Case study · HubSpot
One approachable workspace for building, cleaning and enriching CRM data, launched on schedule and now used by 400+ portals spending credits, up 292% since September 2025.
Preparing CRM data meant moving between separate, technical tools, and few customers used the existing Datasets product. HubSpot needed one approachable workspace to anchor the Data Hub relaunch at INBOUND ’25.
I led UX from early discovery through launch. I defined the design principles and what “approachable” meant, co-led the vision workshops, and owned every AI experience: Smart Columns, the no-code editors and the Breeze assistant. All of them run on one shared, panel-based interaction system.
Business context: HubSpot was relaunching Data Hub, and data preparation (building, cleaning and enriching datasets) was a core pillar of it. Datasets, the existing tool, had low adoption. The goal was a single workspace launched at INBOUND ’25, HubSpot’s biggest annual event, with a target of 5x Datasets adoption by 2026.
Customer problem: Preparing CRM data meant moving between separate tools, and the work felt technical and intimidating to non-specialists. [Who the core users were (RevOps, marketing ops, admins), and what they were doing instead, such as spreadsheets, exports or third-party tools.]
Evidence: [Findings from early discovery and concept research, e.g. number of interviews, where users struggled, and Datasets adoption before launch.]
“[Customer quote about how hard or fragmented data prep was]”
Problem statement: How might we make preparing and enriching CRM data approachable for non-technical users, so that more customers build datasets and get value from AI features?
I owned: The end-to-end UX strategy for Data Studio, and I made the final UX decisions across the related product teams. I co-led the vision workshops, defined the core problems and design principles, and owned every AI experience: Smart Columns, the no-code editors and the Breeze assistant. For each of those, I set the KPIs, interaction model and success metrics.
I partnered on: Workstreams and focus areas with Data Hub leadership; beta and pre-INBOUND milestones with PMs; a single Data Agent experience across Data Studio, CRM and Workflows with the teams that own those products.
I ran early discovery and concept research and mapped how customers clean their data today.
Key insight: [The finding that shaped the product, e.g. users didn’t think of themselves as “data people” and needed guidance rather than power tools.]
I wrote the Data Studio design principles and led an approachability exercise to turn “make it approachable” into specific, testable goals. I also helped leadership set up workstreams and focus areas.
I co-led vision workshops, including a Summit workshop on 5/14/25, and walked Data Hub leadership through the vision flow.
I tested concepts with prototypes, ran the beta, tracked pre-INBOUND milestones, and cleared blockers to keep teams on schedule. [Number of beta customers and what changed because of them.]
How I got buy-in: Vision workshops, a walkthrough for Data Hub leadership, and working prototypes gave teams a shared picture of “what good looks like” before anyone committed to a build. [A moment where a prototype or workshop changed someone’s mind.]
Building, cleaning and enriching datasets now happen in one place, designed around the approachability principles so non-specialists can work with confidence. [One or two specifics, e.g. guided empty states or no-code editors.]
Smart Columns let users enrich or transform data with AI and web search, right inside the dataset. I designed their credit usage, error states and billing experience, so users understand what AI actions do and what they cost.
A panel-based action library covers complex workflows and user states, backed by component audits and a regular design review rhythm. It gives every Data Studio team one consistent pattern, and the same Data Agent experience appears in CRM and Workflows.
Data Studio launched on schedule at INBOUND ’25. Adoption grew quickly: 400+ portals were spending credits in it within months, a 292% increase since September 2025. Customers have spent ~1.3M credits in total, and adding a credit estimate later lifted Data Studio activations ~12% (see the Credit Transparency case study).
Qualitative impact: My manager wrote that the post-launch numbers “reflect real, tangible business impact.” Adjacent teams in Data Platform and Marketing Hub went on to adopt UI patterns from Data Studio.
“[Quote from a beta customer or PM about the work]”