When we onboard new teams, they all swear their workflow is unique. Then we draw their stages on a whiteboard and it's the same five buckets, every time.
Here they are. If your board is missing one, that's usually where work goes to die.
1. Intake
The new thing has just shown up. It doesn't have an owner yet. The clock for the customer just started, but the clock for your team hasn't. This stage's only job is to close that gap fast.
- Repair shop: "Customer dropped off device."
- IT desk: "Ticket arrived in shared queue."
- Agency: "Brief submitted, no PM assigned yet."
- Field ops: "Job request created, not dispatched."
SLA: <1 hour. Anything longer and the customer is already mad.
2. Diagnostic / scoping
The work is now owned by someone, but you don't yet know what the work actually is. You're sizing it. Without this stage, teams jump straight to doing — then halfway through they realize the job is twice the size they thought, and the SLA on the "doing" stage was wrong from the start.
- Repair shop: diagnose the issue, prepare a quote.
- IT desk: triage the ticket, determine severity and owner.
- Agency: discovery / brief refinement.
- Field ops: site visit, confirm scope before booking the actual job.
Common mistake: conflating this with stage 4 (doing). Keep them separate so your "doing" numbers are accurate.
3. Waiting (parts, approval, customer)
The team is blocked. You're not actively doing the work, you're waiting on something or someone external. Most teams don't track this as a stage and it's the single biggest source of dishonest metrics.
- Repair shop: awaiting parts, awaiting customer approval on quote.
- IT desk: waiting on user info, waiting on vendor.
- Agency: waiting on client feedback.
- Field ops: waiting on access, waiting on materials.
Critical: SLAs in this stage should pause when you're waiting on someone outside the team. Otherwise your "at-risk" column fills up with jobs that aren't your fault, and nobody trusts the dashboard.
4. Doing
The actual work. The tech is at the bench. The dev is in the IDE. The designer is in Figma. This is the only stage that scales with skilled-people-time — the rest is operational overhead.
- Repair shop: in repair.
- IT desk: in progress.
- Agency: production / build.
- Field ops: technician on site.
Pro tip: if a job is in this stage for >2× its SLA, something is wrong with the work itself — your tech needs help, scope grew, or the original diagnostic was wrong. The right escalation is to pull a senior person in, not to extend the SLA.
5. Verify & close
The work appears done. But it's not really done until somebody else confirms it. Skip this stage and you'll get repeat tickets, callbacks, and customer escalations that should have been caught in 5 minutes.
- Repair shop: QC the repair, contact the customer for pickup.
- IT desk: verify the fix, get user sign-off.
- Agency: internal review, then client sign-off.
- Field ops: post-job photos, customer signature.
Common mistake: rolling this into "Done." You lose the ability to measure first-time-fix rate, which is the single best leading indicator of customer satisfaction.
What teams add (and shouldn't)
Teams love to add stages. We've seen 14-column boards. Almost every extra column is a sub-state of one of these five — "Diagnostic - awaiting senior tech" is still Diagnostic. Use card metadata for sub-states, not columns. Columns should be the high-level shape of the work, not every possible state a card can be in.
The five-bucket pattern lets the team scan a board in 3 seconds: this many in Intake, this many in Diagnostic, this many Waiting, this many Doing, this many at Verify. A 14-column board takes 30 seconds to scan and gets ignored after week 2.
If you're missing a stage
- No Intake → customer complaints feel random and unfair.
- No Diagnostic → SLAs miss because work is bigger than estimated.
- No Waiting → metrics lie about team performance.
- No Verify → callbacks and repeat tickets eat your throughput.
Most workflows we audit have 3 of these and feel chaotic for the missing 2. Add the missing ones and the chaos drops within a week.