“Have you tried turning it off and on again?” It is the most famous line in IT support, and it works often enough to be worth trying. The trouble starts when it becomes the whole answer.
A reboot clears the symptom. The printer prints, the program opens, the connection comes back. But whatever caused the problem is usually still there, and it will be back next week, or next month, at a worse moment. Real troubleshooting means finding out why it happened and fixing that.
Symptoms and causes
A symptom is what you see: the shared drive is slow, email won’t send, a laptop freezes every afternoon. A cause is the reason behind it: a disk that is quietly failing, a misconfigured setting, a software update that clashes with an old driver, a network cable that was crushed under a desk.
The same symptom can have many causes, and the same cause can produce several symptoms that look unrelated. That is why guessing is expensive. Replace the wrong part, or reset the wrong setting, and you have spent time and money without touching the problem.
Why the quick fix is tempting
Finding a root cause takes longer than a reboot. For a support desk measured on how fast tickets are closed, or a provider on a lump-sum contract where every extra hour is unpaid, the quick fix is the rational choice. The ticket closes, the user is happy for now, and the numbers look good.
That isn’t a judgment on the people doing the work. Most support technicians would rather fix things properly. It is a question of what the arrangement rewards. If an arrangement rewards closing tickets, you get closed tickets. If it rewards systems that stop breaking, you get those instead.
How an engineer troubleshoots
Our method has three steps, in a fixed order: define the issue, then define the fix, and only then implement the fix. Done that way, we know what we fixed, why the fix worked, and why it was broken in the first place.
1. Define the issue, layer by layer
Engineers describe networks using the OSI model, a standard way of splitting network communication into seven layers, each depending on the one below it. In plain terms, working from the bottom up:
- Physical: the cables, connectors, radio signals and power. Is the wire plugged in and undamaged? Is the Wi-Fi signal strong enough?
- Data link: how devices on the same local network find and talk to each other, through switches and wireless access points.
- Network: addresses and routes, which is how traffic finds its way between networks and out to the Internet.
- Transport: the rules that make sure data arrives complete and in order, and reaches the right program on the other end.
- Session and presentation: keeping a conversation between two systems open, and the formats and encryption the data travels in.
- Application: the program you actually use, such as email, a browser or your accounting software.
Working through the layers keeps troubleshooting honest. If the cable is damaged, no amount of reinstalling the application will help. If every layer below the application checks out, the problem is very likely in the program itself. Each layer you rule out narrows the search. Along the way we gather evidence before changing anything (logs, monitoring history, the error the user saw) and ask what changed recently, because many problems follow an update, a new device or a moved desk.
2. Define the fix
Once the cause is pinned down, decide exactly what will fix it, and how we will know it worked. That is the point to test the idea: “if it’s the switch, every computer on that switch will be affected and the others won’t.”
3. Implement the fix, then confirm
Make one change at a time, confirm the problem is actually gone and not just quiet, and write down what happened. The next occurrence then takes minutes instead of hours.
None of this is exotic. It is discipline, and it takes experience to know where to look first.
Recurring tickets are a warning sign
If the same problem keeps coming back, somebody is treating symptoms. Look at your support history, or ask your provider to show it to you, and watch for patterns:
- The same user or computer appearing again and again with “slow” or “freezing”.
- A printer, scanner or application that “just needs a restart” every week.
- Wi-Fi dropouts in the same part of the building.
- Password lockouts for the same people, which can point to an old device still trying an old password.
- Problems that appear after every round of updates.
Each one is a clue. Patterns across tickets often reveal a single underlying cause that no individual ticket would show. That is one reason every request should be tracked from start to finish, rather than handled by phone and forgotten.
Questions to ask your IT provider
- When you close a ticket, do you record the cause, or just that it was fixed?
- Do you review ticket history for repeat problems? How often?
- Can you show me the top recurring issues on our account and what you’re doing about them?
- When something is fixed with a restart, who follows up to find out why it was needed?
A provider who welcomes these questions is probably already asking them. We wrote more about this mindset in proactive IT management.
How FTI does it
FTI is run by engineers, and root-cause work is how engineers are trained to think. Every request to our 24/7 help desk becomes a tracked ticket, with senior engineers behind the first line. Our managed IT runs on Kaseya’s integrated platform, so tickets, monitoring and documentation sit together, and an AI layer reads across them to help our engineers spot patterns sooner. People decide what the cause is and what to do about it. The aim is simple: fewer tickets next month than this month.

