Query
Crafting better ways for users to ask and uncover meaningful insights
(Domain)
Analytics
(Year)
2022
(Services)
Product Design
About
the product
A product built to extract meaningful insights, with Querybuilder powering deep dives.
The platform helps companies analyze and quantify customer voice by aggregating feedback from channels like app stores, support tools, and call transcripts. These feedback are broken into repeated phrases (keywords) that define themes, and reasons that add weight, forming the core parameters for filtering, segmenting, and validating insights at scale.
At the center of this sits Querybuilder, the entry point for deep dives. It powers record lists, Quantify for validation, and Dashboards for monitoring by letting users craft tailored questions to surface relevant feedback. But as the platform grew to support more sources and metadata, the Querybuilder’s UI grew overly complex, pushing users toward manual review or external tools.
Context &
Problem
As the product matured, Querybuilder became the backbone of insight discovery. In Search, it allowed teams to form queries that pulled precise sets of feedback records. In Dashboards, it powered the filters behind recurring views of customer trends. And in Quantify, it enabled complex breakdowns for deeper analysis. But as usage scaled, cracks appeared, its growing complexity and rigid design began to slow adoption, frustrate users, and limit the product’s reach.
Business cost: Low adoption, slower time-to-insight, and reduced trust from product and CX teams.
Mismatch with mental models: The product built for power users, leaves new analysts and cross-functional teams struggling with discoverability and progressive workflows, indicating signs of steep learning curve
Onboarding drop-off: Most users failed to build a first query, creating a conversion bottleneck, which lead users to be more reliant external tools or competitive alternatives
In short: Querybuilder being a foundational feature was sought after heavily at the time but its design debt blocked adoption, scalability, and speed to insight.
Goals
The redesign set out to re-establish Querybuilder as a confident, approachable, and scalable entry point for insight discovery, with goals centered on:
Scalability & future-proofing: Build a flexible framework that could seamlessly support new filter types, evolving taxonomy layers, and AI-driven features without adding clutter.
Clarity in nested queries: Redesign how operators and filters interact especially for complex nesting queries, making it intuitive to construct and easy to scan.
Decreasing the learning curve: Streamline the experience to help more users form their first queries successfully, directly improving conversion from onboarding to active usage.
The overarching goal: drive higher adoption and conversion; enabling more users to confidently form queries while giving power users the flexibility to go deep without friction.
Initial Research
The research began with a one-month observational study of query-building workflows in analytics tools.Some of the Key findings:
Only ~15% of users formed their first query post-onboarding, most dropped after one attempt.
Shared dashboards/feedback records brought several users in, but query creation for the same cohort from scratch, was difficult.
Nested queries and operators caused confusion and errors in most of these sessions.
Core actions like “Execute Query” and “Add Filter” were often missed.
Customer interviews and community feedback reinforced concerns around complexity, discoverability, and learning curve.
Competitive analysis showed best practices:
Progressive disclosure to ease onboarding.
Cleaner vertical layouts for readability.
Clarity-focused query design to balance novice and advanced use.
These insights highlighted the need to shift from a dense, power-user-first model to a guided, progressive approach serving both new and expert users.
Lo-Fi Designs
We explored three directions through low-fidelity wireframes:
Minor cleanup – Simplified visual design and spacing to reduce clutter.
Progressive disclosure – Inspired by FullStory, this version started with simple query blocks and expanded into deeper details as needed.
Vertical structuring – Inspired by Amplitude, this version separated filters, operators, and values into distinct layers for clarity.
Testing these internally revealed strong support for the progressive disclosure approach, as it balanced simplicity for new users with flexibility for advanced queries.
Additional feedback:
Users wanted a more discoverable filter menu, the existing list made it hard to explore new possibilities.
Analytics tied to queries (e.g., just listing reasons) felt too limiting; there was appetite for richer, contextual insights alongside results.
Hi-Fi Designs
High-fidelity designs refined the chosen approach and integrated user feedback:
Progressive query flow: Clean visual hierarchy with step-by-step query building, making even nested conditions clear.
Two-column filter palette: Filters displayed with preview values, improving discoverability and reducing guesswork.
Enhanced inline analytics: Added top reasons, feedback count trendlines, and source distribution directly in the results view, giving users immediate context for their queries.
We validated these designs through:
Internal alpha testing: Teams reported greater confidence in crafting both simple and complex queries.
Beta testing with customer orgs: Users discovered new filters for the first time, leading to richer exploration and more tailored insights.
Outcomes
Adoption growth: Querybuilder usage saw a significant uptick, with first-query success rates climbing from ~15% to ~68%. Conversion to a second query also improved, reaching ~42%.
Customer retention: Newer users who previously relied only on dashboards began creating their own queries, deepening engagement.
Business impact: By directly addressing long-standing usability concerns, the redesign improved customer trust and helped retain key accounts.
Takeaway
What worked: Progressive disclosure simplified query building while still supporting advanced scenarios. The two-column filter palette made exploration easier and more engaging.
What was missing: Demand for custom, saveable filters emerged. While outside the scope of this iteration, it became a roadmap priority.
Key learning: Inline analytics saw limited adoption. Users viewed Querybuilder primarily as a filtering and discovery tool, preferring Quantify or Dashboards for deeper trend analysis. This led us to deprecate inline analytics in a follow-up update.
✨ The Querybuilder redesign re-established it as a core entry point into the feedback ecosystem. By moving from a complex, power-user interface to a guided, progressive experience, it boosted adoption, retention, and trust.
The key principle: simplicity drives confidence. Empowered users built queries more easily, engaged deeper, and relied on the platform as their main hub for customer intelligence.
Query
Crafting better ways for users to ask and uncover meaningful insights
(Domain)
Analytics
(Year)
2022
(Services)
Product Design
About
the product
A product built to extract meaningful insights, with Querybuilder powering deep dives.
The platform helps companies analyze and quantify customer voice by aggregating feedback from channels like app stores, support tools, and call transcripts. These feedback are broken into repeated phrases (keywords) that define themes, and reasons that add weight, forming the core parameters for filtering, segmenting, and validating insights at scale.
At the center of this sits Querybuilder, the entry point for deep dives. It powers record lists, Quantify for validation, and Dashboards for monitoring by letting users craft tailored questions to surface relevant feedback. But as the platform grew to support more sources and metadata, the Querybuilder’s UI grew overly complex, pushing users toward manual review or external tools.
Context &
Problem
As the product matured, Querybuilder became the backbone of insight discovery. In Search, it allowed teams to form queries that pulled precise sets of feedback records. In Dashboards, it powered the filters behind recurring views of customer trends. And in Quantify, it enabled complex breakdowns for deeper analysis. But as usage scaled, cracks appeared, its growing complexity and rigid design began to slow adoption, frustrate users, and limit the product’s reach.
Business cost: Low adoption, slower time-to-insight, and reduced trust from product and CX teams.
Mismatch with mental models: The product built for power users, leaves new analysts and cross-functional teams struggling with discoverability and progressive workflows, indicating signs of steep learning curve
Onboarding drop-off: Most users failed to build a first query, creating a conversion bottleneck, which lead users to be more reliant external tools or competitive alternatives
In short: Querybuilder being a foundational feature was sought after heavily at the time but its design debt blocked adoption, scalability, and speed to insight.
Goals
The redesign set out to re-establish Querybuilder as a confident, approachable, and scalable entry point for insight discovery, with goals centered on:
Scalability & future-proofing: Build a flexible framework that could seamlessly support new filter types, evolving taxonomy layers, and AI-driven features without adding clutter.
Clarity in nested queries: Redesign how operators and filters interact especially for complex nesting queries, making it intuitive to construct and easy to scan.
Decreasing the learning curve: Streamline the experience to help more users form their first queries successfully, directly improving conversion from onboarding to active usage.
The overarching goal: drive higher adoption and conversion; enabling more users to confidently form queries while giving power users the flexibility to go deep without friction.
Initial Research
The research began with a one-month observational study of query-building workflows in analytics tools.Some of the Key findings:
Only ~15% of users formed their first query post-onboarding, most dropped after one attempt.
Shared dashboards/feedback records brought several users in, but query creation for the same cohort from scratch, was difficult.
Nested queries and operators caused confusion and errors in most of these sessions.
Core actions like “Execute Query” and “Add Filter” were often missed.
Customer interviews and community feedback reinforced concerns around complexity, discoverability, and learning curve.
Competitive analysis showed best practices:
Progressive disclosure to ease onboarding.
Cleaner vertical layouts for readability.
Clarity-focused query design to balance novice and advanced use.
These insights highlighted the need to shift from a dense, power-user-first model to a guided, progressive approach serving both new and expert users.
Lo-Fi Designs
We explored three directions through low-fidelity wireframes:
Minor cleanup – Simplified visual design and spacing to reduce clutter.
Progressive disclosure – Inspired by FullStory, this version started with simple query blocks and expanded into deeper details as needed.
Vertical structuring – Inspired by Amplitude, this version separated filters, operators, and values into distinct layers for clarity.
Testing these internally revealed strong support for the progressive disclosure approach, as it balanced simplicity for new users with flexibility for advanced queries.
Additional feedback:
Users wanted a more discoverable filter menu, the existing list made it hard to explore new possibilities.
Analytics tied to queries (e.g., just listing reasons) felt too limiting; there was appetite for richer, contextual insights alongside results.
Hi-Fi Designs
High-fidelity designs refined the chosen approach and integrated user feedback:
Progressive query flow: Clean visual hierarchy with step-by-step query building, making even nested conditions clear.
Two-column filter palette: Filters displayed with preview values, improving discoverability and reducing guesswork.
Enhanced inline analytics: Added top reasons, feedback count trendlines, and source distribution directly in the results view, giving users immediate context for their queries.
We validated these designs through:
Internal alpha testing: Teams reported greater confidence in crafting both simple and complex queries.
Beta testing with customer orgs: Users discovered new filters for the first time, leading to richer exploration and more tailored insights.
Outcomes
Adoption growth: Querybuilder usage saw a significant uptick, with first-query success rates climbing from ~15% to ~68%. Conversion to a second query also improved, reaching ~42%.
Customer retention: Newer users who previously relied only on dashboards began creating their own queries, deepening engagement.
Business impact: By directly addressing long-standing usability concerns, the redesign improved customer trust and helped retain key accounts.
Takeaway
What worked: Progressive disclosure simplified query building while still supporting advanced scenarios. The two-column filter palette made exploration easier and more engaging.
What was missing: Demand for custom, saveable filters emerged. While outside the scope of this iteration, it became a roadmap priority.
Key learning: Inline analytics saw limited adoption. Users viewed Querybuilder primarily as a filtering and discovery tool, preferring Quantify or Dashboards for deeper trend analysis. This led us to deprecate inline analytics in a follow-up update.
✨ The Querybuilder redesign re-established it as a core entry point into the feedback ecosystem. By moving from a complex, power-user interface to a guided, progressive experience, it boosted adoption, retention, and trust.
The key principle: simplicity drives confidence. Empowered users built queries more easily, engaged deeper, and relied on the platform as their main hub for customer intelligence.
Query
Crafting better ways for users to ask and uncover meaningful insights
(Domain)
Analytics
(Year)
2022
(Services)
Product Design
About
the product
A product built to extract meaningful insights, with Querybuilder powering deep dives.
The platform helps companies analyze and quantify customer voice by aggregating feedback from channels like app stores, support tools, and call transcripts. These feedback are broken into repeated phrases (keywords) that define themes, and reasons that add weight, forming the core parameters for filtering, segmenting, and validating insights at scale.
At the center of this sits Querybuilder, the entry point for deep dives. It powers record lists, Quantify for validation, and Dashboards for monitoring by letting users craft tailored questions to surface relevant feedback. But as the platform grew to support more sources and metadata, the Querybuilder’s UI grew overly complex, pushing users toward manual review or external tools.
Context &
Problem
As the product matured, Querybuilder became the backbone of insight discovery. In Search, it allowed teams to form queries that pulled precise sets of feedback records. In Dashboards, it powered the filters behind recurring views of customer trends. And in Quantify, it enabled complex breakdowns for deeper analysis. But as usage scaled, cracks appeared, its growing complexity and rigid design began to slow adoption, frustrate users, and limit the product’s reach.
Business cost: Low adoption, slower time-to-insight, and reduced trust from product and CX teams.
Mismatch with mental models: The product built for power users, leaves new analysts and cross-functional teams struggling with discoverability and progressive workflows, indicating signs of steep learning curve
Onboarding drop-off: Most users failed to build a first query, creating a conversion bottleneck, which lead users to be more reliant external tools or competitive alternatives
In short: Querybuilder being a foundational feature was sought after heavily at the time but its design debt blocked adoption, scalability, and speed to insight.
Goals
The redesign set out to re-establish Querybuilder as a confident, approachable, and scalable entry point for insight discovery, with goals centered on:
Scalability & future-proofing: Build a flexible framework that could seamlessly support new filter types, evolving taxonomy layers, and AI-driven features without adding clutter.
Clarity in nested queries: Redesign how operators and filters interact especially for complex nesting queries, making it intuitive to construct and easy to scan.
Decreasing the learning curve: Streamline the experience to help more users form their first queries successfully, directly improving conversion from onboarding to active usage.
The overarching goal: drive higher adoption and conversion; enabling more users to confidently form queries while giving power users the flexibility to go deep without friction.
Initial Research
The research began with a one-month observational study of query-building workflows in analytics tools.Some of the Key findings:
Only ~15% of users formed their first query post-onboarding, most dropped after one attempt.
Shared dashboards/feedback records brought several users in, but query creation for the same cohort from scratch, was difficult.
Nested queries and operators caused confusion and errors in most of these sessions.
Core actions like “Execute Query” and “Add Filter” were often missed.
Customer interviews and community feedback reinforced concerns around complexity, discoverability, and learning curve.
Competitive analysis showed best practices:
Progressive disclosure to ease onboarding.
Cleaner vertical layouts for readability.
Clarity-focused query design to balance novice and advanced use.
These insights highlighted the need to shift from a dense, power-user-first model to a guided, progressive approach serving both new and expert users.
Lo-Fi Designs
We explored three directions through low-fidelity wireframes:
Minor cleanup – Simplified visual design and spacing to reduce clutter.
Progressive disclosure – Inspired by FullStory, this version started with simple query blocks and expanded into deeper details as needed.
Vertical structuring – Inspired by Amplitude, this version separated filters, operators, and values into distinct layers for clarity.
Testing these internally revealed strong support for the progressive disclosure approach, as it balanced simplicity for new users with flexibility for advanced queries.
Additional feedback:
Users wanted a more discoverable filter menu, the existing list made it hard to explore new possibilities.
Analytics tied to queries (e.g., just listing reasons) felt too limiting; there was appetite for richer, contextual insights alongside results.
Hi-Fi Designs
High-fidelity designs refined the chosen approach and integrated user feedback:
Progressive query flow: Clean visual hierarchy with step-by-step query building, making even nested conditions clear.
Two-column filter palette: Filters displayed with preview values, improving discoverability and reducing guesswork.
Enhanced inline analytics: Added top reasons, feedback count trendlines, and source distribution directly in the results view, giving users immediate context for their queries.
We validated these designs through:
Internal alpha testing: Teams reported greater confidence in crafting both simple and complex queries.
Beta testing with customer orgs: Users discovered new filters for the first time, leading to richer exploration and more tailored insights.
Outcomes
Adoption growth: Querybuilder usage saw a significant uptick, with first-query success rates climbing from ~15% to ~68%. Conversion to a second query also improved, reaching ~42%.
Customer retention: Newer users who previously relied only on dashboards began creating their own queries, deepening engagement.
Business impact: By directly addressing long-standing usability concerns, the redesign improved customer trust and helped retain key accounts.
Takeaway
What worked: Progressive disclosure simplified query building while still supporting advanced scenarios. The two-column filter palette made exploration easier and more engaging.
What was missing: Demand for custom, saveable filters emerged. While outside the scope of this iteration, it became a roadmap priority.
Key learning: Inline analytics saw limited adoption. Users viewed Querybuilder primarily as a filtering and discovery tool, preferring Quantify or Dashboards for deeper trend analysis. This led us to deprecate inline analytics in a follow-up update.
✨ The Querybuilder redesign re-established it as a core entry point into the feedback ecosystem. By moving from a complex, power-user interface to a guided, progressive experience, it boosted adoption, retention, and trust.
The key principle: simplicity drives confidence. Empowered users built queries more easily, engaged deeper, and relied on the platform as their main hub for customer intelligence.