An automated computerized system for evaluating attrition risk for a company, p1 comprising:
- at least one processor executing a plurality of computer instructions stored in memory, causing the at least one processor to perform:
- (1) receiving and processing data comprising a plurality of frustrations;
- (2) independently automatically matching the received frustrations against customer data, and automatically verifying which of said plurality of frustrations are frustrations of actual and verified company customers, who expressly conveyed to the company the one or more frustrations about a company service or product, and filtering out unverified frustrations;
- (3) automatically identifying and selecting as input to a computerized mathematical model one or more key frustrations verified in (2), wherein the identification and selection of each of said one or more key frustrations as being key frustration is based on a calculation of frequency of said frustration in the plurality of verified frustrations in (2), and based on an evaluation of the calculated frequency of the frustration with an evaluation of at least one other factor comprising automatic (a) evaluation of a strength of the company's current relationship with the actual and verified company customer related to the frustration;
- (b) evaluation of verified company customer's engagement with industry;
- (c) evaluation of the verified company customer's satisfaction with the company;
- (d) evaluation of an out-of-category expectation setting, including evaluation of an income-based category engagement levels with other company products, and determination which other companies in the same industry as the company define an industry's role;
- and (e) evaluation of an identity of a primary relationship owner, including identification of a primary company or a product manufacturer;
- (4) for each of the one or more identified and selected key frustrations of the actual and verified customers identified and selected in (3) performing quantitative vulnerability testing using a computerized mathematical predictive model that utilizes a binominal logistic regression analysis based at least in part on the calculated frequency of the key frustration, a calculated determination of uniqueness of the key frustration in connection to other verified and actual customers of the company and customers of competing companies, and a calculated determination of how much the key frustration prompts switching from the company to competitors;
- (5) for each key frustration processed in step (4), automatically combining and evaluating frustrations of other company customers with respect to the same frustration, wherein the frustrations of other company customers are also evaluated by performing said computerized quantitative vulnerability testing using the computerized mathematical predictive model of step (4);
- (6) determining an individual frustration level vulnerability score for each processed key frustration;
- (7) for each processed key frustration, automatically determining company level vulnerability based on a segmentation of individual frustrations according to one or more segments of individuals and calculated segment level average revenue among the actual and verified company customers, wherein the segmentation relates to a likelihood of attrition for different segments of the actual and verified company customers;
- (8) for each processed key frustration, calculating a predicted probability of attrition for each of said one or more segments of individuals and calculating a business value at risk, caused by the calculated probability of attrition for each of said one or more segments of individuals among the actual and verified company customers, wherein the calculation of the business value at risk includes a determination and quantification of losses to competitors;
- (9) automatically generating a list of remedial measures using at least one rule obtained from a database of company rules, and automatically transmitting said list of remedial measures to one or more indicated persons if the calculated business value at risk exceeds a threshold value for the company, wherein the threshold proportionally exceeds a calculated risk for the whole industry;
- (10) storing the calculate business value at risk data and providing a user interface for modifying dates and values in the calculated business value at risk data;
- (11) in response to the user input in (10), automatically rerunning the analysis from step (1) to (10), using a later obtained frustration data and user input, and automatically comparing the company's calculated business value at risk to the company's calculated business value at risk from prior periods and to trends among competitors.