Impact of the Customer Relationship Management on the Performance of the Institutions of Microfinance in Cameroon: The Case of the CAMCCUL Network

The objective of this contribution is to work out the customer relationship management and the commercial performance of the institutions of microfinance (IMF) in Cameroon in order to show its impacts on the performance indicators of the latter. To achieve this goal, we have simultaneously applied the nonparametric method of the Data Envelopment Analysis (DEA) and the censored Tobit model on a sample of 60 IMF belonging to the Cameroon Cooperative Credit Union League (CAMCCUL) network. The different statistical tests carried out have first of all permitted us to estimate the IMF performance levels and then to assess the different factors which influence the commercial performance of the IMF, taking into account the indicators obtained from the literature review. As a general rule, we have realized from our results that the IMF of the CAMCCUL network are globally inefficient as regards the commercial performance, the women influence on the scores of performance being very significant, insofar as the IMF which are in majority made up of women have relatively been more impressive compared to those which are in majority made up of men. Among the factors influencing the performance, we have noted that the computerized management of the customer relationship and the proximity (closeness) of the social contacts exert a positive influence on the performance, while the satisfaction towards the past experiences with the IMF exerts a negative and very significant influence on this one.


INTRODUCTION
The economic expansion of all the nations generally depends on the health and the dynamism of their financial system. The strong development of this sector as well as the numerous innovations in the financial markets during the last two decades are based on the privatization of the key sectors, combined with the process of liberalization of the capital movement and on the accelerations of the technological process linked to the communication and to the transmission of information in real time (Pereira, 2011). This phenomenon, more developed in the developed capitalist economies (Waofo, 2011) has ended in Cameroon from the outset of the 1990s by the hardening of the access conditions to credit, and this is due to the restructuring of banks, following the failure of the structural adjustment programmers (SAP) and the public development aid (Ckouekam, 2008). All these measures have contributed to limit (restrict) more the access to the banking services to the underprivileged sections of the population (the poor). In the concern to correct this imbalance (disequilibrium) and owing to the laws n o 90/053 of the 19 th December 1990 on the freedom of association, and n o 92/006 of the 14 th August 1992 relative to cooperative firms and to the common initiative groups, the microfinance is going to develop and experience a real (genuine) opening in Cameroon (Ngnodjom, 2005). The boom of this new financial industry has been at the origin of the unprecedented disruption of the financial services section, which has seen the hardening of the competition conditions through the threat of the potential competitors who are becoming more and more numerous. Parallel to this arrival of new competitors, the institutions of microfinance are concentrated 1 . The market seems to be saturated; the rate of the proportion of the population having a bank account is at its maximum level and the customer does not hesitate to change bank if he/she is unhappy, or to multiply the number of his/her accounts (Des Garet, 2000). This being, the taking into account of the customer has therefore evolved and his/her satisfaction has then become an important aspect of business. From a product vision, the firm has developed a customer vision. In the perspective of protecting their market shares and after the observation that the development of customer loyalty costs cheaper to the enterprise than the search for new clients, the development of customer loyalty through a good management of its clientele (customers) has increased in scale for, the consumer is today at the center of the strategy of the organizational development of the firms; and that is why in future it becomes necessary for the IMF to invest in a customer relationship of quality, aiming at targeting, attracting and conserving good customers in order to maximize their profits. This is what is called the customer relationship management (CRM). The marketing literature review relating to the relational and transactional approaches attempts to apprehend the development and the amelioration of the customer relationship as source of competitiveness or the gap between this virtue stemming from the exploitation of the relationship and the limit of its treatment in the works justify the problem of the customer relationship management in the firms in general and could find its judiciousness in the Cameroon microfinance institutions in particular. 1 That is to say that they increase in proportion in the market; they are saturated.
As a matter of fact, the IMF today looks more and more for addressing the following issues: how to develop customers' loyalty? What tools should be used? How to develop a profitable and durable customer relationship? (Simonet, 2011). These questions (interrogations) which raise the sensitive problem of the banking marketing could be justified in the environment of the IMF 2 which the number 3 has considerably increased these last years in Cameroon. Moreover, this sector equally experiences a strong competition of the banking sector which operates in an environment impelled (sustained) by a high presence of subsidiary companies of foreign banks (banks of French and English origin) and of the local banks (with national capital). These banking institutions propose offers which are more or less similar and are characterized by a marked proximity with the customers. This bitter competition in the market of the IMF of Cameroon brings up at once the problem of the efficient and rational management of the customer relationship. Furthermore, in spite of the fact that it has been observed that more and more Cameroon firms and precisely the IMF operating in an environment highly influenced by the sophisticated technologies of the information and communication, develop and put in place policies and strategies in connection with the CRM, no study has been yet carried out as to its impact on their commercial performance in the context of Cameroon. Nevertheless, such a study could be of great interests. Therefore, we find it interesting to carry out a study on the following topic: impact of the customer relationship management on the commercial performance of the institutions of microfinance in Cameroon: the case of the IMF of the CAMCCUL network. 2 The environment of the IMF experiences an unprecedented development. Since the liberalization of the financial sector, it has gathered progressively a great number of banking customers (see Simonet's works, 2011; and those of Bekolo and Onomo, 2008) 3 The survey of the Banking Commission of the Central Africa (BCCA) of 2000 has made a census of 652 IMF on the Cameroon territory.
The present article is organized as follows: the first section presents the panorama of the microfinance in Cameroon and the measure of the performance. The section two describes the methodology used and the sources of data. The third section presents the results as well as the suggestions aiming at the improvement of the supply of microfinance services.

Panorama of the Microfinance in Cameroon
According to a study carried out by Creusot (2006), the microfinance under its traditional form (tontine) dates back to more than a century in Cameroon. It has started under the formal form in 1963 with the creation of the first cooperative of saving and credit (« Credit Union » or people's fund) in the English-speaking zone of Cameroon through the impetus given by the Dutch missionaries (these cooperatives are today gathered within the Cameroon Cooperative Credit Union League -CamCCUL-, the biggest network of the IMF of Cameroon. This network has even created a commercial bank for 5 years now: the Union Bank of Cameroon). But the microfinance has not experienced a remarkable expansion and has only been diversified from the outset of the 1990s thanks to the laws n o 90/053 of the 19 th December 1990 on the freedom of association, and n o 92/006 of the 14 th August 1992 relative to cooperative firms (companies) and to the common initiative groups. We should also indicate here that the crisis of this sector at the end of the 1980s and the restructuring of this sector which has followed have led to the liquidation of several banks, the closing of almost all the banks counters in the rural zones and the small towns, and the dismissal of many banks executives. The last-mentioned are going to turn to a new type of employment creating many cooperatives of saving and credit operating or trying to operate as quasi-banks. the institutions which only give credit such as ACEP Cameroon (which the main agencies were at the origin essentially concentrated in the Yaoundé and Douala towns) and the Cameroon Gatsby Trust (CGT); -A great number of projects of development or farm-produce industry with a credit section such as the Company of Development of Cotton (CDC), the South-West Development Authority (SOWEDA), PREPAFEN, etc. This great infatuation for the microfinance is accompanied with a crisis essentially on the cooperative sector, due to the lack of professionalism and to the absence of control. This is going to lead the monetary authority (Ministry of Finances) to take the control of this sector (which up to here was under the exclusive supervision of the ministry of Agriculture), through the decree of the Prime Minister in 1998, which in future subject all the cooperatives to the approval regime and to the control of the ministry of finances. In order to include all the forms of institutions of microfinance and to reinforce the control and the supervision of the sector, a subregional text is going to be adopted by the council of ministers of finances of the Monetary and Economic Community of Central Africa (MECCA), and signed on the 13 th April 2002. This text has come into effect on the 14 th April 2005.

Role of the institutions of microfinance
The activity of the IMF has permitted a genuine recovery of spaces of independence and of the economic expression by a progressive and polar appropriation of financing means. The contribution of the IMF is determining in the support and the integration in the economic activities of the basic operators and of the informal sector. To illustrate these words, the activities of IMF in 2006 are as follows: The 520,000 customers estimated here are often community interest groups (CIG) or small firms employing few persons. The concerned population is actually of the order of 2 to 3 million of persons, each contributing to the satisfaction of the needs of a family. Without being exhaustive, the contribution of the IMF in the Cameroon economy is essential in the following sections:  the creation of thousands of directs and indirect jobs ;  the financing of the small economy ;  the adaptation, to make them popular, of the products of the mass proportion of the population having a bank account ;  the creation of new guarantees ;  The obtaining through new alliances and partnerships of additional resources for the underprivileged strata of incomes.

Notion of performance
The notion of performance occupies a central place in the daily and economic activities. It is often used as means of control or of animation either of an economic entity, or of a financial and economic system as a whole. Nevertheless, this concept gives rise to two categories of interrogations which are thrown to the researchers: what is performance? How to measure it?
The notion of performance is ambiguous (Pesqueux, 1996) and is often perceived as "any result obtained, measured or estimated" (Famose, 1996). The Larousse dictionary (2004) defines it as: "the optimal possibility in encoded data of a material". In a purely economic universe (world), the problems of definition of the performance are at the origin of the difficulties of its evaluation. As such, De La Villarmois (1999) thinks that before tackling the concept of performance, it would be proper to investigate those of effectiveness, of efficiency, of productivity and finally of performance. As a matter of fact, this will permit us to show the representations of the performance and to give details of the implications of the complexity of the concept on the processes of evaluation.

Distinction effectiveness/efficiency
Though having a precisely significance for the economists, the terms effectiveness and efficiency are often used indiscriminately. The effectiveness is often perceived as the capacity to attain an objective. The idea of effectiveness of a unit of production was introduced for the first time by Farell (1957), under the concept of "input oriented measure". According to Farell (1957), a measure of effectiveness is defined as the maximum of equiproportional reduction of all the inputs which permit a continuous production of a given output. The physical and technical effectiveness has to do with the possibility of avoiding the waste in producing as much output as the use of input allows it "input oriented measure", or by using as little inputs as it is required by the production of output "input oriented measure". This effectiveness is measured by comparing the optimal and observed values of the production, the cost, the income, the profit or of all that the production is capable to seek as objective and which is subjected to the constraint of price and suitable quantity. As to the efficiency, it can be defined economically by the ratio output/input. This being said, the increase of efficiency assumes the maximization of the use of resources which passes through the increase of production without any increase of costs, or the delivery of a level of production or of a given service by reducing the factorial endowment (Desreumaux, 1992). As to the banks, the efficiency of a marketing network can be measured by the ratio turnover/the sales force staff. In this manner, the increase of efficiency goes through the increase of the turnover and the reduction of the sales force. This implies the increase of the productivity of the sales force. It appears without beating about the bush that the productivity is an indicator of the efficiency. The terms effectiveness and performance are often used indistinctly. They cover various and particular domains.

The performance indicators of the IMF sector
Among the indicators generally used in the field of microfinance, we can distinguish the following: -the scope (level) indicators: these are the indicators which permit to the institution to measure at the same time the extent of its activities (number of customers served with different types of products) and the degree of the scope (type of customers reached and the degree of poverty of the customers); -the portfolio indicators: the quality ratios of the portfolio give information on the percentage of the unproductive assets which reduce the incomes of the institution and have a negative impact on its treasury situation; - The productivity indicators: the productivity ratios measure the volume of generated activities (output) for a resource or a given asset (input). The productivity is often measured together with the effectiveness which gives some information on the cost per unit produced; -the profitability indicators: the profitability ratios measure the net result of the institution with respect to the measure of its balance sheet; - The viability indicators: these ratios measure the capacity of the institution to cover its charges (expenses) with the products that it generates.

Population of the study and source of data
Our study deals with the population of the IMF affiliated members of the CAMCCUL network, including as a whole 191 IMF. This choice is justified by the fact that this network is the most important one, the oldest and the better organized one as well at the national level as at the sub regional level of the Monetary and Economic Community of Central Africa (MECCA). At the national level, this network covers the extent of the Cameroon territory, with IMF located in towns and in rural zones. It represents about 76% of the deposits and 74% of the credit granted by the microfinance sector in Cameroon (Kobou, Moungou and NGOA, 2009). Our sample is made up of 60 IMF, selected on the basis of the accessibility of the financial and book-keeping information, necessary to the operationalization of inputs and the outputs. For the same reasons, we choose the 2010 financial year. The data that we have used are at the same time primary and secondary data, collected simultaneously to the general direction of CAMCCUL (located in Bamenda), and then completed by the heads of the agencies (or the persons in charge of the agencies). They are qualitative and quantitative in nature.

The adopted method of sampling and the collection of the primary data
The sampling of appropriateness is the one where "all the individuals who have been contacted in function of the sought characteristic are appealed". In other words, it indicates the selected persons in function of the only opportunities which are presented to the researchers without any criterion of choice being defined at the very outset. These opportunities depend on the availability and the accessibility of the heads and the customers of the said IMF. The choice of a sample of suitability in Marketing meets the concern for reducing the bias of the selection effect which depends on the representation of the sample with respect to the population studied. In this manner, the profile of the sample depends on the field chosen on the one hand and on the persons who would have accepted to participate to the survey on the basis of voluntary participation on the other hand.

Procedure of the collection of the primary data
Among all these techniques of data collection, we set our heart on the questionnaire in view of the obtaining of information relating to the satisfaction of the customers, to the computerized management of the customer relationship as well as to the influence of proximity of contacts between the IMF and their customers and their impact on the commercial performance of the latter.

Contents of the questionnaire
Within the framework of our study, the questions are closed in majority. This has as main objective to make easier their analysis. These questionnaires are essentially meant for the persons in charge of the agencies of the IMF of the CAMCCUL network on the one hand, and for their respective customers on the other hand. These questionnaires are organized as follows:  the questionnaire sent to the persons in charge of the agencies is essentially based on the variables related to the concepts of our hypotheses. It subdivided in three parts which are given below: -identification of the company; -the information on the social relations existing between the IMF and their customers; -the information relative to the computerized management of the customers relationship;  as to the questionnaire sent to the customers of these IMF, it has the following elements: -the identification of the customer; -the information relating to the satisfaction of the customers; -the information relative to the social relations existing between the customers and their IMF.

Specification of the variables of interest and the methodology of the data analysis


Choice of the dependent variables
The main explained variable is the commercial performance of IMF. This performance rests on several criteria which are: -the productivity: it is determined through a ratio which measures the volume of activities (output) for a resource or a given asset (input); -the profitability: it measures the net result of the institution with respect to its statement of accounts (balance sheet); -the scope criteria: they permit to the IMF to measure at once the extent of their activities (number of customers served with different types of products) and the degree of scope (type of customers reached and the degree of poverty of the customers); -the portfolio criteria: the quality ratios of the portfolio give information on the percentage of the unproductive assets which reduce the incomes of the institution and have a negative impact on its treasury situation;  Choice of the independent variables As to the CRM variables which are the explanatory variables, we are going to use the following ones: -the satisfaction of the customers with regard to the past experiences; -the computerized management of the customer relationship; -the closeness of the contacts between the IMF and their customers. Globally, it is a question of variables stemming from the operationalization of the concept of the CRM defined previously. We are then going to express our dependent variable in function of the explanatory variables according to the following model: P = f (satisfaction with regard to the past experiences; the computerized management of the customer relationship; the proximity of the contacts between the IMF and the customers).

Method of data analysis
Method and tool of the estimation of the IMF performance levels: the DEA approach To estimate the levels of performance, we use a nonparametric method, the Envelopment Analysis (DEA) method. Two main approaches are generally used to measure the effectiveness of a production unit. It is a matter of the method of productive effectiveness, based on the relation between the principal and the agent, and the method of the productive effectiveness based on the production frontier (Kobou, Moungou and Ngoa, 2009). This approach which interests us here is subdivided into two great methods which are: the parametric method and the nonparametric method. If we consider the parametric method, the IMF can be considered as a unit of production which has as objective to maximize its outputs (interests received size of the portfolio or the amount of deposits and the total amount of loans) for a given level of inputs. This behavior of the IMF can be modeled from their estimated production function, which production function permits to assess the performance of an IMF (Cohn et Cooper, 1997). The limit of this approach is that it is used when the functional form of the production function of a good or a service is known. But the production function of an IMF is not at the very outset known (Domazlicky and Primont, 2006). For these reasons, the nonparametric method will be used to determine the levels of performance of the IMF. The nonparametric method estimates the levels of performance of a production unit from the distance function (KirJavainem and Loikkanen, 1998). We are going to use that method known under the name of Data Envelopment Analysis (DEA) due to its empirical character, but also because it is generally recommended when the functional form of the firm is not known, or when the firm produces several outputs. According to the DEA approach, the frontier is constructed by the technique of the linear programming. The term "envelopment" is used to indicate the hypothesis according to which the production frontier envelops all the observations. The DEA method estimates the relative effectiveness of the comparable production units and generates the levels of effectiveness from the information on the inputs and the outputs of the companies 5 ; in this manner, each unit is considered as a decisionmaking unit called DMU which transforms the inputs into outputs. The DMU (j) (j=1,2,…, m) consumes an amount X{ij} of inputs i (i=1,2,…,k) and produces an amount Y{rj} of outputs (r=1,2,…,s). The efficient units are those which assure the better service while using the minimum of resources. In this case the score of efficiency is equal to 1. The efficiency frontier is made up of unit displaying the scores equal to 1 and comprised between 0 and 1 for the others.
The DEA method has the particularity of presenting two approaches: the inputs oriented approach and the outputs oriented approach. These two approaches are based on two main models: the constant returns to scale model (CRS) and the variable returns to scale model (VRS).

The CRS model
We talk of constant returns to scale when a proportional increase of all the production factors leads to an increase of the production in the same proportion. Here, for each unit DMU k, the form "ratio" of the DEA means to maximize the ratio h k in the presence of r outputs and of i inputs such as: Where r  stands for the weighting coefficient for each output r.
We finally obtain the following formulated programmer:

The VRS model
Contrary to the first model, this model has a particularity. As a matter of fact, the returns to scale are variable if following a proportional increase of all the production factors, the production changes in a different proportion. The program as follows: The input oriented CCR models of Charnes, Cooper and Rhodes (1978) will be used. It is a constant return to scale model. The choice of this model results from the problem stated. For an IMF, we would like to know what inputs should be reduced and in what quantity to increase the performance of the said IMF while keeping the level of outputs unchanged.
The DEAP 2.1 software is used. The inputs and the outputs chosen result from the data obtained from the survey carried out. The table 3 has all the inputs and outputs used in the model.

Factors influencing the performance of the IMF
The DEA method integrates only the discretionary variables, that is to say, those which can be handled per unit of production, and does not take into consideration the environmental variables, also called nondiscretionary variables (Afonso and Aubyn, 2006). However, the differences and the socioeconomic realities of the residence milieu of the IMF can play a central role in the determination of the heterogeneity between the IMF and therefore in the capacity of the IMF to collect resources and to grant loans to the poor (Kobou, Moungou and Ngoa, 2009). In part, the socioeconomic factors and even the psychological ones of the localities of the IMF determine their efficiency. These exogenous and endogenous psycho-socioeconomic factors can for instance include the variables such as the satisfaction vis-à-vis the services rendered by the IMF, the development of the emotional relations with the customers, the follow up of the customers via suitable tools. It is necessary to examine the influence of these factors on the commercial performance of the IMF. The level of performance of an IMF takes the values in the interval  0; 1. We cannot therefore estimate this equation using the ordinary least squares (OLS), but through the censored models such as the generalized and censored Poisson model or the censored Tobit model. As a matter of fact, the censored regression models are particularly recommended when the values of the endogenous variables belong to a precise interval. We say that they are censored. For the case of the factors influencing the IMF performance, the dependent variable "performance" is continuous in the interval  0; 1 (Kobou, Moungou and Ngoa, 2006). In these conditions, the censored Tobit model is not appropriate for the dependent variable does not take the value zero (Maddala, 1983;Greene, 1995). Similarly, the generalized Poisson model cannot be used, for the values of the dependent variables are not natural integers. To circumvent this difficulty, we are rather going to explain the nonperformance, that is to say the ineffectiveness of the IMF, having recourse to the censored Tobit model, since the level of performance of the IMF takes the values zero and positive and are continuous in the interval  0 1. In general, the Tobit model is used when we are in the presence of a great number of observations for which the value of the endogenous variable is zero. In the individual case, the level of nonperformance or of ineffectiveness of the IMF takes the values in the interval  0 1 . The dependent variable will be censored  the vector of the parameters to be estimated; * i Y a latent variable which can be considered as a threshold from which the non-discretionary variables affect the performance of an IMF. In our context, the dependent variable "nonperformance" is continuous and limited at zero. In assuming that the errors are normally distributed, the estimation of the model will consist in the maximization of the logarithm of the likelihood which is written below as follows: where n stands for the number of observations and  for the standard deviation. The variables retained here are those which permit better to apprehend the socioeconomic and psychological realities of the IMF environment, listed above. These variables used in the model are presented in the Table 2:

Interpretation of the results of the estimation of the performance of the IMF through the DEA method
The technical effectiveness levels of the IMF as well as the average level of effectiveness of all the production units are generated with the help of the DEAP software, which estimates the effectiveness of the production units when we have some information on the inputs and outputs. It therefore emerges from these results that the IMF of the CAMCCUL network are globally inefficient as far as the commercial performance is concerned. There are those which the score of efficiency is equal to 1. It is however interesting to bring out the common features between these IMF. In reality, all these IMF are of the first category 6 and are interested as such in a well determined social objective. The percentage of women in borrowings is the highest in these IMF than in the other ones which are less efficient. These IMF are among other structures having a modest size of portfolio, but which the number of employees stands around at seven persons; moreover, the quasi-totality of IMF which are in majority constituted of women are relatively more efficient in comparison with those which members are men in majority. This result is not surprising for, generally, women show more solidarity and use better their savings in comparison with men. They are also the main actress as far as the activities productive of small scale of incomes are concerned (commerce of proximity, micro services, agriculture, breeding, transformation of the agricultural products such as the transformation of cassava into tapioca, etc.)

Interpretation of the results of the factors influencing the IMF performance
After having obtained the score of the DEA model, the next step has consisted in the regressing these scores on the variables susceptible to influence the performance of the IMF; this is done through the censored Tobit model. To achieve this, we have used the Stata software, version 10. The estimation of the determinants of the IMF performance gives the results which are summarized in table 6. For the interpretation of the obtained results, we are going to consider that a negative parameter indicates a negative influence of this parameter on the nonperformance, and as an indirect result, a positive influence on the performance level of the IMF. The estimation model of the performance indexes is globally significant at 1%, for the simple reason that prob>chi2 = 0,000 < 1%. The factors which influence the performance of the IMF are those which the coefficients are significantly different from zero (at 1; 5 and 10%). 6 It is a question of a category of IMF which can only deal with their members. They are cooperatives, associations, etc.
These factors are: SATIF (satisfaction with regard to the past experiences); ROI (relation via the computer tool or the computerized management of the customer relationship) and at last, RSCE (closeness of the social contacts).  For the factor SATIF, its coefficient is equal to 0.025 which is greater than zero and we also have P>t = 0,040 < 5%; this means that the coefficient of that variable is positive and statistically significant at 5% level of significance. The positive sign of the coefficient indicates the fact that the variable SATIF has a positive impact on the nonperformance, and consequently, a negative impact on the performance.  For the factor ROI, its coefficient is equal to -0.005 which is less than zero and we also have P>t = 0,007 < 1%. This means that the coefficient of this variable is negative and statistically significant at 1% level of significance. This variable has a negative influence on the nonperformance and on the other hand has a positive impact on the performance. A 1 unit increase of ROI would lead to a 0.005 increase in the performance.


As to RSCE, its coefficient is equal to -0.051 and P<t = 0,047 < 5%. This signifies that RSCE has a positive impact on the performance. Similarly, an increase of 1unit of RSCE would lead to an increase of the performance by 0.051 units. This result can be justified by three good reasons: -the computerized management of the customer relationship (ROI) offers essential functionalities to the constitution of an integrated and unique view of the customer portfolio, from the first marketing contact, then throughout the commercial process, and last up to the after sale phase. It then helps the IMF and its collaborators to gain in performance, through the improvement of the knowledge of the customer (needs and centers of interests); the development of customer loyalty (supply of a personalized customer service and fitted to his/her expectations) and the dynamism of sales through the principle of the practice of targeted supply. -Moreover, in a microcosm victim of a rough and unprecedented competition, the primordial mission of the IMF does not only reside in the conquest of customers, but especially and precisely in the development of customer loyalty, which constitutes an inescapable guarantee, necessary to the preservation of the market shares. All this leads to a change of strategy for the decision-makers. As a matter of fact, to preserve their competitive advantage, the decision-makers should have at their disposal a relevant database on their customers in order to be better differentiated by the creation and the upholding of a personalized contact; this can constitute a prerequisite to the building of stable and durable relations with the customers. -Moreover, inasmuch as the loss of a customer is prejudicial and that his/her conquest costs five to eight times more expensive to the company, the development of customer loyalty passes through a processes of satisfaction and through a segmented approach which can be easily put in place through the software of the customer relationship management. These software bring not only some pertinent information to the company (IMF), but also, are creators of values, insofar as they permit to reduce the costs, to create a relation of differentiation in adapting the supply to the customer life cycle and by proposing privileged offers (gold customers for example). It also permits to anticipate the depreciation of the portfolio; for thanks to the history of the customer, it is possible to foresee the infidelities, to analyze the purchasing behavior and to deduce the expectations; it also permits to arbitrate the economic ratios in identifying the ranges of the most profitable products or those which generates the less margin. Another variable which exerts a clear influence on the performance of the IMF is the proximity of the contacts between the IMF and its customers. The results of the estimations show that the satisfaction in itself does not necessarily stimulate the performance of the IMF. This can be explained by the fact that the customers feel more in security and more involved when they believe that their concerns and their difficulties are shared by their IMF; in short, when they feel like being taken into account and being considered by their IMF, the satisfaction constitutes no more an ideal for the customers.

CONCLUSION
What is the impact of the customer relationship management on the commercial performance of the IMF in Cameroon? Such is the main question on which we have tried to give an answer. The empirical analyses carried out on the data of 60 IMF of the CAMCCUL network show an average level of performance estimated at 0.382 when the returns to scale are constant; 0.823 when the returns to scale are variable and at 0.434 when the returns to scale are average. Globally, these IMF are technically non efficient as far as commercial level is concerned. But these results hide some disparities 7 linked either to the socio-cultural factors (such as the proximity of contacts between the IMF and their customers), either to the influence of the NTCC (computerized management of the customer relationship) and finally or to the satisfaction of customers towards their past experiences, insofar as a good number of these factors (about two or three) exert a statistical and significant influence on the IMF performance.