< Back to all solutions

Jobs-To-Be-Done Segmentation


Looking for that elusive white space in your category? One approach for identifying innovation opportunities is through jobs-to-be done (JTBD) analysis, a framework oriented toward uncovering the problems consumers are trying to solve when they make a purchase or use a product. Commission a JTBD Segmentation as a standalone research test or add to any custom study to increase the utility and interpretability of your JTBD MaxDiff exercise!

What you'll get

Your Custom Report will include a total-level JTBD comparison illustrating the relative importance of various jobs-to-be-done in your category based on a MaxDiff exercise. Additionally, consumer segments identified in the optimal clustering solution will be visualized, sized, and described according to their differentiating JTBD priorities, as well as demographic, behavioral and psychographic characteristics.

Key metrics
  • MaxDiff results
    Preference likelihood
  • Cluster solutions
    In-sample Accuracy, RLH, BIC, and Calinski-Harabasz measure of purity are used to assess cluster model fit and determine viable cluster solutions for the dataset. Solutions are then examined from a market fit perspective (in collaboration with the client) to determine the most compelling clustering solution for the dataset.
  • Profiled data
    Key differentiating characteristics are identified for each consumer cluster, based not only on their prioritized jobs-to-be-done, but also on any relevant demographic, behavioral, and psychographic metrics.
Common applications

Product development

Identify nuance and variety in category white space, in order to plan new products and/or new product features that will be most compelling to consumers in your category.


Develop compelling and specific JTBD messaging around new products and/or features as they go to market.
  • Relevant respondents complete a MaxDiff exercise in which, over multiple screens, they indicate the most and least important jobs-to-be-done among the subset that is shown. The number of items shown in a subset and the total number of screens shown varies according to the total number of JTBD being tested.
  • Optional
    Respondents may then complete a satisfaction matrix, so that a gap analysis may be conducted between importance and satisfaction/current products’ performance, which will allow your brand to quickly assess high level white space in the category.
  • Optional
    Respondents may also complete a series of profiling questions that will serve to bring the MaxDiff-based clusters to life - these may include demographic, behavioral, and psychographic questions relevant to the category.

Cluster Analysis: JTBD MaxDiff data is examined in two ways: 1) The first uses the categorical responses from the raw MaxDiff choice tasks, in which the selections of most important and least important JTBD choice task alternatives are used as categorical inputs into Latent Class Analysis. 2) The second utilizes the probability scaled scores after Hierarchical Bayesian analysis. For this second treatment of the MaxDiff data, both Latent Class Analysis & K-means clustering are used.