UX Case Study
UX Case Study
Finding the right prospects shouldn't be this hard
Finding the right prospects shouldn't be this hard
How I transformed lead generation from hours of manual research to intelligent, context-aware prospecting that finds the right people in minutes.
How I transformed lead generation from hours of manual research to intelligent, context-aware prospecting that finds the right people in minutes.
2x increase
In filter utilisation (through AI search)
3X increase
In annual plan subscriptions
2x increase
In filter utilisation (through AI search)
3X increase
In annual plan subscriptions
My Role
Product design & strategy
Duration
3 months (phase 1: 2 weeks)
Tools
Figma (handoff), Notion, V0.dev (prototype)
My Role
Product design & strategy
Duration
3 months (phase 1: 2 weeks)
Tools
Figma (handoff), Notion, V0.dev (prototype)
Discover
Discover
The Challenge we faced
Sales teams are drowning in inefficient tools while prospects slip away. The old way of finding leads wasn't working anymore.
The Problem
Sales teams spend 2+ hours per prospect doing manual research across multiple platforms, often missing key opportunities.
Target Users
B2B sales professionals, account executives, and business development reps seeking efficient prospecting tools.
Business Goals
Reduce research time by 70%, increase qualified lead identification by 200%, and improve sales team productivity.
Competitive analysis
We set out to explore the opportunities we have in the current competitive landscape
Features
Lusha
Apollo
Zoominfo
Filter grouping
Simple
Complex
Complex
ICP definition
NIL
Yes
Yes (but not that user friendly)
AI powered Filter
Free text search
AI Filter with complex prompt interaction
NIL
Global search
No
Yes
Yes
Opportunity
✅ Lusha's filter grouping is good, but limited
✅ A way to let the user define an ICP and have the system apply the appropriate filter seems like a good competitive advantage
Key takeaway from discovery
Drowning in a sea of choices
The overwhelming number of filter make it way more difficult to find and apply the right filter while prospecting
Unclear filter relationships
It's very unclear what the relationship is between different filters and how they affect the search results
There is no way to define ICP
Sales team thing in teams of their ideal customer profile, right now, there is no way to define it and create a search criteria from it
The Challenge we faced
Sales teams are drowning in inefficient tools while prospects slip away. The old way of finding leads wasn't working anymore.
The Problem
Sales teams spend 2+ hours per prospect doing manual research across multiple platforms, often missing key opportunities.
Target Users
B2B sales professionals, account executives, and business development reps seeking efficient prospecting tools.
Business Goals
Reduce research time by 70%, increase qualified lead identification by 200%, and improve sales team productivity.
Competitive analysis
We set out to explore the opportunities we have in the current competitive landscape
Features
Lusha
Apollo
Zoominfo
Filter grouping
Simple
Complex
Complex
ICP definition
NIL
Yes
Yes (but not that user friendly)
AI powered Filter
Free text search
AI Filter with complex prompt interaction
NIL
Global search
No
Yes
Yes
Opportunity
✅ Lusha's filter grouping is good, but limited
✅ A way to let the user define an ICP and have the system apply the appropriate filter seems like a good competitive advantage
Key takeaway from discovery
Drowning in a sea of choices
The overwhelming number of filter make it way more difficult to find and apply the right filter while prospecting
Unclear filter relationships
It's very unclear what the relationship is between different filters and how they affect the search results
There is no way to define ICP
Sales team thing in teams of their ideal customer profile, right now, there is no way to define it and create a search criteria from it
Insights from secondary research
127 minutes
Average time per prospect research
73% of search results
Deemed irrelevant by users
Insights from secondary research
127 minutes
Average time per prospect research
73% of search results
Deemed irrelevant by users
Design process
Design process
Key stakeholders
Suresh & Hari
Engineering Lead, Technical feasibility & AI implementation
Abdul & Ameen
Search team, search algorithm & match accuracy
Chester
Sales lead
Rowan
Customer success
Placing AI search contextual to the filter
Explored different designs finalising the one that is simplest to achieve and ship within 2 weeks, the idea is to solve for users pain point and addresses it as fast as possible and avoid over engineering.
AI search workflow
We built a new way of working, but people don't like learning from scratch. Our solution had to feel both fresh and familiar at the same time. The solution is to introduce it to the current workflow
Result relevance feedback incorporated based on engineering feedback
The search team wanted a way for users to given feedback that helps improve the matching algorithm
Improvisation based on feedback
Based on internal testing, we repurposed our saved search feature to let users save their search queries as ICPs. This helps sales teams define and refine their ideal customer profiles. The positive feedback from our sales team led us to roll out this feature for our main users—SDRs and Sales Engineers—so they can use Firmable to build better ICPs.
Key takeaway
Solving for filter complexity
AI search does solve for the immediate need and reduces the UI complexity
Feedback on the accuracy
Getting the feedback from the users will help with the algorithm the team has built
Sales team think in ICPs
Re-purposing AI search so that users can save their ICPs using current system
Key stakeholders
Suresh & Hari
Engineering Lead, Technical feasibility & AI implementation
Abdul & Ameen
Search team, search algorithm & match accuracy
Chester
Sales lead
Rowan
Customer success
Placing AI search contextual to the filter
Explored different designs finalising the one that is simplest to achieve and ship within 2 weeks, the idea is to solve for users pain point and addresses it as fast as possible and avoid over engineering.
AI search workflow
We built a new way of working, but people don't like learning from scratch. Our solution had to feel both fresh and familiar at the same time. The solution is to introduce it to the current workflow
Result relevance feedback incorporated based on engineering feedback
The search team wanted a way for users to given feedback that helps improve the matching algorithm
Improvisation based on feedback
Based on internal testing, we repurposed our saved search feature to let users save their search queries as ICPs. This helps sales teams define and refine their ideal customer profiles. The positive feedback from our sales team led us to roll out this feature for our main users—SDRs and Sales Engineers—so they can use Firmable to build better ICPs.
Key takeaway
Solving for filter complexity
AI search does solve for the immediate need and reduces the UI complexity
Feedback on the accuracy
Getting the feedback from the users will help with the algorithm the team has built
Sales team think in ICPs
Re-purposing AI search so that users can save their ICPs using current system
The Aha! moment
The Aha! moment
Chester searched for Series B fintech VPs of Engineering recently funded and hiring. In 12 seconds: 47 relevant prospects with contact details and buying signals. Tara said: 'This is exactly what I've been looking for. When will this be available for everyone?"
Chester searched for Series B fintech VPs of Engineering recently funded and hiring. In 12 seconds: 47 relevant prospects with contact details and buying signals. Tara said: 'This is exactly what I've been looking for. When will this be available for everyone?"
Design handoff
Design handoff
Outcomes + Next steps
Outcomes + Next steps
Measured results and insights
We saw a 2x improvement in company and people filter, and a steady month-on-month increase in AI search usage.
2x increase
In filter utilisation (through AI search)
3x
More qualified and accurate prospecting
3X increase
In annual plan
subscriptions
Company filter analytics
AI search analytics
Next steps, improve filter grouping and AI search
After release of AI search, phase 2 and 3 are focused on improving how the filters are grouped and introducing AI search globally.
My learnings
Context beats keywords every time
Users don't think in database terms—they think in business context. 'Growing SaaS companies' means more than 'SaaS AND growth AND company.' The most powerful insight was that human thinking is inherently contextual.
User success drives business success
When Chester showed how we get 3x better prospects in 75% less time, customers didn't just renew they upgraded to annual plans and expanded usage. Great UX became our best sales tool.
Technical Complexity vs. Simple UX
We hid the sophisticated AI behind something everyone knows: a search bar. Users describe their ideal customer in their own words, and it just works—no learning curve required.
Measured results and insights
We saw a 2x improvement in company and people filter, and a steady month-on-month increase in AI search usage.
2x increase
In filter utilisation (through AI search)
3x
More qualified and accurate prospecting
3X increase
In annual plan subscriptions
Company filter analytics
AI search analytics
Next steps, improve filter grouping and AI search
After release of AI search, phase 2 and 3 are focused on improving how the filters are grouped and introducing AI search globally.
My learnings
Context beats keywords every time
Users don't think in database terms—they think in business context. 'Growing SaaS companies' means more than 'SaaS AND growth AND company.' The most powerful insight was that human thinking is inherently contextual.
User success drives business success
When Chester showed how we get 3x better prospects in 75% less time, customers didn't just renew they upgraded to annual plans and expanded usage. Great UX became our best sales tool.
Technical Complexity vs. Simple UX
We hid the sophisticated AI behind something everyone knows: a search bar. Users describe their ideal customer in their own words, and it just works—no learning curve required.