The story in 30s
The problem. Entromy's report page, the flagship surface of the business, was essentially a literal mirror of our PowerPoint export—rendering decks that could approach 100 pages in some cases, with each individual slide displayed as a static card in the UI. Over 96% of report views followed a pattern of open, export, leave. Support tickets and executive debriefs exposed a critical commercial threat. First, clients relied almost exclusively on exported PowerPoint and Excel files because scrolling through dozens of static, slide-like cards to locate basic departmental metrics was too painful. When enterprise buyers treat a SaaS platform as a one-time download portal rather than an analytical workspace, renewal conversations become vulnerable to churn. Second, C-level buyers explicitly called the web UI "dated"—a perception gap in Private Equity diligence and board reviews where visual polish is a proxy for data credibility. Underneath that, hardcoded, one-off components inside a rigid slider engine slowed engineering velocity to a crawl, making every customer request expensive to build.
What I did. Instead of pitching a risky overhaul blindly, I framed this around protecting ARR and sequenced the investment iteratively. First, we de-risked the bet with a focused homepage PoC, proving UX modernization could drive a 41% lift in views and a 144% jump in engagement time—unlocking leadership buy-in for the core reporting rebuild. Next, stepping into a dual design and product role for Reports 2.0, I aligned engineering on migrating legacy elements into reusable building blocks before layering on interactivity. I authored the platform's first data visualization guide, replaced the endless feed of slide-like cards with 4 structured tabs and dynamic modules, and negotiated pragmatic scope splits across engineering and QA to ship value fast without stalling the roadmap.
What changed. The PowerPoint and Excel exports stayed right where they were—executives and analysts still rely on them. What changed is the commercial value of the web platform alongside them. Executive debriefs shifted from spending 45 minutes hunting through pages to live, in-app strategic discussions, cutting debrief friction and strengthening renewal positioning. Over half of reporting sessions now run past ten minutes, reusable components cut engineering build time across sibling products (360 and VCP), and client-facing teams call the redesigned UI "a game changer" for C-suite sales and retention conversations. With users finally staying in the app, I conceptualized our next roadmap leap—an embedded AI Advisor—and instrumented PostHog analytics to track its impact.
The thinking: a system before any screens
Before committing engineering sprints to the reporting engine, we needed proof for leadership that better UX would alter entrenched offline habits. Shipping a focused modernization on the homepage first gave us clear commercial validation: views per user rose 41% (11.39 to 16.04) and average engagement time jumped 144% (3m 16s to 7m 58s). Design alone could shift behavior, earning us the mandate for Reports 2.0.
Even with buy-in, the biggest business risk wasn't visual design: it was burning engineering quarters building on top of the hardcoded foundation that caused the bottleneck in the first place. The first real decision was getting product, design, and engineering aligned on migrating legacy elements into reusable components—like turning flat, static lists into sortable tables wrapped in modular cards—before adding deep interactivity. Once that foundation existed, the data visualization guide gave the redesign a shared vocabulary:
- → One model, not six. Every surface follows Shneiderman's mantra—overview first, zoom and filter, details on demand: users see overall context and patterns before specifics, narrow down with filters, and pull detail only when they ask for it, instead of scrolling through up to 100 static, slide-mirrored cards.
- → Depth by role, not by guess. Decision-makers (CEOs, CHROs, PE operating partners) get high-level scorecard views built for fast reads. Analysts (managers, consultants) get detailed, sortable views built for exploration. This mapping came from actual debrief behavior, not assumptions about titles.
- → Same data, same chart, everywhere. Gauges and indicators for scores, line charts for trends, stacked bars for distributions, ranked lists and tables for comparisons, so the same kind of data always looks the same way anywhere in the product.
- → Color means something, or it isn't used. Categorical palettes for unordered groups, sequential for progressions like participation, and divergent palettes reserved for data with a meaningful midpoint, like eNPS running from -100 to 100.
Navigating reality: pragmatic decisions and trade-offs
Moving from a 1:1 PowerPoint mirror to a 4-tab interactive hub (Business Overview, Business Risks, Top Comments, and Interactive Appendix) meant making deliberate scope calls to protect sprint velocity and time-to-market:
- → Visual layouts first, deep modals second. Multi-dimensional drill-downs required heavy backend aggregation queries that risked stalling delivery by months. Aligning with leadership, we shipped the modern card containers and clean navigation first to close the C-suite credibility gap immediately, deferring complex multi-hop modals to later phases.
- → Collapsing repetitive appendix pages within scope limits. I replaced dozens of nearly identical, slide-style appendix cards with two dynamic modules powered by dropdown selectors and sortable tables. When mid-sprint requests for custom multi-line headers threatened our release window, I backed engineering's pushback—locking the spec to stabilize core sorting and scrolling mechanics first and spinning non-essential styling into follow-up tickets.
- → Phasing releases with QA. Bundling eNPS breakdowns, detractor tables, and driver analysis into one monolithic release created a testing bottleneck. Partnering with QA, I broke the rollout into four controlled visual milestones, shipping cleanly to enterprise clients with zero staging regressions.
- → Adapting to backend privacy limits. My initial spec showed all categories in dropdowns, disabling sub-threshold cohorts (<5 respondents) with a privacy tooltip. When engineering flagged that the API couldn't distinguish between suppressed, zero-response, or non-existent categories without a multi-week refactor, I simplified the rule: dynamically render only valid, above-threshold data.
- → Decoupling Web UI from PPT settings. Toggling report settings used to leave clients confused about what changed where, driving recurring support tickets. Rather than forcing the web app and slide deck to match 1:1, we separated their jobs: Report Settings govern boardroom PPT exports, while the web UI stays persistently rich and interactive.
Solution overview
The clearest illustration is the eNPS flow, three levels deep, each one earning the next. Learning this pattern once unlocks the entire report. That mattered enormously for an enterprise product leaders open a few times a year, where every visit is close to a first visit.
- → Overview: enough to stop there. The dashboard card shows the score, its quartile, and movement against benchmarks. A decision-maker can stop here, fully informed.
- → Zoom and filter: one click deeper. Clicking through opens the consolidated Business Risks view: promoter/neutral/detractor breakdowns, detractor insights ranked by gap, and the dimensions where detractors concentrate.
- → Details on demand: invisible until asked for. From there, question-level breakdowns, individual comments with targeted AI summaries, and recommendations open in popovers and collapsible cards—present when asked for, invisible when not.
Business & Product Impact
- → Eliminated the "download-and-leave" churn signal: Over half of reporting sessions now exceed 10 minutes, shifting the platform from a passive file portal into an active workspace where executives and consultants analyze data directly on-screen during debriefs.
- → Sustained adoption & peak usage cycles: Drove multiple "strongest week on record" cycles post-launch, with report exploration clicks and informed exports both jumping ~90% week over week at peak—confirming that in-app discovery fuels broader organizational sharing.
- → Validated product stickiness across surfaces: Built on our initial homepage PoC (which lifted views per user by 41% and engagement time by 144%), cutting navigation bloat by replacing sprawling slide-by-slide pages with structured tabs nav—while org admins, our primary renewal buyers, now drive 93% of report activity.
- → Restored enterprise credibility & R&D leverage: Eliminated "dated UI" objections in sales and Customer Success calls, while the new modular card and table architecture reduced engineering effort to roll out matching upgrades across 360 Feedback and Value Creation Plan (VCP) products.
Roadmap evolution: the AI advisor
Once the modular foundation succeeded in keeping users inside the platform, a new strategic question emerged: how do we shorten the executive path from data exploration to synthesis even further?
As the next step on the roadmap, I conceptualized an AI Advisor—combining automated takeaways with contextual questions and answers so executives could query their results conversationally alongside visual drill-downs. Within months of going live, the advisor was adopted across roughly 65–70% of active report opens—with nearly 45% of sessions driving multi-turn follow-up prompts, showing strong traction as a natural conversational layer on top of the redesigned reports.
Reflection
- 1 The foundation had to come before the features. Migrating hardcoded elements into reusable components, like turning a flat list into a reusable sortable card, wasn't flashy in the final UI, but it unlocked R&D velocity: six sections shipped as one cohesive product, and those exact building blocks immediately scaled to sibling product lines.
- 2 The goal was never to replace the export, it was to justify the SaaS renewal. Executives and analysts still pull PowerPoint and Excel files; those are real boardroom workflows and it wasn't our job to kill them. The commercial win was making the in-platform experience valuable enough—fast navigation, drill-downs, and later the AI advisor—that clients depend on the software itself rather than treating it as a file-download stop.
- 3 Consistency and scope discipline compound. Protecting engineering timelines through phased releases got value into clients' hands months earlier, and repeating one clear interaction model across every tab means that in a product leaders open quarterly or annually, the second section is already familiar by the time someone reaches it.