Search & Stock Discoverability
Problem
Investors couldn't reliably find what they were looking for in stock search across 1,500+ equities.
Impact
1,500+ equities made discoverable via filtered search, with improved result relevance across 1.5L+ scrips.
On an investing app, search is the whole job. If a user can't find the stock they came to buy, the whole app fails at the one thing they opened it for. Discoverability across 1.5 lakh+ scrips was weaker than it should have been.
What I did
I revamped the search logic to make results both broader and more precise. That meant layering in keyword, sector, and product filters so users could approach a stock from whatever angle they were thinking in, and a priority-based ranking system so the most relevant result showed up first instead of buried.
Where it landed
equities
made discoverable via filtered searchscrips
improved result relevanceThe fix was ranking. Get the one stock someone's looking for to the top, instead of just returning more of them.
Teaching the ranking
Dynamic search shipped as part of this work. A user could enter an industry or a keyword and get back stocks they would never have typed the name of. That only helps if people know it works that way, so the rollout came with a public walkthrough of how the app prioritises results, and how industries and keywords can be used to discover stocks.
Search is the one feature where "close enough" means the user leaves.
The ranking logic, the relevance scoring behind it, and the internal search analytics are under NDA.
