My Desk 2.0

Challenge

When someone signed into Trainual, it was unclear what to do next. The team had layered temporary patches over the gap — welcome wizards and product tours nudging customers toward content creation — but the metrics those patches were supposed to move didn't move.

The problem read differently from three different sides of the user base, and that was the point. Each role hit the same empty surface and asked a different question:

Entrepreneurs documenting a business needed a clear action item and a sense of progress against what they'd already done — without it, the platform felt uninformative and the next session didn't come.
Managers trying to read team performance needed visible accountability and a way to see Trainual's value showing up in their team's training — without it, the platform felt disconnected from the work they were already doing.
Technicians needed to be trained on their business's processes but couldn't tell what to work on — without that, the platform felt like a list of nothing in particular.

Three roles, one symptom: the home surface didn't answer "what now?" The hypothesis was that a centralized launchpad with a clear, actionable path could shift the activation metric the welcome wizards weren't moving.

Side-by-side of My Desk before and after. Left: the legacy My Desk — a grid of identical subject cards with no clear next action and no sense of progress. Right: My Desk 2.0 — Up Next panel on the left with a stepped task list, Completions ring on the right showing personal and company-wide progress, plus a mobile-side panel demonstrating the same components at small-screen sizes — The surface change in one comparison. The 'before' answered no question and surfaced everything; the 'after' answered 'what now?' first, then surfaced progress against the answer.

Execution

Modular widget architecture. The page was rebuilt as a composable grid rather than a static dashboard, so the surface could carry different content for different roles and grow into the longer-term product vision instead of getting redesigned end- to-end every quarter.

Two hand-drawn dashboard concepts side by side — left: 'Dashboard Concepts' with Chris's mockup showing a CTA panel, training todos, and a 'WHAT'S NEW' panel. Right: 'DASHBOARD' sketch with annotated zones for Announcements/CTAs, charts, columns, feed, with margin notes for 'What's New?', 'ROI', 'Fav's', 'Get Back to 100%', 'Getting Started' — The discovery sketches. Two early dashboard concepts thinking through which questions a launchpad should answer per role — the modular widget grid fell out of the realization that no single layout could serve all three personas at once.

Wireframes and sketches of the modular widget container system — annotated layout zones showing how widgets snap into a responsive grid, plus three example widget renders (a steps list, a chart, an emoji feed) — The widget component at two altitudes. The container architecture working out where in the grid widgets snap (left), and the shipped Add Widgets picker showing how customers compose their own launchpad from a categorized set (right) — Admin widgets, General widgets, the 'Coming soon!' placeholders for the longer-term vision the architecture was built to absorb.

The shipped Add Widgets picker modal — Admin category with Search Terms, Leaderboard (Coming soon!), and Test Scores. General category with Completion Rate, Training Time (Coming soon!), and Forum Activity (Coming soon!). Three preview widgets on the right showing Completion, Metrics, and Announcements — The widget component at two altitudes. The container architecture working out where in the grid widgets snap (left), and the shipped Add Widgets picker showing how customers compose their own launchpad from a categorized set (right) — Admin widgets, General widgets, the 'Coming soon!' placeholders for the longer-term vision the architecture was built to absorb.

Up Next as the activation component. The biggest pain point was "what do I work on?" Up Next collapsed that question into a single decision — here's your next thing, click it — and the metric set was built around how many people made that click and how fast they finished what came after.
First surface on the Palette System. My Desk 2.0 was the feature initiative the Palette System shipped inside of. Routing the color rewrite through an active product initiative — instead of asking for standalone systems headcount — was what got the foundational work prioritized at all. The new page was the proof the architecture was production- ready.

Two renderings of the same My Desk 2.0 page in different themes. Left: the default light theme with neutral surfaces and the Trainual purple accent. Right: a neon theme with deep purple/black surfaces, vibrant magenta and electric-purple data visualization, and the same component layout — My Desk 2.0 in two themes from the same Palette System architecture. The page on the left is the default; the page on the right is the proof that the theme layer could carry significantly different visual identities without rebuilding components — the Trojan-Horse claim made visible.

Async internal usability testing at scale. Validation ran as a large-scale internal async test rather than scheduled live sessions — appropriate to a distributed team and to a feature whose target behavior was specifically individual engagement, not guided walkthroughs.

A walkthrough of the async internal usability test setup — the format that replaced scheduled live sessions for this initiative.

Outcome

For customers who engaged with the Up Next component — the cohort the initiative was designed to move:

3× higher activity than the comparable non-Up-Next cohort.
20% faster step completion. The practical translation of "I know what to work on" into time saved.
Daily feature retention above 79%. High enough that Up Next qualified as habitual rather than novelty behavior.
Doubled feature conversion rate off the prior baseline.

A walkthrough of the shipped surface — the Up Next component, the modular widget grid, and the cross-theme rendering.

Underneath those numbers, the page shipped a structural change the team carried forward. My Desk 2.0 was the first production surface built on the Palette System foundation — the same architecture the Saguaro Design System eventually composed against. The activation work and the systems work paid each other forward.

Learnings

Subtract, don't surface. The temptation with a "home page" is to surface everything the user might want. The work was the opposite: subtract until one actionable next thing was unambiguous, then measure whether people took it.
Feature initiatives are the right vehicle for foundation work. Asking for resourcing for a "color system rewrite" was a non- starter. Asking to ship a refreshed home page that happened to require the new color foundation got both prioritized. The Palette System case study makes the same point from the other side — My Desk 2.0 is what proved the framing worked.
Three personas, one surface. Designing a single page that answered "what now?" for entrepreneurs, managers, and technicians forced the modular widget architecture. The same architecture is what let the page evolve afterward without another redesign.