Rescuing analytics from the cut list
How user research reversed a decision to strip analytics out of a B2B engagement platform.
User Research
〰️
Information Architecture
〰️
Data Visualisation
〰️
Design Systems
〰️
Design Systems
〰️
User Research 〰️ Information Architecture 〰️ Data Visualisation 〰️ Design Systems 〰️ Design Systems 〰️
Role
Sole designer - research to shipped design
Status
Design complete · in staging
Scope
20+ components · 3 dashboards
Impact
Reversed a decision to cut the feature
Xtremepush is an omnichannel engagement platform used by enterprise marketing teams. Its standard analytics, the dashboards showing campaign, user and event performance, are available to every customer, unlike the paid Intelligence tier. They're a core part of the everyday product and, indirectly, part of how the platform proves its value.
The context
By the time of the platform-wide redesign, standard analytics were running on old technology with a confusing information architecture: hard to find the right metric, charts limited to bar graphs that hid trends, and a UI too dated for Sales to confidently demo.
The problem
The decision I pushed back on
A content audit at the start of the redesign showed the analytics section had low engagement. Based on that, the initial product direction was to strip out the underused analytics and move on.
Low engagement on a badly-structured, dated feature doesn't tell you the feature has no value, it might just tell you the feature is hard to use.
Removing it risked cutting something customers actually relied on, based on a symptom rather than a cause. So instead of accepting the removal, I proposed we find out what users actually wanted from analytics first, then make the call.
My role
Sole designer on this work. I ran all of the research and owned the end-to-end design:
Proposed and designed the analytics survey, and analysed the 74 responses
Led the information architecture for the redesigned analytics section
Designed the full set of 20+ dashboard components across the new structure
Worked with the Product Manager to align findings to requirements, and with developers on feasibility and data accuracy
The pivotal decision
Let the users decide
I ran a survey to test the assumption behind the removal. 74 customers responded. The results directly contradicted the plan to cut analytics.
Analytics weren't unused. They were used, wanted, and underserved by the current design. The survey also did something more useful than just save the feature, it told us which analytics mattered, so we could cut the right things instead of everything.
The insight: users prioritise what people do (events) and what it's worth (revenue), and care least about location data. The original instinct to remove underused analytics was directionally right about some content, but wrong about the section as a whole.
From insight to design
Each survey finding mapped to a design decision.
Three clear dashboards, Campaigns, Users and KPIs, so users could find the metric they needed instead of hunting.
Reorganised the IA
The highest-rated categories got the most prominent placement and the richest views, including campaign comparison and channel performance.
Led with events and revenue
Funnels now run on engagement events, not just tags. This was a specific gap users flagged: “100 people saw this email, 50 clicked, 25% used the promo code. I want to target the ones who clicked but didn't convert.”
Rebuilt funnel analysis
Added channel comparison
So teams could weigh channel performance against each other and make budget decisions.
Deprioritised location data
The lowest-rated category no longer gets equal weight, freeing space for what matters.
Moved beyond bar charts
Chart types that actually surface trends, addressing a core complaint from the audit.
The redesign delivered 20+ dashboard components across the three dashboards, built in collaboration with development against real data constraints.
The outcome
The analytics section was retained and fully redesigned rather than removed — a direct result of the research.
The survey findings shaped the product requirements for the redesign, aligning what got built with what users had actually asked for.
The redesigned dashboards have been demoed to enterprise clients, with positive feedback- the same audience the old UI wasn't good enough to present to.
The work reframed analytics as a trust and upsell lever: strong, well-designed free analytics builds confidence in the platform and open a natural path to the paid Intelligence tier.
Current status
The redesign is built and in staging ahead of a phased rollout; production metrics aren't available yet. The proven outcome here is the research itself: evidence reversed a product decision and redirected the design toward what users actually valued.
The most important decision in this project wasn't a design one, it was choosing to challenge an assumption before it became a deletion. It would have been easy to accept “low engagement = remove it.” Getting the evidence first meant we kept a feature customers depend on, and redesigned it around what they actually use rather than what we guessed.
Reflection
A designer's job isn't only to execute the roadmap. It's to make sure the roadmap is pointed at the right problem.