6.2Recurring issues (playbook)
The essentials in 30 seconds
- A symptom first: describe what is happening, to whom and since when.
- One cause at a time: test the simplest hypothesis before modifying the whole thing.
- Proof after the action: verify that the problem has disappeared and that nothing has regressed.
Nassim’s thread — the incident box
Nassim keeps a simple file: symptom, document or meeting concerned, since when, probable cause, action, test after correction. In three weeks, he noticed that half of the incidents came from incorrectly set notifications or files sent as attachments.
The playbook is not a list of recipes: it is a diagnostic order.
Why a playbook?
The same incidents keep recurring: a meeting without sound, an inaccessible file, a message sent to the wrong place, a channel that is too noisy.
A playbook avoids improvising. It imposes a precise symptom, a probable cause and a testable action. It also allows you to pass on the solution to a colleague.
| Symptom | Likely causes | Actions |
|---|---|---|
| The meeting starts without sound or camera | Wrong device selected, browser permission denied, headset disconnected | Check the audio/video settings, test the hardware, join by the link and not by an old tab. |
| A shared file cannot be found | File uploaded to a chat, personal OneDrive, or another channel | Search by name in Teams, check the channel's Files tab, ask for the link rather than an attachment. |
| Too many notifications and messages | All channels monitored, mobile alerts enabled, mentions too broad | Follow only useful channels, set notifications, use silent hours and targeted mentions. |
| An external guest cannot join | Invitation sent to wrong account, unauthorized external access, expired link | Check address, domain and link. Test with the guest account before the meeting. |
| Two versions of the same document are circulating | Multiple attachments, lack of naming rules, co-editing not used | Centralize in the channel, apply naming, share an editable link and stop local copies. |
| Actions decided in meetings are lost | No manager, no date, report too late or not found | Note actions during the meeting, post a short report, pin it and set a reminder. |
Script ready to say — qualify an incident (30 s)
“How long has the problem appeared, in which channel or meeting, and for whom? I reproduce once, check for the simplest cause, then fix and test.”
To keep to frame an incident with a colleague in a hurry.
Copy and paste template — incident sheet
SYMPTOM: _______________
DOCUMENT / MEETING: _______________
SINCE: _______________
AFFECTED USERS: _______________
REPRODUCED: yes / no
LIKELY CAUSE: _______________
ACTION: _______________
TEST AFTER FIX: yes / no
TRACE PUBLISHED: yes / noShort version
Reproduce, correct, test.
Remote
Request a capture and exact time.
48 hour challenge — apply tomorrow
Take three real or simulated incidents, fill out the form and apply the diagnostic order. Note the cause and the regression test. 30 minutes.
- I describe a symptom before taking action.
- I relate the symptom to a probable cause.
- I test and trace after correction.
Frequently asked questions
What to do if the problem is intermittent?
Note the time, network, device and channel involved. Look for a pattern before changing settings.
Should we recreate the meeting or the channel?
Not as a first instinct. Check link, permissions and account. Recreate as a last resort, warning the participants.
How to prevent the incident from recurring?
Add a simple rule: a link rather than an attachment, one channel per topic, a test before the invitation.
A colleague can't find the meeting materials. What is the first action?
Diagnostic rule: reproduce, isolate, correct, test. If the symptom is not reproducible, note the exact context before changing anything.