Fixing Job Relevance on Naukri.com
Interviews and a survey on Naukri showing how irrelevant matches wear down trust on both sides of the marketplace
✳The problem they went after
Naukri has brand recognition and enormous listing volume, and both sides still come away frustrated. Job seekers field calls for roles that bear no relationship to their profiles. Employers running paid listings report single-digit applicant responses over weeks. The submission reads the keyword-matching algorithms as optimising for volume of connections rather than quality, which produces the appearance of activity without genuine matches.
✳What they actually did
Three in-depth interviews, deliberately chosen for different vantage points: an active job seeker, an employer who paid for a listing, and a community contributor aggregating observations from across the platform. A 20-response survey then quantifies what the interviews surfaced. A structured SWOT frames the opportunity as better AI and machine learning recommendation accuracy combined with stricter employer verification, moving satisfaction from 85 percent to a 90 percent target by end of 2025. Two developed personas close it out.
✳What you can take from it
Worth stealing
- Three interviews chosen for opposing marketplace vantage points
- Qualitative first, then a 20-response survey to size it
- Reads keyword matching as optimising connections over quality
- SWOT converted into a dated satisfaction target
Read this if
Running research on a two-sided marketplace with a small interview budget
The transferable bitThree interviews go a long way when each sits at a different point in the marketplace, because the same complaint arriving from a seeker and from a paying employer is far harder to dismiss than either one alone. Qualitative first and survey second is the right order for a problem nobody has framed yet.
✳The submission itself
Open the PDF →Reading on a phone? The embedded viewer is cramped — open the PDF in a new tab instead.
