Stay Nimble started as a skills-to-jobs matching tool: users could compare their CV skills against in-demand roles, but had no guidance on what to do next. The audience — people in low-paying jobs with long hours and little margin for error — faced high psychological barriers to career change: uncertainty, overwhelm, and low belief that a transition was even possible for them. Career coaches existed, but operated outside the product in Intercom, disconnected from the journey. And the business model was limited to local-authority contracts, with no direct-to-consumer revenue.

I contributed to discovery, UX design, and the end-to-end journey redesign within a broader product team. Within that, I drove the design of the five-step journey structure, the nudge-based re-engagement patterns, and the concept for the B2C coaching-block model.

I focused on the psychology of career change for this specific audience: transitioning roles requires sustained motivation, and our users had very little margin for failure: any friction in the product risked being fatal to their attempt.

HotJar and Google Analytics quant data showed the pattern: users arrived, then left. The qualitative picture explained why: they simply didn't know where to go first. The homepage offered skills comparison and job listings, but no path. It was like handing someone a map of a city without telling them which street to walk down. The insight: users needed a structured journey, not another dashboard.

A five-step guided journey

I proposed structuring the product around five sequential steps — identify your skills, discover your strengths via a psychological assessment, do some learning, book a coaching call, search for jobs — each with a clear next action and no dead ends. The sequence was designed to build confidence progressively, not demand a leap from assessment straight to application.

Nudge-based re-engagement

Given the audience's limited time and low baseline confidence, I designed nudges that prompted users to continue at their own pace if they stalled mid-journey, with coaching offered as a defined block rather than an open-ended commitment.

Drip-fed learning via Open University

We partnered with the Open University to surface relevant courses inline with career recommendations at the learning step, matched to the gap between a user's current skills and their target role, and pacing-friendly rather than requiring a traditional course enrolment.

Embedding jobs and coaching in the journey, not beside it

Indeed-powered job listings appeared within the career page itself, and career coaching — CV reviews, LinkedIn reviews, interview practice — was offered as timed blocks within the same flow. A user could go from assessment to course to coaching to application without leaving the product.

Designing for a B2C pivot

The original model relied entirely on local-authority funding. I contributed to designing the subscription and coaching-block layer that let the same journey be sold directly to users, opening a revenue stream independent of public-sector contracts and extending reach beyond the original geographic catchments.

The first version was passive: it showed skills, compared jobs, and expected users to connect the dots themselves. For a cohort already daunted by the idea of career change, that was enough to make them quit.

I countered with the HotJar data: users were already dropping off precisely because they had no direction. The linear journey wasn't a constraint, it was a guardrail. We kept the original skills comparison, but it became step one of five, not the whole product.

• User base grew from 1,000 to 11,000 (1,000% growth)
• Journey completion rate increased from an estimated 12% to 38% within two months of relaunch
• 87% of users engaged with career recommendations; coach pairings increased 43%
• Helped open a B2C subscription model alongside the original local-authority contracts, with Open University and Indeed partnerships closing the skills-to-employment gap

• Quantify drop-off rigorously before and after the redesign, rather than relying on estimates
• Bring career coaches into the product design process earlier, so their content is structured around the five steps from day one
• Test the subscription model with users before committing engineering resources to billing infrastructure
• Build in flexibility for users to skip steps they've already completed elsewhere, sooner than we did