All work
Professional

2022 · Systems and information design for image quality control · 4MIN READ

Streamlining a Multi-Team QC Workflow

Resolution loop, one issue at a time

01

Issue reported

Any team

Date, style number, colour, view, urgency and source go in one row. Urgent rows turn red automatically until they carry a status.

02

Issue picked up

Retouching / DAM

The team that owns the fix writes its note on the same row - never a new one - so one issue keeps one history.

03

Status set

Retouching / DAM

Pending or Complete, with an update date. Locked entries turn pink until the lock is released.

04

Partner check

External partner

The partner confirms the fix landed. Unconfirmed rows stay visible in the pending filter instead of quietly disappearing.

!

If the check fails, the row goes back to step 02 with its full history attached - the loop that used to restart as a brand new email thread.

Redrawn from the onboarding manual I wrote for the teams. The external partner is referred to by role only.

Shared tracking sheet

UrgentLockedComplete
DateStyleColourUrgencySourceIssueTeam notesStatusPartner check
25-07072...313310UrgentStudioLabel not visible-PendingNo
08-08072...325690UrgentInternalColour differsPushed againCompleteYes
08-08082...312400Not urgentInternalWrong colour-PendingNo

Rules the sheet enforces on its own

  • Urgent rows turn red on their own, and stay red until someone sets a status.
  • Locked entries turn pink, and clear the moment the status becomes Complete.
  • Nothing is ever deleted: an unresolved issue is updated on its original row, which is what killed the duplicates.
  • The update date is mandatory on every status change, so pending age is readable at a glance.
Illustrative rows in the real column model. Style numbers are shortened and no partner or supplier is identified.

One sheet, six views

Retouching

Writes

Team notes, status, update date

Sees

Its own tab, filtered to pending

DAM

Writes

Team notes, status, locked in PIM

Sees

Its own tab, plus everything locked

Copywriting

Writes

Issue, urgency, source

Sees

Copy-related issues only

Fitting

Writes

Issue, urgency, source

Sees

Fit-related issues only

Product

Writes

Issue, urgency, source

Sees

Its own reports and their status

External partner

Writes

Partner notes and partner check

Sees

Complete rows awaiting confirmation

Each team got its own tab and saved filter, so nobody had to read a column that was not theirs - the part that made adoption possible across non-technical teams.

Process summary

How the work was made

01

Surveying how the process was really used

I asked every party involved - Retouching, DAM, Copywriting, Fitting, Product and the external partner - how they were reporting issues today and where it broke down. The answers did not describe one process; they described six, each with its own channel and its own idea of what 'done' meant.

02

Designing one shared tracking system

One sheet with a fixed column model and three status stages: Pending, Complete, Partner check. Urgency and status colour themselves automatically, each team gets its own tab with input permissions and saved filters, and an unresolved issue is updated on its original row rather than reported again.

03

Writing the onboarding manual

The system was only worth as much as its adoption, and the users were not technical. I wrote a step-by-step manual - one instruction per screen, in the teams' working language - covering data entry, urgency, status, locking and filtering, so anyone could start using the sheet without being trained on it.

04

Opening a feedback loop

The manual ends with an explicit invitation: tell me what to change, including in the other teams' tabs. Treating it as a living tool rather than a one-off delivery is what kept it in use as each team's needs shifted.

Participants

Five internal teams - Retouching, DAM, Copywriting, Fitting and Product - plus an external retouching partner, surveyed first and onboarded afterwards.

Result

Adopted and used simultaneously by all six parties, ending duplicate issue tracking and giving each step of the resolution a named owner.

What I would test next

Measuring time-to-resolution per issue and checking whether the urgency flag actually shortens it, rather than assuming visibility is enough.

Outcome

  • One shared tracking system replacing mail and chat reporting across six parties
  • Duplicate reporting eliminated by the update-the-original-row rule and per-team filters
  • Clear ownership of every stage: who reports, who fixes, who sets status, who confirms
  • An onboarding manual that let non-technical teams adopt the system without training sessions

What I took from it

That a spreadsheet can be a product if you treat it like one. The survey mattered more than the build: once I understood what each team was already doing, the system only had to formalise it. And writing the manual taught me that a design is finished when someone who was not in the room can use it unaided.

Microsoft ExcelGoogle FormsMicrosoft 365
Service DesignJourney MappingUser ResearchProcess DesignInformation Design

Next case study

Personal Finances