Inclusion Behaviour Uplift
30-45%
Cohorts Delivered
14
Locations
4 Cities
Behavioural Transfer Window
90 Days
Cohort
Middle and senior managers across four metro cities
Scale
15,000-person institution
Methodology
Unseen Verdicts, Elephant in the Room, REEL|Life
Duration
14 cohorts over 18 months
The problem before ProventusHR
The organisation had a well-resourced diversity function and a strong stated commitment to inclusion. What it lacked was observable evidence that the commitment was producing behaviour change at the manager level, where the actual day-to-day experience of employees was being shaped. Exit interview data showed that departing women consistently cited a feeling of being seen but not heard, acknowledged but not sponsored, and held to a different standard than their male peers. The bank had run sensitivity training previously. The feedback from those sessions was generally positive. The behaviour change data at 90 days was not. The HR leadership team came to ProventusHR with a specific ask: stop training people to say the right things, and build something that actually changes what they do when they are not being watched.
What the diagnostic surfaced
ProventusHR conducted discovery interviews with thirty managers and twelve HR business partners before designing the intervention. The finding was not that managers were malicious or wilfully discriminatory. The finding was that the biases operating in the institution were almost entirely invisible to the people holding them. Affinity bias was producing identical hiring and sponsorship decisions consistently. Managers who genuinely believed they were meritocratic were, when presented with matched profiles, systematically rating male candidates higher on strategic thinking and female candidates higher on execution. Nobody was doing this deliberately. That was precisely what made it so difficult to address through awareness alone. Awareness of bias in the abstract does not shift the specific decisions a manager makes in a hiring panel or a performance calibration meeting. What shifts those decisions is an encounter with the pattern in oneself, under conditions where denial is made difficult.
ProventusHR designed a two-intervention architecture. The first intervention, Unseen Verdicts, was a full-day session using a structured cinematic reflection methodology. A curated sequence of film scenes placed participants inside situations where bias was operating at a level too subtle to identify on a first pass but unmistakable on the debrief. The debrief was designed to produce not agreement with a principle but recognition of a pattern. The second intervention, Elephant in the Room, ran two weeks later with the same cohort. It built the behavioural architecture: specific, observable commitments to one or two inclusion behaviours that participants would practice over the following 90 days. Pre-programme 360 data was collected on eight inclusion-related behaviours using a validated tool. The same tool was administered at 90 days. Reporting managers were briefed on the programme objectives and asked to validate specific observed behaviour shifts at 30 and 90 days.
Fourteen cohorts ran across four cities over eighteen months. The design allowed for city-specific contextualisation , the dynamics of inclusion in a Hyderabad corporate banking floor and a Mumbai retail branch are not identical, and the facilitators adjusted the film selection and scenario framing accordingly. Several cohorts included significant numbers of male managers, by design. ProventusHR has consistently found that inclusion interventions targeting women alone produce limited systemic impact. The bank agreed, and the cohort composition reflected the full manager population rather than a protected sub-group.
Post-programme 360 data at 90 days showed a 30 to 45 percent improvement across the eight measured inclusion behaviours. The range reflects variation across cohorts rather than a single aggregate figure. Three cohorts in corporate banking showed the highest gains, which the facilitators attributed to the design having been most precisely targeted at the specific bias patterns surfaced in the discovery phase for that function. Manager validation rates at 90 days were above 80 percent across all cohorts. A participant from a senior management cohort wrote: I have been in diversity training for twenty years. This is the first time I left a programme thinking about my own behaviour rather than about other people who need to change.
Bias work fails when it positions the participant as a problem to be fixed. It succeeds when it creates the conditions for the participant to discover, on their own terms and with their own data, a pattern they did not know they had. The shift from awareness to behaviour requires an encounter with specific evidence, not with general principles. The encounter has to be designed, not facilitated. The design is the intervention.
Engineering Consulting · Leadership Capability Development
CRAFT 1.0: building project leadership from the site up
5 months · CRAFT, ExperienceLearning, REEL|Life, Computer Simulations
Health and Pharmaceuticals · Executive Coaching
Post-merger MD coaching: from operational control to strategic influence
9 months · Marshall Goldsmith Stakeholder Centred Coaching
FMCG and Consumer Products · Women Leadership and DEIB
Women emPower: accelerating senior women beyond awareness
6 months · Women emPower, ExperienceLearning, Individual Coaching