As curvas de Stobachoff

Neste postal "Segmentação retrospectiva dos clientes" referi um artigo de Storbacka "Segmentation Based on Customer Profitability – Retrospective Analysis of Retail Bank Customer Bases" onde é mencionado o Stobachoff Index.
.
Uma pesquisa na net permitiu-me encontrar uma interessante referência em "The strategic value of customer profitability analysis" de Erik M Van Raaij publicado por Marketing Intelligence Planning (2005) Volume: 23, Issue: 4, Publisher: Emerald Group Publishing Limited, Pages: 372-381 (obrigado ao Baidu) e outra no livro de Robert C. Blattberg, Byung-Do Kim e Scott A. Neslin "Database Marketing - Analyzing and Managing Customers" do qual retirei o extracto que se segue para memória futura na minha base de dados:
.
"Van Raaij et al. (2003) report the first experience of a business-to-business company with incorporating customer value analysis into their marketing planning. The company, which we will call “DBM,” was a multinational firm in the market for professional cleaning products. It sold directly to end-users such as in-flight caterers and professional cleaning services, as well as through distributors. It divided its market into sectors such as healthcare, lodging, or dairy. Sales and profits had been leveling off after years of growth and DBM was worried about new competitors. Further non-product costs (e.g., costs to service customers) had been increasing. The company desired to assign these costs to individual customers and calculate customer profit.
DBM undertook a six-stage process to calculate profit at the customer level and then develop strategies based on the results:
1. Select active customers
2. Design the customer profitability calculation model
3. Calculate customer profit
4. Interpret the results
5. Develop strategies
6. Establish an infrastructure for future applications."
...
"The company used the customer profit curve to plot what they called a “Stobachoff” curve. This is simply the equivalent of a cumulative lift curve. It orders the customers according to profitability, and then plots the cumulative profit accounted for by these customers as one progresses from the highest to lowest profit.

shows that in this example, 75% of the customers are profitable (the curve increases up to about that point) while 25% are unprofitable. Given that the top 75% of customers accounts for 120% of the profits, the remaining 25% really drag profits down.
In this case there are a lot of profitable customers but they are subsidizing a relatively small number (at least a minority) of unprofitable customers. Note that by adding fixed costs through overhead, the firm may be distorting the true profitability of the remaining 25% of the customers. Some of these may be incrementally profitable." (Moi ici: E qual será o perfil na sua empresa? Basta recordar Byrnes para ficar com os cabelos em pé. E a minha experiência a trabalhar com PMEs, pouco ou nada habituadas ao conceito de clientes-alvo também não ajuda a melhorar o retrato)

"In the low dependence, low subsidizing cell, all customers are profitable and roughly equally so. In the low dependence, high subsidization cell, most customers are profitable but there are a few unprofitable customers who drag down total profits. In the high dependence, low subsidizing cell, there are only a few profitable customers and the rest of customers are unprofitable but not highly so. In perhaps the most dangerous case is the high dependence, high subsidization cell. In this case, there are a few highly profitable customers,
and many highly unprofitable customers. This is dangerous because if those few highly profitable customers should defect, the company would suddenly be losing a lot of money."