four-week course covers the basics of extracting and analyzing large quantities of textual data from social media. The main tools taught are Python and R. It covers a limited number of social media channels, including Twitter, YouTube, Flickr, and Yelp. But the working knowledge of APIs acquired through the course should readily extend to other social media channels.

Books

4. Freakonomics. This is a book you probably don’t expect to see on this list. It doesn’t teach any analytical methods, but it is a great demonstration of how we can find answers to questions through smart and creative use of data. I strongly believe that the abilities to ask interesting questions and translating numbers into insights are just as important as crunching numbers and running sophisticated statistical models. This book is a good example of the data to insight process.

5. Quantitative Models in Marketing Research. Compared with freakonomics, this book represents the other end of the spectrum and focuses more on crunching numbers (i.e., data modeling). This concise book is organized by data types (e.g., continuous variables, binary variables, etc.) and describes the most common ways of modeling each data type. Each topic is organized similarly, in the order of general model description, estimation, diagnostics, an application example, followed by more advanced issues on the topic. I like that each chapter comes with an illustrative example, and you can download the data used in each example to try the analysis yourself. For some topics, the book offers the analysis code in Eviews, another common statistical package commonly used in economics. The book does requires knowledge of basic statistics and matrix algebra to digest more easily.

6. R for Marketing Research and Analytics. This book shows you how to use the open-source software R to tackle the most common marketing analytics problems, such as segmentation, basket analysis, and data summary and visualization. It does assume some familiarity with the underlying concepts and focuses mostly on their implementation in R. I like that the book offers clear R script examples in each area, making it a very hands-on learning experience.

7. Web Analytics 2.0. Published in 2009, some might consider this book a dinosaur in the digital age. But I love this book and think it is as relevant as ever. Yes, the specific examples using the various web analytics tools may be outdated. But the basic approach behind how to identify and interpret key web metrics is quite universal and timeless. I like the book’s very interpretation-oriented approach, focusing on what the different metrics mean or sometimes do not mean. This is a great starter book if you want to become familiar with the world of web analytics.

Websites

8. Kaggle.com. Devoted to data science, Kaggle.com offers a variety of resources for current and future data scientists. It has over 10,000 datasets on a wide variety of topics, to give you hands-on experience with a dataset of your interest. You can take free short online courses on important data science tools such as R, Python, SQL, and machine learning. If you feel good about your data analytic skills, you can even participate in various Kaggle competitions. At the time of writing, there are 16 active competitions running with prizes ranging from knowledge and swag to $100,000.

9. Occam’s Razor blog. Blog from the author of the Web Analytics book mentioned earlier, Avinash Kaushik. Just like the book, the author delivers sensible advice on how to do analytics right, with a razor-sharp focus on identifying what is important. It is clear from this blog that more is not always better when it comes to marketing analytics.

10. Big Data, Plainly Spoken blog. This blog by bestselling author Kaiser Fung discusses the use and interpretation of data in diverse domains, including economics, education, and politics. What makes the blog stand out is its making sense of data in layman’s words. Just like the Freakonomics book, you can learn a lot about translating data into meaningful information here.

 
I hope you find the list helpful. If you are aware of a good learning resource that is not listed here, please feel free to add a comment below or drop me a note. Happy learning!

The Basics of Behavioral Segmentation

Posted on

What is Behavioral Segmentation?

There are lots of things that you can try to learn about your customers. For example, you might want to gather information on how old they are, where they live, and whether they are married and have kids. You’re probably also using information about how often they buy from you and when they last bought from you to predict their likelihood of buying again in the future. Knowing such information can help you optimize your loyalty program and target the most enticing rewards and marketing communication to a particular group of consumers.

Traditional customer segmentation has often been based on demograhpics, geographics and socioeconomic status. However, as Gibson Biddle, former VP of Product at Netflix pointed out in his Habit Summit talk earlier this year, common demographics tend to predict consumer preferences rather poorly. Instead, what people do are often more telling about their taste and what they may like or dislike. Together with an increasing amount of available data about consumers, this has prompted rising interests in behavioral segmentation, that is, segmentation based on what people do. Over the course of this year, I have worked with a CASC business partner on a customer segmentation project using loyalty program data. The practice has been really insightful in revealing hidden groups of customers for great targeted marketing opportunities that may not have been apparent on first look. Continue reading “The Basics of Behavioral Segmentation”