Before/After: Rebuilding Search UX for an Enterprise Data Startup

Tech · 6 min read

QueryWorks initially exposed its analytics platform through a SQL-forward search bar that favored technically fluent users but locked out analysts and business users. Usability testing showed that non-technical users struggled with syntax, leading to heavy reliance on centralized analyst teams and slow decision cycles. The product team chose to layer a natural-language interface over the query engine with guided templates for advanced queries.

Designers created a hybrid experience: a conversational search input that translated plain-language questions into optimized queries, combined with an expert panel that allowed raw SQL editing for power users. The UI also included smart suggestions, clarifying tooltips, and a persistent example library tailored to common business questions. Prototypes were validated in-context with cross-functional teams to ensure result accuracy and transparency.

After launch, the percentage of ad-hoc queries performed by non-technical users increased by 35% and average analyst turnaround time decreased by 42%. Confidence in self-service analytics improved—feedback polls showed a 28-point rise in perceived autonomy. To maintain trust, the team added an explainability panel showing how natural-language prompts mapped to query logic and data lineage.

This redesign shows how enterprises can democratize complex tools through layered interfaces that respect both novices and experts. The key is transparent mapping between natural language and technical operations so users understand and trust outcomes while easing the load on specialized teams.