Home › Topics › Monitoring Remote & Home-Office Workers › Remote Support Playbook

A Remote-Worker Support Playbook: Fixing Problems You Can't Touch


In the office, the hardest part of a "my network is slow" ticket is the walk to the desk. Remotely, there is no desk. You cannot see the router blinking in a cupboard, swap the cable, or ask the person at the next desk whether they have the same problem. Everything you would learn in thirty seconds of looking has to come through a phone call. The person is frustrated, not technical, and worried that you are about to blame them.


A playbook fixes that by deciding what to ask, what to collect, what to try, and when to stop. This one is for a small help desk supporting people you will never visit. It includes six intake questions that scope a problem in two minutes and a check script users run and paste back. It covers the data you can pull without touching the laptop and the common home-network fixes in words a user can act on. It gives rules for escalating to the ISP, your own team, or a shipped box. It also includes a way to remember each home setup and a triage table a new technician can follow on day one.


In this guide

  • Intake: six questions that scope the problem fast
  • The self-service check users run and paste back
  • Remote data collection: what you pull yourself
  • The common home-network fixes, in plain language
  • Escalation: theirs, yours, or a shipped box
  • Documenting home setups for repeat visits
  • The triage flowchart, as a table
  • Keeping the playbook current
  • Where to go next

Intake: six questions that scope the problem fast


The first two minutes of a remote ticket decide the next hour. Ask these six questions in this order, every time, before touching anything. Each eliminates a whole branch of the investigation, and together they usually point at a segment of the path before any measurement is taken.


  • 1. What exactly is slow or broken? "The VPN" is a road, not a symptom. Get the application and the behavior: the file share takes a minute to open, the video call freezes for a few seconds at a time, the CRM login spins. Each lives on a different road and fails differently.
  • 2. Since when, and what changed? A problem that began yesterday afternoon coincides with something: a new router, a moved desk, a software update, a new neighbor, a roommate who started working from home. Continuous since the person started remote work is a different and usually simpler problem.
  • 3. Everything, or one application? One application points at that application's road and its far end. Everything points at the shared segments: the laptop, the Wi-Fi, the router, the ISP.
  • 4. Wired or Wi-Fi? If wired, skip every wireless fix. If Wi-Fi, ask where the router is relative to the desk and whether a cable is possible at all.
  • 5. Where are you right now? Home, a hotel, a coffee shop, a relative's house, a phone hotspot. The environment sets the expectations and the available fixes; a hotel network needs a workaround, not a router restart.
  • 6. Have you tried the hotspot test? Tethering the laptop to a phone and repeating the task. If it fixes the problem, the house or its ISP is the culprit; if not, the far end is. Many users have done it already, and the answer is worth more than most measurements.

Write the six answers at the top of the ticket in one line each. A technician picking up the ticket later should see the shape of the problem without re-interviewing the user.


The self-service check users run and paste back


Users cannot read a routing table, but they can double-click a shortcut and paste what appears into a ticket. Give them a small script that prints the numbers you most need. These are gateway round trip and loss, round trip and loss to your VPN edge, whether an internal name resolves, and the wireless state. The illustrative version below is for Windows; the shortcut that launches it belongs on every remote laptop, named something friendly.


# quick-check.ps1 — run me, then paste everything below into your ticket
"Time: $(Get-Date -Format 'MM-dd HH:mm:ss')   Computer: $env:COMPUTERNAME"
$gw = (Get-NetRoute -DestinationPrefix "0.0.0.0/0" |
       Sort-Object RouteMetric | Select-Object -First 1).NextHop
"Gateway ($gw):"
ping -n 10 $gw | Select-String "Lost|Average"
"VPN edge:"
ping -n 10 vpn.example.com | Select-String "Lost|Average"
"Internal name:"
Resolve-DnsName intranet.example.com -ErrorAction SilentlyContinue |
    Select-Object Name, IPAddress | Format-Table -HideTableHeaders
"Wireless:"
netsh wlan show interfaces | Select-String "SSID|Signal|Channel|Receive rate"
$isp = try { Invoke-RestMethod "https://echo.example.com/isp" -TimeoutSec 5 } catch { "unknown" }
"Public network: $isp"

A pasted result, which takes ten seconds to read:


Time: 03-14 10:14:02   Computer: LT-0417
Gateway (192.168.0.1):
    Packets: Sent = 10, Received = 10, Lost = 0 (0% loss),
    Minimum = 1ms, Maximum = 4ms, Average = 2ms
VPN edge:
    Packets: Sent = 10, Received = 7, Lost = 3 (30% loss),
    Minimum = 35ms, Maximum = 410ms, Average = 122ms
Internal name:
intranet.example.com  10.0.4.20
Wireless:
    SSID : home-net    Signal : 91%    Channel : 44    Receive rate (Mbps) : 866
Public network: cable-provider

The gateway is perfect: the house is not the problem. The VPN edge is losing three packets in ten with round trips swinging from thirty-five milliseconds to over four hundred. So the trouble is between the house and your door, or at the door. The public network is the user's usual provider, so this is not a hotspot day. Next, check whether other remote users see the same thing now. Look at the concentrator and the office circuit. If both are quiet, this is the ISP or the path, and the user needs evidence and a workaround. The full decision table for these numbers is in sorting home Wi-Fi from corporate blame.


Remote data collection: what you pull yourself


While the user runs the script, collect what you can from your side. None of it touches the laptop, and together it turns the user's one snapshot into a timeline.


  • The endpoint check rows for the last day. If the laptop runs the lightweight endpoint check, you already have gateway, tunnel, DNS, and signal numbers every few minutes. Look for when the numbers changed, not just what they are now.
  • The concentrator's session log for that user. Connect and disconnect times, the source address for each session, and any reconnect storm. A tunnel that dropped six times this morning explains "things keep stalling" on its own. And a source address that changed at nine tells you the user moved to a hotspot.
  • The office side at the same minutes. Concentrator sessions and CPU, office circuit utilization, and any alert that fired in the window. If the circuit was pegged, stop investigating the user.
  • Other users' rows for the same window. Three people on the same ISP with the same pattern is a provider problem; thirty people on every ISP is yours.
  • The VPN client's log, with permission. Ask the user to send it or read the last few lines aloud. It names reasons for disconnects that the concentrator only records as "peer went away."

Tip: put the endpoint timeline, the session log, and the office graphs for the same window on one screen before you call the user back. Most remote tickets are solved by noticing that two lines changed at the same minute. And that is hard to see across three windows and two tools.


The common home-network fixes, in plain language


When the evidence says the house, the fix is something the user has to do with their own hands. None of them assumes an ethernet port at the desk, because there is no desk. The playbook's job is to describe each one in words that work over the phone, with the reason attached, because people follow instructions they understand.


Band selection

Home Wi-Fi uses two or more frequency bands. The lower band reaches farther but is crowded and slow; the upper bands are faster but fade through walls. Some routers show them as two separate network names; many combine them into one and choose for you, not always well. Tell the user: "If you see two network names, use the one with 5G or 5GHz in its name when you are near the router. If you see one, moving closer usually convinces the router to switch you to the faster band."


Channel congestion

Neighbors' networks on the same channel share the airtime, which shows up as a fine signal with terrible loss. Most home routers pick a channel automatically at startup, so a restart is also a chance to land on a quieter one. Tell the user: "Your signal is strong but the air is busy. Restart the router; it will pick a new channel when it comes back, and we will check the numbers again."


Router placement

Routers end up where the cable enters the house: a cupboard, a basement, behind the television. Wi-Fi wants height, open space, and a central spot. Tell the user: "Out of the cabinet, up on a shelf, away from the TV and the microwave, as central as the cable allows. Even a meter can matter."


Mesh systems and their backhaul

A mesh kit adds satellite units that relay traffic to the main router. When the relay itself is wireless, each hop roughly halves the available speed. Under those conditions, a satellite placed too far from the main unit relays a bad connection beautifully. Tell the user: "Put the satellite about halfway between the router and your desk, not at your desk. If there is any way to run a cable between the two units, do it; a wired link between them fixes most mesh problems."


DNS settings

Leave the router's DNS settings at their defaults unless there is a reason not to. The VPN client sets the resolver it needs when the tunnel comes up. "Helpful" changes on the router can override it and break internal names or slow every lookup. Tell the user: "If anyone changed the DNS settings on the router for speed or filtering, set them back to automatic and we will test again."


Extenders and powerline adapters

A cheap wireless extender rebroadcasts the signal it receives. That halves the speed and often creates a second network name the laptop clings to long after the user has walked back to the router. Powerline adapters send data over the house wiring and work well on one electrical circuit and badly across two. Tell the user: "Extenders and powerline plugs are often the cause rather than the cure. Try turning the extender off and connecting to the main router directly. If that is better, we will talk about a proper second access point or a mesh kit."


The ISP modem reboot

When the router is separate from the modem, the modem holds its own state and its own connection to the provider. In that setup, a stuck modem produces a perfect gateway with a dead internet. Tell the user: "Unplug both boxes. Count to thirty. Plug in the modem first and wait until its lights stop blinking, then plug in the router. Order matters."


Two more that save calls: ask the user to pause large uploads and cloud backups during meetings. Also ask them to check whether the router has a "gaming" or "prioritization" mode switched on that puts the work laptop last in line.


Escalation: theirs, yours, or a shipped box


A playbook needs a stopping rule, or every ticket becomes a project. A remote ticket can go to three places, and the evidence decides which.


At Northgate Retail one remote buyer's ticket stayed open eleven weeks. It gathered forty comments, a shipped adapter, a shipped router, and a suggestion that she move her desk. She did. Her gateway had been clean from the first paste-back; the loss began two hops past her ISP. A stopping rule would have sent it to the provider in week one. The review called the ticket "thorough."


To the ISP, through the user

Loop in the provider when the gateway is clean or the loss or delay begins beyond the house in a path test. Also loop them in when the pattern recurs at fixed hours or the modem keeps losing its connection. You cannot call the provider; the user can, and they will do better with a script. Give them the time window and the symptom in one sentence. Include the fact that a wired test showed the same problem (providers ask), and the path test output. Say plainly that this is the provider's to fix, and offer the hotspot as a workaround while they wait.


To your own network team

Escalate internally when several users on different providers show the same tunnel symptoms in the same window. Also escalate when the concentrator or the office circuit was busy at the time. Do the same when the tunnel is clean and an internal application is slow. This is your incident now, and the courtesy is to tell remote users quickly so they stop rebooting routers. The guide on monitoring both remote paths covers the concentrator-side checks that should have raised the alarm first.


To a shipped box

Suppose the verdict is the home network, the plain-language fixes have been tried, and the numbers are still bad. In that case, ship hardware rather than continue the conversation. Start with a wired adapter. For roles that depend on calls, ship a company-owned router or mesh kit. Decide the criteria once and write them in the playbook, so the help desk can apply them without asking a manager.


Documenting home setups for repeat visits


Remote workers call back. The same house has the same router in the same cupboard. And a technician who knows that starts the next ticket where the last one ended. Keep a short home profile for each remote worker alongside the laptop's asset record, in network terms only. Nothing about the family, the address, or the devices in the house beyond what the fix needs.


Asset: LT-0417                  ISP: cable, provider-supplied modem/router combo
Desk vs router: one floor up     Wired run: not possible; adapter shipped, unused
Wi-Fi: single network name       Company kit: mesh satellite at top of stairs
Known: microwave beside router; evening slowdowns are the provider's
Last verdict: home network. Satellite moved; gateway 3 ms, no loss since.
Promised: check-in after next ISP call

Update it at the end of every ticket, not the start of the next. The "last verdict" and "promised" lines save the most time. They stop the technician re-running fixes that already failed and forgetting a follow-up promised in good faith.


The triage flowchart, as a table


Flowcharts do not survive in ticketing systems; tables do. Start at step one and follow the answers; each row names the evidence that decides it and where to go next.


StepQuestionEvidenceIf yesIf no
1Is the laptop itself healthy?CPU, memory, disk, power mode, uptime from the endpoint rowGo to step 2Desktop support: reboot, free space, close the runaway process, then retest
2Is the gateway clean?A few ms, no loss, in the paste-back or the endpoint rowGo to step 3Home-network fixes; hotspot test to confirm; hardware if fixes fail
3Is the user somewhere unusual?Public network shows mobile, hotel, or a different providerSet expectations; workaround; note in ticketGo to step 4
4Is the tunnel clean?Tunnel round trip minus gateway is steady and region-appropriate, no lossGo to step 6Go to step 5
5Are others affected, or your edge busy?Other users' rows; concentrator sessions and CPU; office circuitYour incident: escalate internally, tell remote usersISP or path: evidence to the user; hotspot workaround
6Is an internal application slow?Application response from the endpoint vs tunnel round tripServer team: the network is exoneratedGo to step 7
7Is a cloud application slow from the endpoint?SaaS reachability from the laptopGo to step 8Nothing reproducible now: leave checks running, ask for the next occurrence
8Is it slow from the office probe too?The same check from your networkThe vendor: status communication; us-or-them playbookThe user's direct path: ISP evidence; consider routing that app through the tunnel

The order is deliberate. The laptop comes first because a sick machine fails every later step for the wrong reason. The gateway comes second because it is the cheapest number to get and the most common answer. The path questions come last because they need your data as well as the user's.


Tip: add a ninth row for your own environment's most frequent special case. That might be the branch office that everyone in one region tunnels through, or the application that only works in full tunnel. It might be the ISP whose evening congestion you have learned to recognize. A playbook that knows your specific network is the one the help desk will actually keep open.


Keeping the playbook current


A playbook decays quietly. The VPN policy changes and half the application map is wrong. A new mesh kit becomes the standard and the extender advice is out of date. The ISP that used to be reliable is not. Build a small maintenance habit into it.


  • Step 1: Name an owner. One person is responsible for the playbook, and their name is on it. Shared ownership is no ownership.
  • Step 2: Update on triggers, not on a calendar. Any change to the VPN routing policy, the standard hardware kit, or a critical application's hosting updates the playbook the same day.
  • Step 3: Read the verdicts periodically. Skim the closed remote tickets. Fixes that never worked come out; a fix that solved a dozen tickets moves up; a provider that keeps appearing gets its own row.
  • Step 4: Retire what nobody uses. A step nobody follows is noise that hides the steps that matter.
  • Step 5: Test it on a new technician. Hand the playbook to someone who has never supported a remote worker and listen to where they get stuck. That is the next edit.

Keep a change note at the bottom: what changed and why. When a user says "last time you told me something different," the note is how you find out which advice was right.


Where to go next


The process has six questions, one paste-back script, and a short list of what to pull from your own systems. It also has fixes in the user's language, three escalation doors, a home profile, and a table to follow. That is a complete support process for people you cannot walk over to. It is not the walk to the desk, but it is close. The visibility model explains the order of the questions. The guide Lightweight Endpoint Checks for Remote Laptops supplies the rows the triage table reads. The guide Home Wi-Fi vs. Corporate Blame holds the decision table for step two. All of it is collected at Monitoring Remote & Home-Office Workers.


Put this into practice with SysPeep Monitor

Step five of the triage table asks whether your edge was busy at the time. That answer should already be on a graph. SysPeep Monitor keeps it there. It discovers your VPN concentrator, firewall, and edge router. It polls sessions, CPU, and circuit utilization over SNMP, and alerts before the tickets arrive. It does this from one Windows machine with a single web console and lightweight remote probes for other sites. The playbook is yours to write; the free trial covers the half you can measure.

Download Free TrialLearn About SysPeep