Robota.ua

How I did it, and what I used:

• Framed the real problem first
  The company had fewer active employers and published vacancies than its main competitor. Instead of guessing why, I scoped the investigation specifically to the segment that mattered: employers posting 0–5 jobs/month, and how they choose between platforms.
  
  

• Prioritized ideas with data, not opinion
  After sampling users, used a Value Proposition Canvas and 2-factor analysis (user benefit vs. implementation complexity) to rank which ideas were worth building first, rather than designing everything at once.
  
  

• Built and tested 5 competing hypotheses
  Multi-company account switching, richer candidate previews (status, experience, key-skill tags), an improved job creation form, better resume structure, and a recommendation system. Each was built as a clickable prototype in Figma and tested against real personas (Veronica, Mariia) with structured usability sessions.
  
  

• Iterated on real feedback, not assumptions
  Account switching tested well overall, but one respondent flagged that company identity wasn't visually distinct enough in the header/side menu when switching, a fix that fed straight back into the design. The candidate-preview hypothesis (status, experience, tags) resonated strongly, directly solving Mariia's pain of manually opening 100 resumes to find the same info.
  
  

• Delivered validated, not just designed
  Packaged prototypes, testing notes, and prioritization rationale so the product team could make an informed call rather than a subjective one.