Productising a core business stream
Transforming offline PDFs and CSVs into a new, productised in-app experience
Context
Demand for Sylvera's early-stage market expertise was surging, but the solution lived entirely offline in CSVs and PDFs — making it hard to scale or sell as a renewing subscription. Our five-person squad (GM, PM, myself as designer, & 2 engineers) set out to bring these insights into the platform as a core, interactive feature.
Our first step was exploring how text-heavy, ~50-page reports could be turned into experiences worthy of a modern SaaS platform.
Details when needed, not upfront
To reduce cognitive load and move away from the PDFs' "wall of data" approach, I surfaced key insights through charts and interactions, deferring technical detail to tooltips and footnotes for on-demand access.
Tools not text
The report's recommended actions were redesigned into an interactive planning tool within the SaaS platform, enabling users to model trade-offs, forecast outcomes, and make data-informed decisions to strengthen project performance.
Making the cut
The exploratory phase ended with a tough reality: an engineer-light squad and impending quarterly goals. To deliver as quickly as possible, we focused the MVP on bringing a CSV of early-stage project data into the platform as an interactive table. Relatively unglamorous, but the right decision:
- The content would be continuously added-to, driving logins and justifying subscriptions.
- We could reuse existing frontend components, giving the best return on limited engineering resource.
Design system for rapid delivery
Leveraging a flexible, front-end-aligned table component I'd recently designed for the Sylvera design system, I was able to help deliver a scalable UI for the catalogue in record time. I streamlined the layout by merging related CSV fields into multi-element columns — minimising horizontal scrolling and making side-by-side comparison effortless.
Column order was optimised around the decisions users needed to make once they'd narrowed their list to a shortlist, ensuring the most relevant data was always in view.
Ideally, we'd have shipped an alternative card view for scannability, but speed-of-delivery had to win out.
The best UI is the one you don't think about
Catalogue filters used sliders — often a source of frustration. To avoid this, I defined their behaviour in detail, from increment scale to micro-interactions. The result was a seamless flow that kept users focused on exploring opportunities, not wrestling with controls.
Macro-components for faster dev cycles
Where possible, the early-stage project page design mirrored the mature-project structure while accommodating its unique data points. I used the opportunity to harmonise all existing project page designs, distilling seven distinct page types into a single flexible design-plus-frontend responsive component. This cut design and engineering overhead and sped up later dev cycles.
Project reflections
Wins
- Our OKR was three early-stage sales in the quarter. With the productised experience launched, we more than doubled that — closing seven deals and proving the value of the strategy.
- The harmonised project page component cut dev and design cycles and improved UX consistency across the platform.
- The project was such a success that the company org structure was overhauled to replicate the success of our team.
Learnings
- The original design vision was overly ambitious, forcing large scope cuts to deliver an MVP within the quarter. I could have anticipated this and invested less polish and detail in the initial exploratory designs.
- Embedding filters in column headers would have made the table cleaner and easier to use. The above-table placement came from the original vision, which allowed for switching between table and card views with shared filters above both. In hindsight, I could have anticipated that competing priorities might drop the card view, and focused on perfecting the table experience from the start.
