Is Customer Loyalty A Predictor Of Profitability?

[tweetmeme source=”pricingright”] Much has been said and written about the need for customer loyalty. The need to focus and attain customer loyalty is intuitively clear to all marketers. Some of the key arguments for customer loyalty include

  1. Reduced Customer Acquisition costs – Since it costs $X to acquire new customers, any customer you hold on to saved you $X. For example, it takes mobile providers $350 to acquire new customers and there are similar metrics for most products.
  2. The Loyalty Effect: Longer a customer stays longer they keep paying you. There was a book by the same name that claimed up to 75% increase in lifetime value of a customer if they stayed longer.
  3. Cross-Sell & Up-Sell: Since you keep your customers and come to know more about them it creates additional revenue opportunities through cross-sell and up-sell opportunities.
  4. Price Tolerance: Loyal customers keep buying from you because they are delighted by your product and are less sensitive to prices.  Some even claim that loyal customers do not even bother to use coupons and promotions, thereby saving you money.
  5. Decreasing Cost to Serve: The more you understand your customer’s usage behavior and needs fewer the mistakes in servicing them and hence lower the cost to serve them.
  6. Bump From Word of Mouth: Loyal customers are also your best marketers, they are happy to write online reviews and promote your products to all their friends and web communities. This means they generate additional incremental revenue.

All these factors seem plausible and the “gut feel” says these must be true.  If even a subset of these six factors are a work, customer loyalty must be a very good predictor of sales growth and profitability.

We should be able to validate the following models

Sales Growth =   Constant  +   ß1 * (Customer Loyalty)

Profitability =  Constant  + ß2 * (Customer Loyalty)

(ß1 and ß2 are the weights of  customer loyalty )

In a study published in circa 2000 in the Total Quality Management journal, researchers studied precisely these two models for a large set of products and services. The result?

Loyalty is a poor predictor of both sales growth and profitability. Their R-square values are 6% for profitability and 2% for sales growth. (For services the number goes to 14.7% and 7.8% respectively). That means only a tiny fraction of the changes in sales growth and profitability are explained by changes in customer loyalty.

Loyalty has positive impact on sales growth but more strikingly, for products, the impact on profitability is negative, which means higher the loyalty lower the profitability. This means any attempt to “buy loyalty” with price cuts does bring you loyalty but at lower profitability.

The net is, what seems too obvious isn’t so. This is not to categorically dismiss need for loyalty but the positive effects of loyalty are clearly overrated. If their effects are so low then there is a high opportunity cost to improving them. You cannot put all the  wood behind the loyalty arrow!


Correlation means two variables are associated and the extent of association si expressed as correlation coefficient. It ranges from -1 (low,high)  to +1 (high,high). A value of 0 means no correlation.

Predictability, R-square, means one variable is a predictor of other. It is measured as a square of correlation coefficient. So two variables that have a correlation coefficient of 0.8 have a predictability of only 0.64. R-square is usually expressed in %, so 64% means 64% of changes in dependent variable are explained by changes in predictor variable. That said, correlation does not mean causation. There are other factors to consider including but not limited to statistical significance of weights of variables, omitted variable bias, etc

Consumer Learning Curve – Package Sizes

Do consumers learn  from experience about package resizing? How steep is the learning curve? Once they find that they paid the same for a smaller amount or when they served themselves much more than they intended into a short and wide container, do they apply the lessons learned next time they purchase the same product? Do their learning carry over to other similar purchasing scenarios?

To answer these questions we need to do experiments on the same group of customers and see whether their buying pattern changes after the customers learn about size of the packages.

I talked about the use of short and wide containers and self serve model at a frozen yogurt chain and how it leads to over serving. I made a claim that was purely based on the consumer comments in the Yelp. Most consumers commented that they were shocked and warned about over serving. But those comments really indicate their attitude and not the actual behavior.

If we want to measure consumer learning we should follow a group of consumers and measure the mean amount of frozen yogurt they served themselves, the first, second, third fourth fifth … times. Only if we find statistically significant difference can we claim that learning occurred.

So I retract my claim that this method of price realization  will not work. The question should not be whether customers served themselves less amount the next time rather whether the fact that they paid a high price will lead the customer to avoid the store altogether.

Did you see the movie on the opening night?

This is an update of an old post, updated for iPad queues.

Once I overheard two parents talking about Obama’s election. It was clear that both were Obama supporters and happily donated money to his campaign. One asked the other, “But did you support him from the beginning”.  Then it dawned on me the existence of an implicit pecking order in the minds of  consumers of certain brands, Obama, Apple, In N Out, Star Wars.

Does it matter that the other parent supported Hillary before the primary and then decided to throw her support behind her party’s nomination?

Does it matter I did not stand in line to get iPad?

Why does it matter to one segment of customers when the other became the fan of the brand?

I think that the mystique surrounding certain brand makes  the early adopters feel that they somehow own it.  The low uptake during this phase gave them the aura of exclusivity. As in the early days of Obama or in the case of Apple. But as the brand becomes more popular and reaches the exponential growth phase the early adopters find it hard that they don’t anymore enjoy the exclusivity.

The brands needed these passionate early adopters  early on, but they nee the revenue from the large middle even more. Different needs at different stage of product lifecycle.  In the end they end up treating all of them as the same.

Hence the early adopters  tell themselves stories about being the first to support Obama before he was popular or stand in line for four seven hours to get the magical new device.

All in an attempt to find the lost mystique, to tell ourselves that we are somehow different and unique from the rest, even though we buy and use the most  popular product which is now available to all.

I am only taking the crumbs – Niche Strategy

How do you enter a market with large incumbents without being crushed by them? Market entry lesson from a hair stylist, Arrojo: (via WSJ)

WSJ: You sell and use P&G’s Wella brand in your studio and represent the brand at beauty shows. Why isn’t P&G upset about you launching your own competing brand of Arrojo product?

MR. ARROJO: I talked candidly with the P&G folks. They want market share of the beauty industry in the States, and Nick wants to keep building his business and brand. I’m a little dot compared to their business. If I sold to other salons tomorrow, [he sells primarily online, in his salon and through QVC] I’d get them upset perhaps. I’m not planning to go down that road at present. Sure I might make some more money selling shampoo, but I wouldn’t get to travel and the exposure and to be visible on stage. Don’t bite off the hand that fed you; don’t think you can rule the world.

In the rest of the interview Arrojo talks about Branding, customer loyalty and retaining customers even when their regular stylist quits. Arrojo charges $500 for a haircut, but when I read this interview I think he can pass for one of those management gurus who charge $50,000 for a speech.