Click here to preview the new Fast Company

Want to try out the new

If you’d like to return to the previous design, click the yellow button on the lower left corner.

When Loyalty Programs Are A Waste Of Money

In times like these, cost cutting and streamlining are compelling business strategies. Business executives sometimes have second thoughts about their loyalty programs and ask our consultants whether it’s really worth the cost.  

The answer, however, depends on how well the program is aligned with the type of customer base being served. 

A loyalty program, or frequent-buyer program, rewards customers with points, miles, or other credits that can be redeemed for discounts and free products. Loyalty programs have become ubiquitous in a variety of industries, from airlines and groceries to credit cards, packaged goods, mobile phone carriers, coffee and restaurant chains, and retailers of all kinds. But they are probably over-used; last year in the U.S. alone, researchers tallied more than 2 billion loyalty program memberships, which means the average U.S. household belongs to about 18 different loyalty programs.  

Loyalty programs do cost money, not just in terms of the rewards themselves but also the administrative burden, so it’s not uncommon for executives to question their decision to implement one. Evaluating a program, however, should be based on two issues: First, under what circumstances will a loyalty program generate incremental repeat business? And second, how valuable are its other benefits, including the chance to gain insight into individual customer needs and preferences?

In 1996, Martha Rogers and I published our second book together, Enterprise One to One, and we tackled the first question head on. We concluded that a loyalty program can directly increase customer loyalty when a business’s customer base had two characteristics:

1. Just a few high-value customers do the vast majority of business; and

2. Customers’ needs are fairly uniform, meaning there isn’t much product differentiation in the category.  

(Recently my Twitter buddy Arie Goldshlager alerted me to the fact that in 2011 two Yale marketing professors, K. Sudhir and Jiwoong Shin, were given the John D. C. Little Award in marketing science for having proved Martha’s and my argument mathematically.) 

The long and the short of it is that paying customers for their loyalty is more likely to generate a direct profit when your customers have similar needs but highly different values. The airline industry is a great example. At an airline, the top 1% or so of flyers generate a substantial majority of the profits, and yet customers are fairly uniform in terms of what they actually need. Aside from seat or meal preferences, customers all want the same basic thing—to get safely and reliably from Point A to Point B—and pretty much any airline that flies a route can do the trick. 

So airlines can profit by purchasing customer loyalty directly, but if your business is not characterized by a similar kind of customer base, with a small minority of extremely high-value consumers who have relatively undifferentiated needs, then it might not make as much sense for you. 

Which brings us to the second issue: What are the other business benefits a loyalty program can generate? The real secret to the vast majority of loyalty programs has to do with compiling and using customer-specific data. By encouraging customers to identify themselves (in order to get their benefits) you can track their purchases and interactions, and then use insights from this data to tailor your offer, your product or your service to individual tastes. In effect, rather than rewarding customers for their patronage, you’re rewarding them for identifying themselves.

This strategy, however, will be more compelling when you don’t already have a natural mechanism for linking customers’ purchases to their identities, and when your customers are highly diverse in their needs. Grocery retailing is a good example here. There are about 40,000 different products on the shelves of a typical U.S. supermarket, but the average household will stock less than 1% of them, and every shopper buys a different assortment, with different brand preferences, types, and sizes. Moreover, unless shoppers somehow identify themselves at the cash register, the grocer has no practical way to keep a record of any individual shopper’s purchases. By using a loyalty program to identify individual customers and track each customer’s transactions, however, a grocer can compile enough data to make personally relevant offers.  

Tesco is a U.K.-based grocery retailer that does exactly that, for instance. One of the world’s most sophisticated users of customer data, Tesco launched its Clubcard loyalty program in 1995, and it now sends a quarterly newsletter to 16 million U.K. Clubcard members in 9 million different versions! Every Clubcard member gets a highly customized set of discounts and offers, and the company claims a response rate on the newsletter of some 25%. Analysts have estimated that the program could be generating more than £100 million in incremental revenue for the company every year. 

In recent years, Tesco has been expanding beyond the U.K. with a number of innovative offerings in different countries, and I think if there’s one business model that might eventually give WalMart a run for its money in the U.S., it would be Tesco’s. WalMart offers "everyday low prices" to every shopper, but some day Tesco could offer "personalized lower prices," based on each individual shopper’s own needs.

So if you want to avoid wasting money on your company’s loyalty program, ask yourself a few basic questions:

  • How much of your business comes from the top 1% or 2% of your customers?
  • Is it possible to identify and track your customers’ individual purchases even without a loyalty program?
  • Do your customers have diverse needs and preferences? and
  • Are you prepared, organizationally, to treat different customers differently?

For more leadership coverage, follow us on Twitter and LinkedIn.

[Image: Flickr user Mcscrooge54]

Add New Comment


  • Brett Perlman

    I have been involved with providing businesses with loyalty program solutions since 1997. 
    Over the years, there has certainly been an evolution. Today, more than ever, technology has become affordable and even the smallest of businesses are capable of and are implementing their own loyalty program.  Most are offering rewards from their own high-margin inventory, to keep the program costs down.

    The long-term cost benefit of having vs not having a loyalty program is difficult to truly measure.  However, from personal observation, one trend has evolved over the years, and that is many consumers are influenced by loyalty programs and many expect them.  Not having one would seemingly put a business at an immediate disadvantage.  I think another consideration that is not easily measured is the benefits achieved when a business starts focusing on knowing their customers and engaging on a more personal level with their customers; something that happens as a result of loyalty programs.

    Brett PerlmanPreferred Market Solutions, LLC

  • Amberoon


    Great article. With today's technology it is possible to create quantitative models that provide a lot more insights on micro-segmentation of customers based on their propensity to buy a product. We can now take loyalty data combined with social media and customer support data to get insights on "nano-segmenting" customers based on their customized basket of preferences. Some of these technologies were not available or very expensive as recently as 5 years ago.

  • Emil Koutanov

    Interesting article. But what (or rather, how much) does it take to have a good customer loyalty program these days?

    Recently there has been a move away from traditional magnetic card, stamp or punchcard based schemes to online loyalty programs. While these schemes vary, the common element is a push toward eradication of a traditional card, in favour of an electronic equivalent, such as MazeCard.

  • David Gee


    You raise some excellent points. Contrary to popular belief all customers are not created equal. The cost of providing superior customer is expensive and at some point those costs need to be analyzed (unfortunately most companies do not go through the exercise). Identifying those customers that most closely align with your corporate objectives and identify what will stimulate the desired behavior is critical. If a firm does decide to move forward a comprehensive analysis needs to be done of the lifetime profit of costs vs. the cost of loyalty program launch, maintenance (administration and rewards). At the end of the day running a pilot or rolling out to a segment and maintaining constant analysis through A/B testing is the only true way to measure loyalty program ROI.

    Dave Gee
    CEO-Bungee Loyalty Programs

  • Don Peppers

    With respect, Jon, it's not hard to write a book or two about the conditions under which a loyalty program can be justified, but I forced myself to boil it down to a few paragraphs, and I stand by what I said.

    Before you suggest (in your second paragraph) that the prerequisites for genuine success that I described are "unprofessional" you really should read the literature.  I suggest starting with Chapter 3, "Mapping the Strategy," in our book Enterprise One to One (1997).  Then, if you still disagree, you should check out Sudhir and Shin's paper and feel free to take issue with their fairly sophisticated mathematical equations proving our point.  You can find their paper here:

    As for all the other things to look for in a loyalty program (your first paragraph), you certainly do have a point, but all these benefits (higher frequency, more spending, etc) can be summarized with a single principle:  A loyalty program must generate incremental revenue and/or an increase in total customer lifetime values in excess of the rewards and administrative investment inherent in the program, and it must do this every year.  To prove the worth of the program you have to set aside a control group of customers not solicited for it, in order to isolate the incremental results.  This can be tricky, but for the most part companies don't even bother to do this, instead plunging ahead with their program and "proving" it with a global result that produces a fake ROMI based not on the program itself but on general economic conditions.

    If you DO set up a control-based system to measure results, what you'll still find is that in order to change customer behaviors to get the results you listed, you must appeal to customers in some way that is relevant to their needs and preferences.  This will be easier to do the more those needs and preferences are diverse.  If their needs and preferences are not diverse, then you don't need to undertake the expense of individualized marketing in the first place, right?  You can just advertise and promote the same benefits to all.  

    I would also recommend our CRM textbook for graduate business students, Managing Customer Relationships (2nd edition, 2011), at pp. 106-109.  There is a sidebar here titled "Real Objective of Frequency Marketing Programs" and it covers many of your points.  

  • Jon Underwood

    These are smart people who get paid lots of money, but this is an extremely narrow and limited answer, to the point of being unprofessional. A good loyalty program should 1) Increase the profitability of existing customers, which happens when they A) Come back more often (increase in frequency), B) Spend more per visit (increase in average ticket), and C) become more loyal due to the rewards (increased retention/less attrition). They also can 2) attract new customers. If NONE of these things happen, your program is a complete waste of money (as many are), and the business should drop the program. However,  If ANY of these things happen, then the company should compare the cost to the increase in revenue per person to determine the individual ROMI, and ADJUST THE OFFER if the return is too low. When more than one of these things happen, then almost assuredly the company will lose more money than they save if they drop their loyalty program.

    This bit about: 'a loyalty program can directly increase customer loyalty when a business’s customer base had two characteristics: 1. Just a few high-value customers do the vast majority of business; and 2. Customers’ needs are fairly uniform, meaning there isn’t much product differentiation in the category.' is just plain nonsense; the reason loyalty programs work is because they deliver incremental revenue increases throughout multiple customer segments (often for different reasons), and the costs are associated with performance, as customers who do not participate do not cost anything. Yes, the top 1% - 2% can really move the needle and drive data, but this simplistic formula given above isn't close to being a proper way to evaluate loyalty program effectiveness. In fact, what it reveals is a pre-determined way of working, framed to influence companies to over-value the services they offer. It's a free country, but your reputation just took a big hit in my book.