What happens when an MSP leaves: a real infrastructure rescue.
What undocumented, neglected small business infrastructure actually looks like, and what getting it back under control involves.
May 2026 — Kurth Bemis
A few years ago I took over as the in-house technology manager for a manufacturing operation in Claremont, NH after their managed service provider ended the relationship. The details of why they parted ways are not mine to share, but the situation I walked into is worth describing, because it is not unusual. I have seen versions of it with almost every new client who has come to me from a prior provider.
What I found
The infrastructure was running. That is probably the best thing I can say about it on day one. Machines were on, email was working, production was not stopped. But almost nothing was documented, and several things were running on configurations that nobody had intentionally set up. They had just accumulated over time.
A few of the specific things I found in the first two weeks:
- The domain and several vendor accounts were registered to an email address that belonged to a former employee who no longer worked there. That former employee was not reachable.
- Backups existed, but nobody had tested a restore in over a year. When I tested one, it failed. The backup job was running and completing successfully in the logs, but the backup was not usable.
- A server running a business-critical application was running on hardware that was past its end-of-life and had not received OS patches in eighteen months. It was one hardware failure away from taking a core function offline with no clear recovery path.
- Three employees who had left the company in the previous year still had active accounts on several systems.
- Nobody knew the network topology. I spent a full day mapping it. Some of what I found was not intentional: equipment that had been added and forgotten, addresses that conflicted, a wireless access point hidden above a ceiling tile that nobody knew was there.
None of this is catastrophic in isolation. Put it all together and you have a business that was one bad day away from a serious problem with no clear path to recovery.
What getting it stable looked like
The first month was almost entirely discovery and documentation. Before changing anything, I needed to understand what was there. I mapped the physical network, inventoried every device, documented vendor contacts and account credentials, and got everything into a central documentation system. That work is not glamorous and it takes time, but everything that comes after depends on having an accurate baseline.
Once the picture was clear, the fixes were largely straightforward: patch the server, replace the aging hardware on a planned timeline, fix the backup configuration and verify restores, close the former employee accounts, and transfer domain and vendor account ownership to email addresses the company actually controlled.
The monitoring infrastructure went in early. If something started trending toward failure, I wanted to know about it before it became an outage. By the end of the second month, the environment was stable, monitored, documented, and backed up properly. Not upgraded, not transformed, just sound.
Why this keeps happening
MSPs that serve small businesses are usually structured around a service delivery model that works better for larger clients. Small accounts get less experienced technicians, less frequent proactive attention, and documentation that lives in the MSP's systems rather than yours. When the relationship ends, you get whatever they decide to hand over, which is usually not complete.
The documentation problem is the most persistent. I have never taken over a small business environment from a prior provider where the documentation was accurate and current. Not once. It is not necessarily bad faith on the provider's part. Keeping documentation current is unglamorous work that does not bill well and does not come up in a sales conversation. It tends not to happen.
This is part of why I treat documentation as a deliverable, not a side effect. Everything I set up or find goes into a system on your hardware. When I am done with an engagement, you have a complete record of your environment, regardless of what happens next.
If this sounds familiar
If you are currently with a provider and have a nagging feeling that things are not quite right, or if you have recently parted ways with a provider and are not sure what you actually have, the free assessment is exactly where to start. I will come to your location, spend a few hours looking at what is there, and give you an honest read on the state of things. No commitment required.
Get in touch or call (603) 826-6070.