The PIMPmyBOSS method: from answers to profiles and compatibility
V1 uses a deterministic, versioned engine. Answers are calculated server-side, active parameters are documented and historical results are never rewritten by a future version.
The engine separates four objects: profile, compatibility, relationship quality and lived score. Both profiles distribute 100% across five reference centres according to weighted proximity; compatibility is built from ten mirror pairs and an explainable distance.
Two families of ten axes, two families of five archetypes.
The employee side describes ten professional preferences. The manager side describes ten lived behaviours. Employee profiles are Star, Pilot, Specialist, Creative and Collective; manager components are Transformational, Directive, Persuasive, Delegative and Participative.
PIMPmyBOSS presents employee profiles as proprietary archetypes. On the manager side, it combines four styles from the Hersey-Blanchard situational family with a Transformational component linked notably to Bass; this is a PIMPmyBOSS synthesis, not one official academic taxonomy.
A profile distributes 100% across five centres according to proximity, rather than putting someone in a single box.
The engine calculates the weighted distance between the ten answers and each of the five centres. It converts those five distances into positive proximities and normalises them into five shares totalling exactly 100%. Dimensions that genuinely distinguish the centres carry more weight; no artificial floor is added.
The initial centres are an explicit, versioned theoretical prior. They are not presented as learned from a real population and a new version will never recalculate historical results.
Ten explainable pairs instead of an opaque matrix of one hundred crossings.
Ten mirror pairs are active and equally weighted in the current version. The other ninety crossings remain neutral until sufficient justification and data support activating them cleanly.
The RMS distance from those ten pairs is transformed by 100 × (1 − RMS / 10). The transformation is transparent, monotonic and explicitly non-empirical.
An insight must be tied to a specific gap.
The active insight set only displays an insight when an absolute gap reaches at least 3 points. It keeps at most three and selects the wording from the actual direction of each gap.
Micro-validations — right, partial or wrong — will be used to evaluate insight relevance. They never instantly change an individual result or engine parameters.
An engine change creates a new version instead of rewriting the past.
Questionnaires, centres, pairs, calibration, privacy parameters and insight sets are versioned. Every completed cycle keeps the exact version that produced its result.
Calculations keep full precision server-side; rounding only happens at the display boundary.
Frequently asked questions
Important answers should remain understandable without animation, internal jargon or dependence on a personal result.
Is profile scoring performed in the browser?
No. The front collects answers, but profile components, compatibility and other scores are calculated exclusively server-side.
Why doesn’t answering 5 everywhere necessarily produce 20% everywhere?
Because the point 5 across ten axes is not necessarily equally distant from the five centres. The engine preserves those proximity differences instead of forcing artificial symmetry.
Can the reference centres change?
Yes, when recalibration is defensible with real-world data and separate validation. Any change will create a new version and will not rewrite historical results.
Does PIMPmyBOSS use a hidden ‘good manager’ threshold?
No. Styles are not ranked and compatibility encodes no good, average or bad threshold.
The model becomes clearer when the pages connect.
See what this looks like in a real relationship.
See the journey in practiceEvery criterion is defined before it is interpreted.
What is established, and what still needs validation
PIMPmyBOSS brings management frameworks together and builds its own framework of professional preferences. Mathematical consistency alone does not demonstrate predictive validity. The five employee profiles are not a validated psychological test.
The active method uses versioned theoretical centres. Fixed-choice feedback and the overall lived-experience rating help assess the relevance of interpretations. Empirical validation requires real data, a protocol and a separate sample; it must not be claimed before it has been carried out.
An example to explain the formula
Purely illustrative example: if all ten aligned differences are 0, compatibility is 100%. If they are all 5, it is 50%. If one difference is 10 and the other nine are 0, it is about 68%. The calculation therefore accounts for one large difference without treating it as ten large differences.
Stability and Adjustment are compared in opposite directions: the management value used is 10 minus Adjustment. The lived-experience rating and the four quality signals remain outside this formula.
Sources and the scope of the method
These studies inform mechanisms of motivation, speaking up and management. They do not validate PIMPmyBOSS scores or archetypes. PIMPmyBOSS is responsible for the synthesis and its parameters.