Rostering, time and attendance · working file 1 of 2
Interview synthesis
Four methods, one service. The interviews, the role-branched survey, the restaurant visits and the support-queue analysis all landed on the same problems, which is what made each finding defensible in the vendor conversation.
Method, four converging
Each method was chosen to cover a gap in the others. The survey reached the people who were never interviewed. The visits turned reported problems into observed ones. The ticket queue showed what people gave up on and rang about. Every role was covered by at least two methods, and where a survey branch returned nothing the interview round covered the role instead.
Role spread
| Role | What the evidence added |
|---|---|
| Team members | Phone-first behaviour; no view of their own breaks or clock-ins; some did not know the platform showed pay at all. |
| Shift supervisors | Most adjusted forgotten clock-ins themselves; break handling at the five-hour boundary was where the rules got bent. |
| Rostering managers | Copy-week rostering; the five-hour auto-break misfiring; the deployment chart printed and rewritten by hand. |
| Restaurant managers | Rostering craft; labour-cost estimates that drifted at scale; drops, swaps and the workaround gallery below. |
| Area coaches and a franchisee | Per-restaurant labour compiled by hand; payroll and platform figures disagreeing; exceptions spot-checked manually. |
| IT support | Registration failures; integration latency; open shifts excluded from hours; version deployment. |
| People-systems | Implementation history; feature enablement varying by restaurant; the training gap; compliance mapping. |
| Payroll, equity and franchise | Pay-period rigidity; the phantom-shift correction; leave balances updated only at the pay run. |
The survey's area-coach and franchisee branches returned nothing. That was read as a sampling signal rather than a finding: those roles do not fill in surveys. The triangulation rule caught it, and the above-restaurant interview round, with a franchisee, two area coaches and payroll, covered both roles.
Pain-point taxonomy
Every pain point was classified twice: once by solution type, once by backlog status. The two-way split separates what the vendor should own from what the business should own from what needs a different owner entirely.
| Solution type | Share of pain points | Meaning |
|---|---|---|
| Technology | 41 per cent | Platform features, configuration, mobile parity |
| Education | 26 per cent | Training and communication, a category with no delivery vehicle at the time |
| Integration | 21 per cent | Data flow between systems: latency, syncs, feeds |
| Process | 11 per cent | Rules and workflows, including the rules that create workarounds |
| Other | 1 per cent | Unclassified at triage |
| Backlog status | Owner |
|---|---|
| Already planned | On the vendor's release calendar, itemised |
| Solvable but unplanned | Vendor or business, with no backlog item yet |
| Outside the platform's scope | Education, process change, other systems |
| Unclassified | Reviewed at the half-yearly gate |
Workaround archaeology
The visits made the findings physical. Every recurring workaround is a missing feature made visible, and the gallery became the prioritised design brief.
| Workaround | Where it lives | What it proves |
|---|---|---|
| Paper deployment chart transcribed from the platform once or twice a day | Back office, near the schedule | Deployment generation is missing |
| Whiteboard above the time clock for breaks | Kitchen wall | No break visibility for team members or managers |
| Spreadsheet roster kept alongside the platform | Manager's desk | The readable roster does not exist in the product |
| Printed break sheets | Shift handover | Break compliance depends on paper memory |
| Missed pay added as an empty shift in the next week | Payroll correction | Reopening a pay period is effectively unavailable |
| Shared logins to fix one's own timekeeping | Manager devices | Managers cannot edit their own records |
Working file from employment at KFC South Pacific, anonymised: no restaurant, colleague or vendor is named, and the figures are the ones the case study states. Back to the case.