Anita has contacted IT times since May. of those trace to a single cause — an NVMe drive that has been failing since her first ticket on 11 May and has not been replaced. Along the way she was routed to Networking times, abandoned a chat at 11:40pm during quarter-close, and lost days waiting on an SAP approval that sat unactioned in a queue. She has never filed a ticket about the application crashes her laptop has logged.
Written from 59 signals · 4 episodes
Lifeline
Ninety days, every channel she used, laid over what her laptop was doing underneath.
How each contact travelled, and where it was mis-routed.
Journey flow
Every contact she made, from where it started to how it ended.
Root cause
One failing component, and everything it produced downstream.
One cause. Fourteen contacts. Nine of them avoidable.
Chapters
4 episodes in this period, newest first.
08 JUL — 09 JUL
1 day
Resolved
Root causeInferred
Entitled workflow was undiscoverable in search
What we got right
What we got wrong
Time spent — hers vs ours
0.3h hers
0h ours
Virtual agentEnterprise searchWorkflow
6 linked signals
How she got here
Searched enterprise knowledge
2 attempts, then asked
0.1 hrs
Opened the virtual agent
Answered by the agent
0.1 hrs
Raised it in the portal
two records raised
0.1 hrs
Ran the entitled workflow
Automated, no human touch
0.1 hrs
Closed out
Resolved, no recurrence
1 day
Coral marks a dead end.
29 JUN — 01 JUL
1 day
Worked around
Root causeInferred
VPN gateway instability on consumer networks
What we got right
What we got wrong
Time spent — hers vs ours
2.1h hers
0.7h ours
Service portalVirtual agentEnterprise search
5 linked signals
How she got here
Searched enterprise knowledge
3 attempts, no result
0.1 hrs
Opened the virtual agent
Abandoned at 23:41
0.8 hrs
Advised a workaround
Cause never investigated
1 day
Coral marks a dead end.
15 JUN — 06 JUL
21 days
Resolved
Root causeStated
Approval step stalled with the SAP role owner
What we got right
What we got wrong
Time spent — hers vs ours
5.9h hers
23.1h ours
Service portalVirtual agentEnterprise search
16 linked signals
How she got here
Searched enterprise knowledge
12 attempts, then asked
2.3 hrs
Opened the virtual agent
one session, escalated
0.1 hrs
Raised it in the portal
three records raised
3.5 hrs
Closed out
Resolved, no recurrence
21 days
Coral marks a dead end.
05 MAY — 03 AUG
90 days
Recurring
Root causeInferred
Degrading NVMe drive in LT-CHI-4471
What we got right
What we got wrong
Time spent — hers vs ours
17.5h hers
17.8h ours
Service portalVoiceVirtual agentDevice telemetryEnterprise search
27 linked signals
How she got here
Searched enterprise knowledge
11 attempts, no result
2.2 hrs
Opened the virtual agent
Abandoned at 23:41
0.8 hrs
Called the service desk
Ticket raised on the call
0.6 hrs
Routed to networking
Wrong team, three times
9 days
Reimaged, drive not replaced
Recurs six weeks later
repeats
↻ Returns to the start
Coral marks a dead end.
The Unspoken
Fourteen contacts. Eleven and a half hours she never told us about.
7 silences · recorded by our own systems · never raised with us
Unreported fault
11 MAY 26
CauseInferred
The device agent logged every one. Only a fraction reached a ticket, and none after May.
5.1 h latent
Unused workflow
11 MAR 24
CauseInferred
The entitlement was never surfaced to her — not in search, not in the portal, not by an agent. A chat mentioned it by accident.
3.5 h latent
Unfound search
15 JUN 26
CauseStated
An article on exactly this exists — KB0041188 — but it is tagged to SAP Financial Reporting, so her query never reached it.
2.2 h latent
Abandoned chat
30 JUN 26
“never mind, I'll just restart again”
CauseStated
The session closed without an escalation record, so the platform counted it as contained. Nobody followed up.
0.8 h latent
Unused knowledge
04 NOV 25
CauseInferred
Tagged to SAP Financial Reporting when it belongs to SAP Approvals. The content was right; the shelf was wrong.
counted above
Unused workflow
20 JAN 25
CauseInferred
Nothing tells an employee what they are already allowed to run. The catalog waits to be searched.
counted above
After hours
08 JUN 26
CauseStated
None of this time is in a timesheet. She fixed her own tooling in the evening so the day could be spent on the close.
counted above
“never mind, I'll just restart again”
Abandoned chat · 30 JUN 26 23:41 · the last thing she said to us that night
What We'd Do Next
Four moves, each one answering something the biography already proved. Together they return 18.8 hrs of her quarter and close the cause behind 9 of 14 contacts.
01Proposed
Replace LT-CHI-4471 now, don't wait for the refresh cycle
Disk health on LT-CHI-4471 fell to 4% while the same fault produced 61 crashes and 9 support contacts. It was reimaged, not replaced, and it recurred. The chapter is still open.
61 crashes on LT-CHI-4471
disk health 4%
9 contacts, still recurring
Owner — Endpoint Ops
9.1 hrs saved
02Proposed
Re-tag the SAP approver knowledge article to the correct service
She searched for the same answer 11 times and got nothing back. The article that answers it exists — it is filed against a service her queries never touch, so enterprise search had no reason to return it.
11 zero-result searches
KB0041188 tagged to the wrong service
19 days without the role
Owner — Knowledge
2.2 hrs saved
03Proposed
Enable her for the contractor-onboarding workflow
The self-serve workflow was already hers to run. Nobody told her, so she opened a ticket and waited. The catalogue held the answer and the portal never surfaced it to her.
entitled, never triggered
raised as a ticket instead
9 contacts after 19:00
Owner — Service Owner
3.5 hrs saved
04Proposed
Add a routing rule: disk and crash symptoms never route to Networking
Three storage-fault tickets were routed to Networking on creation and reassigned out of it every time. The symptom words were in each description; the rule that reads them is not there yet.
3 mis-routes to Networking
storage fault, network queue
14 contacts in the quarter
Owner — Triage
4.0 hrs saved
Action log
0 entries · append-only
No action taken yet. Every decision on this page is recorded here.