Clients historically relied on Operations teams to change existing bookings, creating unnecessary calls, slower resolution and significant internal workload.

I identified which amendments could safely move to self-service and led the design and delivery of complex assigned and unassigned workflows, including the business rules and edge cases required to protect already-confirmed workers.

​

I also reworked the measurement model so digital adoption was compared only against amendments clients could realistically self-serve, rather than against all manual activity.

Impact

• 48% of eligible assigned amendments completed digitally

• 30% of eligible unassigned amendments completed digitally

• ~900 manual edits replaced in June 2025

• ~90 operational hours saved in June

• confirmation +34%; cancellation −38%

Following major company redundancies, I inherited a booking-amendment programme whose committed scope and timeline no longer matched the engineering capacity available.

​

Rather than reducing the disagreement to “Engineering needs more time,” I developed six rollout options and compared them across time-to-value, testing, customer exposure and release risk.

I recommended separating the lower-risk unassigned-booking workflow from the significantly more complex assigned-booking journey.

​

Outcome

• leadership adopted the hybrid rollout strategy

• lower-risk functionality reached clients earlier

• higher-risk assigned workflows were tested with a controlled client group before broader release

• the phased approach ultimately delivered the self-service outcomes above without exposing the entire client base to the highest-risk changes at once

Leadership proposed removing a booking control that allowed clients to restrict shifts to healthcare workers who had previously worked at their service.

Before the change, I analyzed whether the control was unnecessary complexity or whether it represented a genuine client need.

​

The evidence showed:

• it was used in ~32% of booking-creation journeys

• it was the 5th most-clicked action on the booking screen

• it generated roughly 1,400 interactions per week

• for some care settings, continuity was a real operational requirement rather than a preference

​

I recommended a graduated approach that reduced habitual use while preserving continuity where it genuinely mattered. The broader removal went ahead.

After the change, HFC’s coverage rate fell from roughly 75% to 65%, alongside reduced workforce visibility and a shift toward more targeted assignment. Because several changes landed together, I treated this as a strong signal rather than attributing the full decline to one feature.

​

I then led a redesign around the underlying client need: help clients quickly find people they trust without unnecessarily excluding the wider workforce.

The new experience simplified staff selection, opened targeted invitations to new/local workers and made the same selection model easier to use across block bookings.

​

Impact

• HFC coverage recovered from ~65% to ~79% in early 2026, with my changes contributing to that recovery

• one-click staff selection reached 27.3% of per-booking actions

• recommended staff-card interactions increased 447 → 3,098 (+593%)

• HFC booking creation was +0.8% vs the final full pre-change week, while total bookings were +1.8% in the first complete post-release week

By May 2026, HFC still accounted for 65% of all client staffing hours and achieved 79% coverage, slightly ahead of manually created demand at 76%. Yet HFC-created hours had fallen from roughly 80k to 29k per week (-63%) since early 2024.

​

Yet HFC-created hours had fallen from roughly 80k to 29k per week (-63%) since early 2024.

I decomposed the decline across booking channel, volume, booking length, coverage and product engagement, then combined the analysis with 10 interviews across clients, Operations and commercial teams.

The evidence showed that poor coverage was not the main problem — coverage had actually improved while digital demand continued to fall. The strongest recurring issue was client trust in workforce quality and confidence that concerns would be acted on.

​

Outcome

• reframed the problem from “improve coverage” to rebuild genuine client preference for the digital channel

• identified workforce quality, trust and complaint closure as the strongest product opportunity

• secured executive backing for staff feedback and quality as a strategic priority

• defined a cross-product Staff Feedback, Quality & Complaint Management model connecting client feedback, workforce signals and operational action

The founder wanted to extend booking amendments so workers could counter-offer a different time when a client edited an assigned booking.

Instead of treating the request as a commitment, I analyzed whether the problem was large enough to justify the engineering investment.

​

The data showed that workers were already accepting around 92.5% of edit requests requiring a response, while only ~5% were declined. The existing decline/reassignment path already handled the edge case.

I mapped the workflow and showed that adding a negotiation layer would introduce additional booking states, client delay, operational complexity and engineering cost for a very small user problem.

​

Outcome

• the founder agreed not to build the feature

• protected limited engineering capacity for higher-value work

• demonstrated that successful Product work can mean preventing unnecessary complexity, not adding another feature

• Moved booking amendments into digital self-service, reaching 48% adoption for eligible assigned changes and 30% for unassigned changes, replacing ~900 manual edits and saving ~90 operational hours in June 2025

• Turned a high-risk delivery disagreement into a structured rollout decision, creating six options and securing leadership agreement on a phased approach balancing time-to-value and customer risk

• Redesigned client staff selection after a high-risk product change, contributing to HFC coverage recovering from ~65% to ~79% while the new one-click journey reached 27.3% of per-booking actions

• Diagnosed a 63% decline in demand through Newcross’s primary digital client channel despite improving coverage, shifting product strategy toward client trust, workforce quality and complaint closure

• Used product data to stop unnecessary complexity before it was built, demonstrating that prioritization sometimes means protecting engineering capacity rather than adding another feature