From Brent Tuttle, VP – Strategic Partnerships
Is the “Mood Factor” Skewing Your CLV?
Advertisers have been using a very solid scientific method to optimize their individual ads on television, in newspapers, magazines, etc., for well over over 100 years. Whenever there is significant money spent on acquiring customers, it is vital to test the different advertising campaigns and ‘paths to purchase’ consumers are responding to in order to maximize the return on advertising costs.
Recently, the internet and a plethora of new tools has emerged that now allows analysis-loving CMOs and other marketing executives to easily and precisely measure the number of views an ad receives, clicks through to a website, conversion rates of vistors becoming subscribers, etc.
Unfortunately, the exactness of the calculations to determine a customer’s lifetime value is often skewed by company methods of handling customers. Instead of deriving an exact value, the unique personality qualities of customer service representatives often causes calculations to be inexact.
Any company using ‘live’ agents to servcie customers will have inconsistencies that affect the actual statistics regarding refunds, chargebacks, save sale revenues, etc.
While the cost of acquiring a customer can be plotted with 100% accuracy, overall revenues from each customer can be skewed up or down by the “mood factor” when customers interact with live humans in customer service. Whether these humans were handling phone calls, email respones or support tickets, they are rarely consistent enough to provide the same experience for every customer.
In fact, depending on the level of consistency (or lack thereof), the actual customer lifetime value (CLV) calculation could be skewed 10% or more either way—by rep, by hour or even by day. These inconsistencies make it difficult to judge the true dollar value of each customer and can lead to wrong decisions.
Typical CLV Calculations
There are literally hundreds of calculations to determine CLV. Some are complex, like the one below taught in a Harvard CLV Analysis workshop.
Unfortunately, in every case, the calculations assume the data being collected is static and not influenced by the mood of the customer service rep or team.
Even Further Complexity
Many companies make things even more complicated by running tests to influence customers as they go through the purchasing experience.
For example, a company selling a golf product runs tests to see if traditional green and white colors makes customers perceive the product has a higher value opposed to when the product has a matte black finish.
Instead of simply focusing on the front-end experience customers have with their product, the company should test communications with customers when customers try to leave, cancel, return, refund, etc. to see if exit offers or down sell offers generate more revenue. Post-exit strategies include testing emails, postcards, catalogs, and other tactics used to entice consumers to buy again or buy more.
The Flaw Caused by Humans
While most of these data points will be absolute, the ‘human factor’ can skew the results where sample set revenue values are completely invalid. What we are talking about is Mary.
One day, Mary is feeling overly generous and grants refunds to practically every customer, thus skewing the refund percentage for the green and white box test. The next day, after a fight with her boyfriend, she’s angry and doesn’t give anyone a refund. This skews the chargeback ratio and losses for another test you have running. Then, for two weeks, she’s fine and gets almost all the customers she interacts with to stay with your program at a reduced save sale rate.
REGARDLESS OF HOW CAREFUL YOU ARE, REGARDLESS OF HOW SCIENTIFIC YOU ARE, REGARDESS OF HOW MANY (OR FEW) TESTS YOU RUN IN FRONT OF THE “MOOD FACTOR”, YOU CANNOT BE SURE YOUR CUSTOMER LIFETIME VALUE (CLV) CALCULATION IS VALID.
A Simple Recurring Billing Model Conundrum
|CAMPAIGN / SOURCE||PRICING||CONTENT||SAVE OFFER||VALUE|
|Traffic Source 1||$12 a month||Long Copy||Save Sale Offer 1 X Mary||$59|
|Traffic Source 1||$12 a month||Long Copy||Save Sale Offer 2 X Joe||$75|
|Traffic Source 1||$15 a month||Long Copy||Save Sale Offer 1 X Rick||$66|
|Traffic Source 1||$15 a month||Long Copy||Save Sale Offer 2 X Mark||$61|
|Traffic Source 1||$15 a month||Short Copy||Save Sale Offer 1 X Lucy||$84|
|Traffic Source 1||$15 a month||Short Copy||Save Sale Offer 2 X Patty||$77|
With different live agents either handling live chats, email tickets, or answering phones directly, you can see having just 5 different individuals with their own “mood factor” can skew results by as much as 20%.
How to Avoid the “Mood Factor” in CLV Calculations
In order to validate your CLV numbers and the retention tests you have running, it is a good idea to replace the human factor for at least a portion of your customer interactions. This allows you to remove the “mood factor” and any skew on your results to achieve a more statistically sound calculation.
This is possible with an Interactive Voice Responder (IVR), integrated IVR or Web Service platform. The exact same message is delivered every time with the same tone and same options.
Running a percentage of your customers through this path allows you to determine a near exact CLV. Having an exact number will allow you to confidently make decisions about what to do with your profits and what actions to take to become more profitable.
Using an automated platform such as RevGuard’s OCO ® IVR or OCO ® Web Platform allows you to test various save sale, down sell and other options that significantly enhance revenues, reduce refunds and reduce charge backs. In most cases, it is less expensive to systematically present customers with common self-service requests instead of spending money for live Customer Service Representatives (CSRs) to handle routine requestes. This saves money and gives your live agents time to work on more complex and difficult service requests.