Methodology
The project methodology involved a significant mindset shift, moving from a segment-focused model to a behaviourally-driven journey model. In this new approach, we concentrated on understanding entrepreneurs' personal paths, intrinsic motivations, relationships, and transitions throughout their journey.
We analysed the company’s existing data to identify critical drop-out points along the journey. This analysis helped define our target profiles and shaped the key research questions for further exploration through qualitative behavioural research.
We conducted fieldwork in three locations—Moscow, Bratsk, and Orenburg—to capture a range of social and entrepreneurial contexts, interviewing 12 to 15 entrepreneurs at each site. The research design included individual, dyad, and group interviews with entrepreneurs, in-context observations, and informal conversations at community meetings, points of sale, and other relevant settings. Additionally, we interviewed both local and international experts on entrepreneurial platforms. During the interviews, we moved beyond a traditional question-and-answer format by co-creating artifacts such as emotional collages, network maps, and visual journeys.
We synthesised the research findings to develop behavioural personas, each characterised by unique primary and secondary motivations for joining, as well as barriers to entry and engagement. Each persona included a one-line motivation statement that served as a key prompt for ideation when designing new propositions to keep them engaged.
Additionally, we formulated design principles and actionable insight statements, connected to 'How might we?' questions that facilitated ideation discussions. Each insight was linked to the relevant persona(s). These served as inputs for stakeholder workshops, where we collaborated with key company stakeholders and selected entrepreneurs to co-create and prioritise new concepts that addressed the identified opportunities.
Results
The concepts generated during the workshops were prioritised by the core team, with six selected for further development. Of these, two were implemented as a Live Model Test (LMT) in Kazakhstan.