← Back to case
Research report · PRIME framework

Share price simulator — discovery sprint

What we set out to solve, how we approached it, what the evidence revealed, and the moves the team is making next.

P — Problem

People open the simulator twice a year and walk away with more questions

The current simulator answers one question — “what if the share price moves X%?” — but participants arrive with five. They want to know when they can sell, what tax looks like, what happens if they leave, how much of their grant is real today, and whether to act now or wait. The slider front door doesn't fit the way people think.

“I open it, look at the slider, and immediately go ask HR what it means for me. I don't think the tool was meant for me.”

— Participant 03, Norway, 2 years tenure
R — Research

Five inputs, twenty-seven participants, one transaction-list-style synthesis

The discovery sprint ran for two weeks across April 2026. One stakeholder interview opened the sprint with the product lead on the Equity Platform. Six user interviews followed the BGOALC framework, scoped to Norway, UK, and Germany participants with one to four years tenure. A sixteen-response survey validated the behavioural patterns surfaced in interviews. Four usability test sessions tested the rebuilt co-pilot prototype against the four most-asked questions: what's it worth, when can I sell, what if I leave, and should I sell.

I — Insights

Three patterns the evidence keeps returning to

Insight 01

The slider front door makes people feel like the wrong audience

Participants don't think in percentages — they think in life events (“when can I sell?”, “what if I leave?”). The first interaction asking for a percentage signals “you're not the user we built this for,” which is why people close the tab.

Insight 02

“Twice a year” means zero memory of how the tool works

Recall between sessions is near-zero. Every visit is a cold start. Tooltips and onboarding don't survive the gap. The tool has to make sense without prior knowledge each time.

Insight 03

Refusal-to-recommend reads as respectful, not evasive — if the redirect is concrete

When the co-pilot said “I can't recommend, but I can compare scenarios for you,” every usability participant took up the redirect. None read the refusal as the tool being unhelpful. The condition is the redirect being specific.

M — Material

Evidence at a glance

Key numbers that give the findings their weight, drawn from all five research inputs.

4/4
Usability participants took up the co-pilot's refusal redirect
6/6
User-interview participants opened the tool twice a year or less
13/16
Survey respondents said tax-by-country context would increase confidence
3/4
Usability participants asked about timing before they asked about price
11/16
Survey respondents wanted notifications when a window opens or closes
High
Volume of HR tickets about “has it actually unlocked?” — flagged, not quantified
What people ask the co-pilot for vs. what the old tool currently provides
Users ask for it Currently provided
Timing & sell windows Tax for my country What happens if I leave Window alerts
E — Execution

What the team is doing next

Prioritised against user need, business impact, and engineering effort. P1 ships in the next sprint; P2 in the following one.

  • P1
    Replace the slider front door with a question box. Default surface becomes “ask anything about your shares.” The slider stays available, but as a secondary tool. Wires directly to the co-pilot prompts already validated in usability.
  • P1
    Co-pilot answers timing questions before price questions. When the tool can't tell which the user wants, it leads with timing. Three of four usability participants asked about timing first.
  • P2
    Country-aware tax explanations as a default panel. Surface the participant's actual country and show one comparison alongside (UK or NO, depending on contract). Don't bury it behind a toggle.
  • P2
    Notifications for unlock/blackout window changes. Email + in-app, both opt-in by default. Eleven of sixteen survey respondents asked for this directly.
  • P3
    Quarterly recap email. Closes the twice-a-year gap by bringing the tool to the participant when something has actually changed. Test against open-rate before scaling.
Summary

If you read nothing else

People open the simulator with life-event questions and a near-zero memory of how the tool works. The slider front door fits neither. The rebuild leads with a question box wired to a co-pilot that explains timing before price, surfaces country-specific tax by default, and refuses to recommend — clearly, with concrete redirects. The team is shipping P1 in the next sprint.