05/18/2026
A managed services engagement that started in late 2020. Here is what 5 years of running it has actually changed for the client, and what it has changed about how we staff this work.
Managed services is a category where the marketing language and the operational reality are often very different things.
We are not going to share client identifying details in this post. What we will share is what 18 months of running a managed observability and cloud operations engagement has taught us about what good looks like in this category, versus what most firms describe.
What changed for the client:
- Mean time to detect on production incidents reduced significantly across the period measured monthly, not claimed annually
- Alert volume reduced by over 70% in the first six months. Maintained, not regressed, in the following twelve.
- Their internal team has grown into capabilities they did not have at the start. The engagement made them stronger, not more dependent.
What we changed in how we staff this work:
- Quarterly architecture reviews where we are actively recommending things that reduce the client's spend with us. Counterintuitive. Effective.
- Clear runbooks for everything. The day we leave (whenever that is), the client's team can run the operation independently.
Why this matters:
The default model in managed services is to maximize contract length and complexity. Ours is to make the client more capable over time, even if that means the engagement evolves or eventually ends. We have lost some renewal cycles to this approach. We have built deeper, longer relationships with the clients who valued it.
If your organization is evaluating managed services partners or evaluating whether your current one is actually making your team stronger, that is a conversation worth having.
What is the one thing your current managed services partner does that you wish more partners did? Comment below.
