Resolve: Plan, Implement, Verify, Document
Where you are: Module 3 — The Universal Seven-Step Methodology · Lesson 3.3 · Based on Chapter 3 (§3.4–3.7 and the worked example "Email not sending") of Dr. Tahseen Al Doori's Troubleshooting Essentials.
What you'll be able to do
- Write a plan of action that states the change, the risks, the rollback, the impact and any approval needed (Step 4).
- Explain the difference between a backup, recorded settings and a rollback plan, and why all three come before the change.
- Implement one change at a time, rolling back or escalating when necessary (Step 5).
- Verify a fix with the user and document it so that it becomes reusable knowledge (Steps 6 and 7).
Watch: Resolve: Plan, Implement, Verify, Document
The video takes you from a confirmed cause to a closed ticket. It explains each of Steps 4 to 7 and then walks through the book's own worked example, an email client that will not send. Keep these questions in mind:
- Why does the plan in the email example still mention backups, even though none is needed?
- What two things must be checked in Step 6 besides the original fault?
- Which part of the email record prevents the problem from happening to other users?
Why Steps 4 to 7 matter
A confirmed cause is not yet a fixed problem. Between knowing the cause and closing the ticket, a hurried technician can lose data, break something else, or make a change that cannot be undone. Steps 4 to 7 exist to stop that. They are also the steps where the Module 2 principles "protect what matters", "change one thing at a time" and "write it down" are applied most directly.
Step 4: Establish a plan of action
Before anything is touched, the book asks you to decide:
- exactly what will be changed — not roughly, exactly;
- what could go wrong;
- how the change will be undone;
- the impact on users and other systems;
- whether organisational policies allow it;
- whether approval is needed, because the change affects others or costs money;
- that data is backed up and current settings are recorded.
For a small change the plan may be two lines in your notes. For a change to a shared system it may be a written document approved by a manager. The questions are the same either way.
Backups and rollback planning
Three safeguards give you a way back. They are different things, and a good plan names all three (or states why one is not needed).
Backup
A copy of the user's data, made before the change and stored somewhere other than the system you are changing. A backup you have not checked can be opened should be treated as missing.
Recorded settings
Notes, screenshots or exported configuration showing exactly how the system is set up now, so it can be restored precisely rather than from memory.
Rollback plan
The written steps that return the system to its recorded state if the fix fails or makes things worse. A rollback is undoing a change to return to the previous working state.
Why prepare the rollback before the fix? Because the moment you need it is the worst moment to invent it: the system has just got worse, the user is waiting, and stress causes skipped steps. A rollback written in advance turns a crisis into a routine action.
Safety and authority. If the fix involves opening equipment, electrical work, failing storage, other users' accounts or systems you do not administer, it may be beyond your authority or skill — escalate rather than proceed. Never ask a user to tell you their password; use the approved reset process. Practise risky changes only on a training machine or virtual machine, never on your primary device.
Step 5: Implement the solution or escalate
Apply the plan carefully, one change at a time, following vendor instructions where they exist. Test after each change, so you always know what each change did.
- If the fix makes things worse, roll back to the recorded settings. That is what the plan is for.
- If the solution requires skills or permissions not available to you, escalate with full notes, so the next person continues from where you stopped instead of starting again.
The output of Step 5 is either the change itself or an escalation. Both are legitimate outcomes.
Step 6: Verify full system functionality
Verification confirms two things: that the original problem is resolved and that nothing else has been broken. The book lists four actions:
- Repeat the action that originally failed. If printing an invoice failed, print an invoice.
- Check related functions. A change to an email password, for example, might affect calendar or file synchronisation as well.
- Ask the user to confirm the system works for them. A problem is not closed until the user has confirmed the fix — they know their work better than you do.
- Implement preventive measures where appropriate, so the problem does not return.
Step 7: Document findings, actions and outcomes
Record the symptom, the cause, the actions taken and the result. Documentation completes the cycle in three ways: it supports future troubleshooting, it reveals recurring problems that no single ticket would show, and it turns individual effort into shared knowledge.
A practical way to record an incident is the book's problem log template (Appendix B; you will use it in this module's lab and find it in the Course Toolkit). Its fields include the problem statement, exact error message, recent changes, theories tested and results, root cause, solution applied, who verified it and when, and preventive measures.
Worked example: email not sending (from the book)
| Step | What happened |
|---|---|
| Identify | From 09:00 today the user's email client reports "Sending failed". Incoming mail still arrives, other staff can send without trouble, and the user changed their password the day before. |
| Theory | The most likely explanation is that the client is still using the old password for the outgoing mail server. |
| Test | A message sent through webmail with the new password goes through, so the account is fine and the fault lies in the client. |
| Plan | Change the password saved in the client; no data is at risk, so no backup is required. |
| Implement | The stored credentials are updated. |
| Verify | A test message goes out and arrives, and the user confirms everything is back to normal. |
| Document | The cause and fix are recorded on the ticket, and the password-change guide gains a reminder to update saved passwords in email clients. |
Three details are worth noticing:
- The plan considers backups explicitly, even when the answer is "not needed". Deciding is part of the plan; forgetting is not.
- The test isolates the fault to the client without changing anything: the webmail result proves the account and password are fine.
- The documentation includes prevention. Updating the password-change guide means the next user who changes their password is less likely to raise the same ticket.
Try it
(Fictional scenario.) A user's laptop began showing the wrong time after a trip abroad, and secure websites now display certificate warnings. You have confirmed the cause: the laptop's time zone was changed manually during the trip and never set back.
On paper, write Steps 4 to 7 for this fault:
- A plan stating the exact change, what could go wrong, the rollback, the user impact and whether a backup is needed (and why).
- The single change you will implement.
- At least three verification checks, including the original failing action and the user's confirmation.
- A problem log entry with symptom, cause, action, result, verified-by and one preventive measure.
Completion criteria: the plan names a rollback (for example the recorded original setting); only one change is made; verification includes reopening a secure website and the user's confirmation; the record contains all four of symptom, cause, actions and result.
Knowledge check
Which item does NOT belong in a Step 4 plan of action?
A fix makes the problem worse. What should you do?
Which pair of checks does Step 6 require besides fixing the original fault?
In the email example, which action is a preventive measure?
Answers explained
- C — the method changes one thing at a time; a second simultaneous fix loses cause and effect. A, B and D are all required parts of the plan.
- A — Step 5 says roll back to the recorded settings. B multiplies unknown changes, C hides the problem, and D breaks the rule never to ask for passwords.
- B — the book asks you to check related functions and get the user's confirmation. The other pairs are not part of Step 6 as written.
- D — the guide note helps future users avoid the fault. A is verification, B is the implementation, and C was the Step 3 test.
Key takeaways
- Step 4 decides exactly what changes, what could go wrong, how to undo it, the impact, policy and approval — before anything is touched.
- A backup, recorded settings and a rollback plan are three different safeguards; plan all three, or state why one is not needed.
- Step 5 applies one change at a time; roll back if things get worse, escalate with full notes if the fix is beyond you.
- Step 6 repeats the failing action, checks related functions and ends with the user's confirmation.
- Step 7 records symptom, cause, actions and result, turning one fix into shared knowledge.
Sources
- Dr. Tahseen Al Doori — Troubleshooting Essentials, University Edition (2026), §3.4–3.7, worked example "Email not sending", Appendix B (Problem Log Template); cross-references to §2.2 and §10.4 (security reminder).
