Customer Retention or Customer Migration - McKinsey on ROI

MyCustomer.com
Share this content

Perhaps most CRM programmes have, as a major part of their justification, improvements in customer retention. Personally, I've always liked to see retention within a broader context which one might call 'customer value maximisation', as customer retention will be impacted by previous inter-actions with the company. Recently, I was browsing the McKinsey Quarterly web-site (a great source of business-related articles at www.mckinseyquarterly.com) and came across an excellent article which explored this theme much more coherently than I have, and so I thought I'd share some of that thinking with you in this editorial. (Note that if you're new to the McKinsey site, you'll have to register before being able to access the detailed articles.)

I was pleased to see that I knew at least one of the authors – Tim Gokey, who I had the pleasure of working with back in the mid-90's when we were still defining what CRM was all about. Tim, and a McKinsey colleague, Stephanie Coyles, have undertaken a two-year study of 1,200 households to find out about companies in 16 industries covering, amongst others, airlines, banking, and consumer products. The study shows that improving the management of customer migration as a whole, by focusing not only on defections, but also on changes in customer spending can have as much as ten times more value than focusing on defections (customer retention) alone. Of course, this is particularly relevant in companies or industries with high frequency of customer purchases. Gokey and Coyles found at least four more factors which are of importance: the frequencies of other kinds of interactions, such as service calls; the emotional or financial importance of a purchase; the degree of differentiation among competitors' offerings; and ease of switching.

So what do the authors mean by customer migration, compared with customer retention? They propose that in addition to measuring levels of customer satisfaction, and customer retention, companies should measure customer migration – from the satisfied customers who spend more, to the downward migrators who spend less. This step is important because large amounts of value are at stake. They quote, for example, one retail bank, where 5% of current accounts customers defected annually, taking 10% of the bank's current accounts, and 3% of its total balances. But every year, the 35% of customers who reduced their balances significantly cost the bank 24% of its total balances, while the 35% who increased their balances, raised its total balances by 25%. In other words the 1% net growth covered a 'churn' in balances of 49%. This effect was not restricted to retail banking, but affected all 16 industries. For example a phone company found that 90% of its loyalty opportunities came from customers dropping features such as second lines or call waiting.

The main differences compared with a customer retention programme are two-fold: firstly, if you manage migration, you start trying to change customer behaviour much earlier in the downward path to customer attrition; and secondly, you have the opportunity to manage customers upwards, as well as downwards.

So how do we manage customer migration? Gokey and Coyle's research identifies a 6-part segmentation that again appears to apply across the 16 industries they examined. There are three 'loyal' segements, and three 'downward migrators'. If we start with the 'loyal' segments, the research highlights that customers are loyal because they are emotionally attached to their current provider, have rationally chosen it as their best option, or don't regard switching as worth the trouble. Note that this last segment ties in well with the research from the Centre for Customer Studies that we've highlighted in a previous editorial and presentation.

The downward migrators again have one of three reasons for spending less; their lifestyle has changed so they have developed new needs; they continually reassess their options and have found a better one; or they are actively dissatisfied, often due to a single bad experience. It should be clear, that there are three underlying attitudes that impact both positive and negative customer migration: emotions, inertia and deliberation. Deliberative customers are usually the largest group, representing on average 40% of all customers, perhaps not as high as one might expect. This average is, in a sense, misleading, as one of the more interesting findings of the research is that the proportion of different types of customer varies significantly by industry type. For example, deliberative customers who change their spending patterns because of factors like convenience, account for more than 70% of reduced spending by purchasers of casual apparel, but only 33% of reduced spending by mobile-phone customers. This emphasizes another point, I'm keen to push – CRM varies by industry and company, depending on the relationship customers have with particular companies.

So if we can identify these loyalty profiles in our customer base, how can we use them? Perhaps the most important point is that it encourages companies to realize that it is about understanding and managing all six loyalty segments, and that different tactics are required to manage each segment. If this is combined with a measure of customer value (both current and future) the profile helps a company prioritise its loyalty-building activities based on the value of each opportunity.

Perhaps a couple of examples from the article will help:
"One financial institution aimed all its loyalty efforts at increasing its customers' satisfaction. It did so measurably, made major investments to cut down on service failures (such as unanswered phones), and reduced the number of closed accounts. But the effect on overall growth was marginal, and the company's loyalty profile shows why: customers are spending less of their money at such financial institutions mainly because their needs are changing - for example, they might be sending their children to college. Deliberative behaviour, based on factors like prices and features, ran a close second. Here, as in most industries we studied, downward migration due to dissatisfaction represents a small proportion of the total loyalty opportunity."

The second example I know from personal involvement, as I've mentioned before on this site. In the car-hire industry, deliberators' concern with price is expanded to include other factors, particularly the time spent by a busy 'user-chooser' waiting for their car at the airport. The Hertz No 1 club has a number of features built in to save time for that user-chooser whenever they rent a car. Frequent travelers are encouraged to base their decisions about which car rental company to patronize, not only on price, but also on ability to save time.

Of course, the real bottom-line, is does this have any impact on the ROI from CRM. I can do little better than quote the authors themselves:
"By using a coordinated approach to understanding the many facets of loyalty and by tailoring tactics appropriately, several companies have reduced downward migration and defection by 20 to 30 percent. Consider the case of the retail bank that saw 50 percent of its annual balances in play because of migration. Even assuming that more than half of the change was due to factors the bank couldn't influence, and even if it managed to recapture only 20 percent of the rest, its top-line growth, currently in single digits, would rise by three to five points. Similarly, department stores using a coordinated approach have cut downward migration by 25 to 30 percent, leading to top-line gains comparable to a 1 to 2 percent increase in same-store sales, a key retailing metric generally in single digits. For banks and the many other industries with modest price-to-earnings ratios, increasing growth rates by the clearly feasible target of 3 percent should lead, within three years, to a 25 percent increase in market capitalization.
By focusing on migration and by understanding the motives that underlie customer loyalty, companies can increase that loyalty in a meaningful way."

There's some interesting stuff here, and there's more on the McKinsey site, so I encourage you to visit the original article. As always we'd like to hear your comments. Make them below or email me.

Regards,

Richard Forsyth

Replies

Please login or register to join the discussion.

avatar
02nd Apr 2002 22:25

The McKinsey article, your comments, Richard, and the two folks who responded make a very interesting read.

May I also suggest tracking down the work of Susan Fournier of Harvard Business School. In my opinion, she is doing the necessary primary research to lay the groundwork for a solid CRM model to emerge. Her work on Relationship Management is well worth a look.

Thanks (0)
avatar
02nd Apr 2002 17:36

This is indeed a useful article, and it has been making the rounds of my CRM colleagues. Beneficial as it is in offering tags and definitions for groups of at-risk customers, and giving detailed examples of retained customers that are not necessarily loyal and don't exhibit loyal behavior, the information isn't particularly new. For many years, savvy marketers and researchers have been looking at customers with attrition and risk behaviors as a definable group or life-stage, sometimes representing a precursor to turnover (such as we see in churn analysis and models), sometimes representing opportunities to rebuild customer trust and spending levels. What the McKinsey article did was break loyal/risk/potential churn behaviors into six customer segments, which is a helpful way of understanding them according to needs and supplier perceptions.

The authors seem a bit conflicted about the role, or effect, of satisfaction in understanding loyalty behavior. There has been very little demonstrated relationship between the two concepts and metrics, and the authors appeared to acknowledge this in their article. As you did in your editorial, I had also noted the paragraph which discussed how a financial institition targeted its loyalty efforts on increasing customer satisfaction, only to find that higher satisfaction levels had only marginal effect on growth. This was proven a number of years ago in a major study conducted by Juran Institute. Earlier in the article, the authors stated: "A broad measure of satisfaction can tell a company how likely customers are to defect." Then, they went on to say that satisfaction levels don't necessarily tell a company what makes customers loyal and that they must understand the other drivers of loyalty to have an influence on migration, or risk.

There are, as the authors point out - and reflective of the results of extensive work by Jan Hofmeyr and Butch Rice - several important dimensions, or varients, in loyalty levels and behaviors. These dimensions must be understood, through profiling of needs and need direction, by marketers for any CRM or customer loyalty program to be effective. As they state early on: "Differentiating and measuring degrees of loyalty is an evolving craft." How well we all know.

Michael

Thanks (0)
avatar
08th Apr 2002 10:47

Raymond,

Thanks for the pointer to Susan Fournier's work, Raymond. Do you (or anyone else) have any online links to her work??

Best wishes,

Richard Forsyth

Thanks (0)
avatar
By Shilpaa
05th Mar 2003 21:31

I am a mature student currently doing my dissertation. I am wriiting on CRM and ERP Integration. I read in a book that Segmentation is now an old strategy and it emphasised on the 3D's model od Discovery, Dialogue and Discipline. I wonder is segmentation still the strategy being used.
Also is segmentation used for capturing the right customer, customer value or the ROI.

I hope someone will be able to clarify this for me. Thanks

Thanks (0)
avatar
By admin
05th Apr 2002 04:45

The most important information from this McKinsey article is that the current definition of Customer loyalty is no longer sufficient to describe customer relationship. Among these 6 loyalty segments, only the Emotive Loyalists could be claimed as the real loyal customers: most “loyal” customers are not loyal anyway.

In addition, if we look at the customer relationship from our customers’ standpoint, Customer Loyalty is definitely not the relationship they think they have with companies. By claiming Customer Loyalty, companies think they OWN their loyal customers. However, when customers think about their relationship with companies, they often think in the opposite way, that is: they OWN those companies. Listen how customers express their relationship with their service providers: MY bank, MY barber, MY doctor, MY favorite coke, etc. Of course, usually you only hear a “Loyal” customer making such ownership claims.

What does this mean? It simply means that the conventional concept of Customer Loyalty describes only how companies think about customer relationship. But our customers never think about their relationship with us in the same way as we think. If we don’t understand how our customers think about our customer relationship, we are unlikely to properly manage our customer relationship.

A further issue about Customer Loyalty is that we hardly investigate it under market competitions. We have accepted this fact from market competitions: competing companies are sharing customers. That is, most customers have relationship simultaneously with competing companies in a market. We describe such situation as “share of wallet” or “share of customer”. It actually describes a situation of “Share of Loyalty”. A wallet could be shared, a customer also could be shared. But Loyalty could not. When Loyalty is shared, Loyalty losses all its meaning.

Customer Loyalty is no longer a marketing concept that could be used to properly describe the customer relationship (or at least most customer relationship) under market competitions.


Regards,

Tim Lee
CMO Consulting International
http://www.WebCMO.com

Thanks (0)
avatar
By Shilpaa
05th Mar 2003 21:31

I am a mature student currently doing my dissertation. I am wriiting on CRM and ERP Integration. I read in a book that Segmentation is now an old strategy and it emphasised on the 3D's model od Discovery, Dialogue and Discipline. I wonder is segmentation still the strategy being used.
Also is segmentation used for capturing the right customer, customer value or the ROI.

I hope someone will be able to clarify this for me. Thanks

Thanks (0)
avatar
By admin
05th Apr 2002 04:45

The most important information from this McKinsey article is that the current definition of Customer loyalty is no longer sufficient to describe customer relationship. Among these 6 loyalty segments, only the Emotive Loyalists could be claimed as the real loyal customers: most “loyal” customers are not loyal anyway.

In addition, if we look at the customer relationship from our customers’ standpoint, Customer Loyalty is definitely not the relationship they think they have with companies. By claiming Customer Loyalty, companies think they OWN their loyal customers. However, when customers think about their relationship with companies, they often think in the opposite way, that is: they OWN those companies. Listen how customers express their relationship with their service providers: MY bank, MY barber, MY doctor, MY favorite coke, etc. Of course, usually you only hear a “Loyal” customer making such ownership claims.

What does this mean? It simply means that the conventional concept of Customer Loyalty describes only how companies think about customer relationship. But our customers never think about their relationship with us in the same way as we think. If we don’t understand how our customers think about our customer relationship, we are unlikely to properly manage our customer relationship.

A further issue about Customer Loyalty is that we hardly investigate it under market competitions. We have accepted this fact from market competitions: competing companies are sharing customers. That is, most customers have relationship simultaneously with competing companies in a market. We describe such situation as “share of wallet” or “share of customer”. It actually describes a situation of “Share of Loyalty”. A wallet could be shared, a customer also could be shared. But Loyalty could not. When Loyalty is shared, Loyalty losses all its meaning.

Customer Loyalty is no longer a marketing concept that could be used to properly describe the customer relationship (or at least most customer relationship) under market competitions.


Regards,

Tim Lee
CMO Consulting International
http://www.WebCMO.com

Thanks (0)
avatar
02nd Apr 2002 22:25

The McKinsey article, your comments, Richard, and the two folks who responded make a very interesting read.

May I also suggest tracking down the work of Susan Fournier of Harvard Business School. In my opinion, she is doing the necessary primary research to lay the groundwork for a solid CRM model to emerge. Her work on Relationship Management is well worth a look.

Thanks (0)
avatar
08th Apr 2002 10:47

Raymond,

Thanks for the pointer to Susan Fournier's work, Raymond. Do you (or anyone else) have any online links to her work??

Best wishes,

Richard Forsyth

Thanks (0)
avatar
02nd Apr 2002 17:36

This is indeed a useful article, and it has been making the rounds of my CRM colleagues. Beneficial as it is in offering tags and definitions for groups of at-risk customers, and giving detailed examples of retained customers that are not necessarily loyal and don't exhibit loyal behavior, the information isn't particularly new. For many years, savvy marketers and researchers have been looking at customers with attrition and risk behaviors as a definable group or life-stage, sometimes representing a precursor to turnover (such as we see in churn analysis and models), sometimes representing opportunities to rebuild customer trust and spending levels. What the McKinsey article did was break loyal/risk/potential churn behaviors into six customer segments, which is a helpful way of understanding them according to needs and supplier perceptions.

The authors seem a bit conflicted about the role, or effect, of satisfaction in understanding loyalty behavior. There has been very little demonstrated relationship between the two concepts and metrics, and the authors appeared to acknowledge this in their article. As you did in your editorial, I had also noted the paragraph which discussed how a financial institition targeted its loyalty efforts on increasing customer satisfaction, only to find that higher satisfaction levels had only marginal effect on growth. This was proven a number of years ago in a major study conducted by Juran Institute. Earlier in the article, the authors stated: "A broad measure of satisfaction can tell a company how likely customers are to defect." Then, they went on to say that satisfaction levels don't necessarily tell a company what makes customers loyal and that they must understand the other drivers of loyalty to have an influence on migration, or risk.

There are, as the authors point out - and reflective of the results of extensive work by Jan Hofmeyr and Butch Rice - several important dimensions, or varients, in loyalty levels and behaviors. These dimensions must be understood, through profiling of needs and need direction, by marketers for any CRM or customer loyalty program to be effective. As they state early on: "Differentiating and measuring degrees of loyalty is an evolving craft." How well we all know.

Michael

Thanks (0)
avatar
02nd Apr 2002 20:14

I'm sure the McKinsey work helps us to think through the complexity of behaviours which our customers exhibit. It reminded me to re-read three seminal HBR articles which hit a number of these points a few years ago:
1 Achieve your customers’ full potential
Grant & Schlesinger, Oct 95
2 Why satisfied customers defect
Jones & Sasser, Dec 95
3 Learning from customer defections
Reichheld Mar 96

These articles hit all the main issues including partial defections (varying spend trends) and variation of satisfaction / loyalty by industry.

The bit I particularly like is "getting the right people to learn the right lessons" from the last of these three articles. It advocates the need for the board to listen to what the customers are telling them and to push harder to understand the root causes of customers issues and wants. Something my friend Bill Price, erstwhile of Amazon, strongly advocates to drive out counter productive contact with customers.

When consultants research the customer base of client companies, I always wonder why the company itself isn't doing that all the time - its so evidently valuable and avoids those "fact-free" board room debates. The trouble is they usualy are but not listening to what the customers are saying.

And if the big research budget isn't there, talk to the frontline staff - they can nearly always tell you what the research will say - ok, with some emotional bias !

Putting in this kind of "analytics and decisioning" into your organisation takes bravery, listening and acting on the results is sometimes, unfortunately, unpopular with colleagues but very popular with customers !

Thanks (0)