<B>Customers: angels or devils?</B>

Eduardo Navarro. Managing Partner. Improven Consultores

31 May 2004

It is natural to treat customers alike, as if they were all the same. But this is a big mistake, the author argues, for some customers are angels and others are devils.

The angels offer high profit margins with relatively low commercial costs and high consumption.

The devils purchase little, give us lots of problems and bargain right down to the last cent. Usually, although we have devils with whom we lose money, the angels compensate, and in the end our profit-and-loss account comes through in the black. But what would happen if we were to turn all our devils into angels? What effect would it have on results?

Of course, the answer depends on each particular case. Even so, the effect is significant, because by simply rethinking the company’s customer strategy we could greatly improve results. In a practical case of one of our customers, we discovered the following data (See Figure 1):

Number customers<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

% total turnover

% contribution to profits
















Here we clearly see that the highest contribution to the company’s profits (and a good part of turnover) comes from 12 customers (approximately 3.5 percent of the total number of customers); whereas 301 customers with a lower turnover reduce profitability (9.2 percent).

Is it logical to dedicate the same resources to the 12 customers that represent almost 40 percent of turnover as to the 384 that represent 19 percent? If we consider the company’s structure - and this is more apparent with an activity-based cost system – we find many processes and subprocesses that are independent from the size of the order or the customer.

This means they are highly profitable for large orders/customers and not very profitable otherwise. Examples of said subprocesses from a commercial standpoint include some internal logistics costs or sales administration costs.

Segmentation can be simple

If we agree that almost all the elements in our organization - from product development to price, and including distribution or competitive positioning - depend to a greater or lesser extent on our customers and on the level at which we manage to cover their expectations, then why not make greater efforts to discover and separate our customers into segments?

Of course, it is often complicated to manage customers as individuals, and even moreso when your customer list reaches 10,000 - particularly with retail trade. This is where segmentation comes in. Segmentation consists of grouping customers together according to their needs or the characteristics that condition their purchases. Segmentation can be very simple and based on aspects such as turnover, geographical area or distribution channel; or it can be more or less complex and based on customer behavior. It is common practice to make segmentations as simple as the one in the following table:

Customer type

Potential turnover range


> 100,000?


> 30,000? and < 100,000?


> 6.000 ? and < 30,000?


< 6,000?

Although it is a good starting point, this profile is inefficient in an environment as complex as today’s. For example, with a customer in the chemical sector, we define a segmentation model based on customer needs as follows:

· Those who look for convenience: Buyers who purchase from different suppliers and whose main concern is service;

· Price mercenaries: Buyers whose main motivation is price. They can be large customers with a high purchasing frequency, but at the same time can have high loss levels due to appearance of a competitor with a better price. This customer is particularly interesting for organizations with cost differentiation strategies;

· Brand buyers: Buyers for whom the brand name is the main reason for their purchase.

· Quality essential: Buyers for whom quality is the basic parameter;

· Those who look for relations: Customers with few suppliers and a high level of relations and loyalty. This type of customer usually has a high acquisition cost but a low rate of loss.

Although the segmentation model defined above can be used as a starting point, the specific segments must be identified for each particular case. For a segmentation model to be valid, other parameters must also be taken into account: customer profitability, customer satisfaction, purchasing frequency, and average order size. It can also be crossed with a customer scoring system, such as RFM (Recency Frequency Monetary).

An element as important as sensitivity to price - and therefore profitability - is directly related to the segment (this case is special and as such extrapolation will not always be direct):


Those who look for convenience

Price mercenaries

Brand buyers

Quality essential

Those who look for relations

Sensitivity to price











Segmentation is useful for other elements, such as customized direct marketing, allocation of resources, customer relations management and the proposal of value for the customer. We all have a great number of angels among our customers, but we also have devils.

Consequently, having an appropriate methodology implanted to identify and manage them accordingly will greatly help the company's profitability.


#IECampus, the Campus of the Future

See video
Follow us
IE Agenda
Most read
IE Business School | María de Molina 11, 28006 Madrid | Tel. +34 91 568 96 00 | e-mail: info@ie.edu


IE Business School

María de Molina, 11. 28006 Madrid

Tel. +34 915 689 600