Neon Market

A tool people used once a quarter and dreaded

The existing goal management tool was built primarily to serve reporting needs. It captured the data leadership wanted, but the experience of setting and updating a goal was slow, deeply nested, and disconnected from the daily work it was supposed to describe.

The result was predictable: people filled goals in at the last minute to satisfy a deadline, updates went stale within weeks, and the data leadership relied on was often out of date the moment it was reported.

Design question

How do we redesign a system of record so that it stays accurate — by making the act of updating a goal genuinely useful to the person doing it, not just to the person reading the report?

How I approached it

Finding where people gave up

At this scale, behavioral data told the first story. I mapped funnel drop-off across the goal creation flow and found the majority of abandonment happened at a single step: alignment — choosing which parent goal a goal rolled up to. I paired that with interviews across individual contributors, managers, and org leaders to understand why.

What I learned

• Alignment was a guessing game. People could not find the right parent goal, so they picked something plausible or gave up.
• Updating felt like paperwork. The update flow asked for narrative text but gave nothing back to the user.
• Goals were invisible between cycles. If you never see a goal for three months, it stops being real.
• Managers duplicated everything. They maintained shadow trackers because the tool did not show team-level progress usefully.

A system of record only stays accurate if it's also useful

Every prior improvement attempt had focused on reminders and deadlines — pushing people to update. But the root cause was that updating produced no value for the person doing it.

The reframe: make the goal view the place where you can actually see your own progress, and updating becomes something people do because it helps them, not because they were nudged. Accuracy becomes a side effect of usefulness.

Rebuilding the alignment model

Alignment was the highest-leverage problem, so it got the most exploration. I tested a searchable flat list, a guided org-tree browser, and a recommendation-first approach that suggested likely parent goals based on team and prior cycles.

Testing what we chose

Suggestions-first tested strongest, but only when users could see why something was suggested. Without that context people distrusted the recommendation and fell back to manual browsing — so I designed the reasoning to show inline with each suggestion.

Three shifts that changed behavior

Suggested alignment

Instead of browsing an org tree, users see a short list of likely parent goals with the reasoning shown. Manual search remains available, but most users no longer need it.

Lightweight progress updates

Updates shifted from long-form narrative to a quick status plus optional context. The tradeoff was deliberate: slightly less detail per update in exchange for far more frequent, more current data.

Team progress view

Managers get a real team-level rollup with trend over time, which removed the reason to maintain shadow spreadsheets.

What changed

• Steps to create and align a goal dropped roughly 45%.
• On-time goal updates rose about 30% per cycle.
• Alignment abandonment fell substantially, improving rollup data quality org-wide.
• Patterns from the team view were reused in adjacent planning tools.

Note: metrics shown are directional and rounded. Detailed figures are confidential.

What I'd do differently

Designing for 700k users meant I could not test my way to every answer — I had to reason from behavioral data and validate narrowly. That worked, but I was slower than I should have been to trust the quantitative signal early on.

The bigger takeaway: when a tool is mandatory, low usage quality is easy to misread as a discipline problem. It was a value problem. Fixing the value fixed the behavior.