Why loyalty programs are failing and how to fix yours

An article in MediaPost today highlights why few companies are delivering upon their loyalty programs. In a nutshell, consumers want personalized rewards, whether it be through their grocery store cards, airline frequent flyer program, or other points-based systems.  But marketing departments are not able to personalize because they are unable to collect the data needed to personalize offers and internal organizational hurdles prevent companies from unifying their data across multiple silos.

I was fortunate to spend a few years at American Airlines to perform analysis on the AAdvantage program– to learn firsthand how the granddaddy of loyalty programs operated. Some challenges and how we overcame them:

Incomplete customer data
The website, reservations deck, and gate agents all had separate customer databases. If you were a smart customer, you could complain at all three locations and earn triple the points. So if a bag supposedly fell on your head from opening the overhead bin, you could get miles for the inconvenience, and the agents at the airport, on the phone, and from the website wouldn’t know that you were already compensated. 

The solution– create a unified customer database. This is quite expensive, and has political issues to solve, but is well worth the effort.  When you can create a single view of the customer’s activity, you can measure their overall profitability and create programs to incent the right kind of behavior.

Mass blasting (spamming) customers

Email marketing is so easy and inexpensive that SVPs of Marketing are tempted to spam customers.  After all, if sales increased by changing the newsletter frequency from monthly to twice a month, why not weekly? We’ve seen this approach taken at a number of large brands.  Not only is this a customer turn-off, but goes against the whole point of offering personalized offers that are most appealing to where a customer is in your lifecycle and their stated preferences.   Part of status is being remembered for your particular likes (mints on your pillow and extra towels if you’re a hotel customer), not to get the same message delivered to everybody.

The data crypt

What used to be called a data warehouse, then later called a datamart, then called business intelligence, then enterprise analytics is really just the same thing with a new name each time. Even if you can embark on a $20 million project to consolidate customer data into one spot, the bigger issue is actually getting it out.  Most marketing managers believe they have to speak a special prayer to the high priests of IT to be granted access to the database. The inability to easily access the data– no matter what expensive analytics tool you might have bought– prevents the marketing manager from being able to do the analysis necessary to create a tailored menu of rules. Without a menu of actions and corresponding points, and then being able to adjust payouts based on what’s working, a loyalty program bleeds.  And the more data you have in your scoring model, the more complex it is to calculate status, as you’re dealing with an increasing number of empty fields for variables and bad data.  So perhaps you were able to get a data append from a third party– let’s say Acxiom– and now you have gender, income, and the type of car they drive.  How is that affecting the ways to earn and burn points in your loyalty program?

IT doing Marketing and Marketing doing IT

Managing a loyalty program is really an exercise in user psychology, but it requires some technical execution.  You’re really looking at video game design and trying to influence their behavior.  An average IT administrator is not going to understand this, nor is a traditional brand marketer that has done media buys for 20 years. If you have a loyalty program in-house, ask yourself who is running it.  

The answer is finding folks who are well-versed in data analysis (crunching SQL statements in the database) and also understand user psychology.  Writing SQL is for engineers, you say?  Look at Amazon, where they require marketing managers to know how to query a database.  It’s not that hard.  How are you going to structure and adjust rules for earning and burning points by customer segments if you’re not able to go in there and hands-on be able to run reports? 

Summary– the weakest link

Well there you have it– all it takes is one break in the chain and your loyalty program has a problem.  If you aren’t able to collect user data– transactions and preferences, you can’t create personalized offers. If your marketing people can’t properly access the data, that expensive database just sits there. If you have the wrong people trying to solve the problem, it’s a guy with a hammer who thinks everything looks like a nail.

I’d say that brands would be well-served to hire folks from Nintendo, Blizzard, and other companies in the gaming industry to help revamp their loyalty programs. This is nothing more than a video game.


Leave a Comment

Scroll to Top