Optimizing benefits selection in view of both member population and organizational preferences

A system comprising one or more processors configured to perform operations for anticipating member satisfaction with each of a plurality of product offerings and applying satisfaction scoring to product offering options, the operations being a) stored as programming instructions to a non-transitory computer readable medium and/or b) programmed to logic circuits of at least a portion of the one or more processors, the operations comprising:

• storing demographic information for each member of a member population belonging to a first organization;
• based on
• i) a set of demographic features of a population of individuals, the set of demographic features comprising, for each respective individual of the population of individuals, one or more of a salary, a salary range indication, a type of employee, a regional location, an age, or an age range indication of the respective individual, and
• ii) historic selection data comprising a) sets of survey data submitted by at least a first portion of the population of individuals, each set of survey data comprising, for each respective individual of the population of individuals, indications of satisfaction with each characteristic of a plurality of product characteristics of a product previously selected by the respective individual of the first portion of the population of individuals, and b) sets of member data of at least a second portion of the population of individuals, each set of member data corresponding to membership of a respective individual within a given organization of a plurality of organizations, each set of member data including a plurality of characteristics of a selected product offering of the plurality of product offerings selected by the respective individual within the given organization,
• training at least one machine learning model using the historic selection data to predict a probability of a given person selecting a given product offering of the plurality of product offerings based on a plurality of characteristics of the given product offering and at least one demographic feature of the given person, and
• periodically updating the training of the at least one machine learning model as new historic selection data is acquired, including promoting, within the training, the importance of a portion of the historic selection data acquired within a most recent calendar year;
• accessing the demographic information;
• using the demographic information, dividing the member population into a set of demographic divisions based on one or more groupings of demographic features of members of the member population and into a set of parameterized demographic divisions based further on characteristics of a current insurance plan held by each member of the member population, wherein
• the one or more groupings of demographic features each include one or more demographic features of the set of demographic features;
• applying the at least one machine learning model to the set of parameterized demographic divisions to obtain, for each respective parameterized demographic division, a probability of the members of the respective parameterized demographic division selecting each product offering of the plurality of product offerings;
• calculating, using the probability of the members of each respective parameterized demographic division selecting each product offering, an employee perception score corresponding to each product offering of the plurality of product offerings for the respective parameterized demographic division;
• using the set of demographic divisions, identifying a set of peer entities to the first organization sharing similarities in the set of demographic divisions, wherein, for each demographic division of the set of demographic divisions, each peer of the set of peer entities differs in distribution of the members across the one or more groupings of demographic features by no more than a threshold value corresponding to the respective demographic division;
• determining, from the set of peer entities, at least one perception benchmark score representing a level of the member satisfaction with the product offering options provided by the set of peer entities to respective members of the peer entities, wherein determining each perception benchmark score of the at least one perception benchmark score comprises:
• identifying at least one peer perception score of each peer of the set of peer entities, and
• calculating an average or median of the peer perception scores of the set of peer entities to determine each perception benchmark score of the at least one perception benchmark score;
• and
• presenting, for review by a representative of the first organization, a graphical user interface with at least one adjustable filter control, the graphical user interface comprising at least a portion of the plurality of product offerings and at least one of a relative comparison or an absolute comparison of the employee perception score to the at least one perception benchmark score.