Consolidation Wave
TechCentral is out with a piece that should feel uncomfortably familiar if you run a managed services business. They covered an executive dinner hosted by Acronis where MSP leaders described a common pattern: more tools, more dashboards, more endpoint agents, and yet less confidence in what is actually happening across client environments. The reporting is blunt about the day-to-day reality. Teams are spending more time bouncing between consoles, reconciling alerts, and stitching together a picture after something goes wrong. And the story frames it as an observable operational condition: the stack itself has grown into something that creates drag, not clarity.
At the infrastructure layer, The Register reports on a survey from Virtified of 450 VMware users across 14 countries, all from organizations with more than 500 employees. The headline number is simple: half of VMware users say they plan to reduce their usage by 2028. The story attributes that intent to Broadcom’s push toward a bundled private cloud package, Cloud Foundation 9, which respondents describe as too costly or too complex for what they want. It also notes that migrations are already under way, and that some organizations are looking at their virtual machine fleets with an eye toward downsizing or shifting portions to alternatives like Nutanix or Microsoft. Again, we do not have to infer anything here. The observable fact is that a large share of the installed base is signaling an active move to reduce footprint.
And then there’s Apple. In Apple’s own newsroom, the company announced “Apple Business,” a new all-in-one platform launching April 14th, 2026, in more than 200 countries and regions. Apple says it is combining built-in device management, business email and calendar with custom domain support, and directory services into a single offering. Apple also says the platform will connect to marketing and discovery, with local advertising on Apple Maps coming later this summer in the U.S. and Canada. The on-the-record message from Apple is that this is meant to make it easier for businesses of all sizes to “run and grow” with fewer moving parts.
Coherence Gap
Stacks are under pressure, customers are declaring intent to reduce and consolidate, and major vendors are shipping platforms designed to absorb more of the workflow into one place.
What’s powering this shift is a breakdown in operational coherence—organizations can no longer reliably align policy, workflow, and execution as environments grow more complex. And when that alignment fails, organizations gravitate toward whatever reduces decision-making, enforces guardrails by default, and makes the system easier to run day to day.
Auvik’s 2026 IT Trends Report puts numbers on that internal mismatch. Their data shows that AI enthusiasm is widespread, but operationalization is rare, and the bigger tell is the disconnect between leadership intent and front-line reality. Auvik highlights that most organizations think they are ready, yet they still cite fragmented visibility, limited time, and governance gaps as the blockers that keep initiatives from becoming real operations.
Then look at how Microsoft is positioning Copilot Cowork in Microsoft 365. This is not “here’s another AI tool.” It is AI designed to live inside the suite where work already happens, carrying state across files, calendars, and tasks, with enterprise boundaries wrapped around it. That architecture is not just a product choice. It is a response to the coherence problem. If organizations struggle to stitch together workflows across disconnected tools, the platform moves to own the stitching.
The SMB market shows the same pattern in compressed form: adoption is ahead of alignment, so buyers look for someone to impose order.
So the mechanism is simple: when organizations cannot keep their own operations coherent, they choose environments where coherence is prepackaged.
Margin Leak
For MSPs, this shows up as a margin problem driven by unpriced volatility. When AI-driven environments are not tightly governed, the work doesn’t become predictable—it becomes erratic. Edge cases, inconsistent outputs, and ‘it worked yesterday’ tickets turn support into a volatility problem, where labor hours swing based on exceptions, not volume.
If your agreements do not define what is covered when AI-generated work misfires, you inherit ambiguity. And ambiguity always resolves into free labor.
Forrester’s reporting on workforce AI readiness shows why those exceptions are going to show up. They are very direct that many employers are rolling out workforce AI tools without training people to use them competently, and that foundational skills are barely improving. That is the recipe for negative productivity: staff use the tools, get inconsistent results, do not know how to validate them, and then work has to be redone. In an MSP context, those retries and cleanups land in your queue. They show up as escalations, project overruns, and “can you just take a look” requests that blow up the economics of fixed-fee agreements.
Business Insider’s look at IBM’s consulting model gives the other half of the picture: IBM is not treating AI scale as something you “deploy and forget.” They built real-time monitoring for AI agents, and they measure outcomes like compressing investigation time from 45 minutes to a few minutes. The proof point is the operating assumption: automated work only stays cheap when oversight and control are built in, continuously, as part of delivery. Without that, the cost leaks out through exception handling.
So the fork is simple. The MSP either becomes the provider that simplifies and governs the automation layer, with clear standards and managed controls that keep exceptions rare and billable, or the MSP gets trapped absorbing complexity on the back end while the client negotiates as if AI should have reduced the price.
Why Do We Care?
The question MSPs need to answer right now is not “should we adopt AI” — it is “who owns the coherence layer in our clients’ environments, and are we positioned to be that provider or to be displaced by it.”
If you believe clients are buying tools, you will keep adding complexity while your margins become more volatile. The MSP that positions as the entity that reduces decisions — through governance, through stack simplification, through defined outcomes — captures the relationship. The MSP that keeps adding tools to demonstrate capability is moving in the wrong direction.
If an MSP misunderstands this, they will keep adding tools to signal capability while their cost structure becomes more volatile and their contracts fail to capture that risk.
What to Consider
- Audit every fixed-fee agreement for AI-generated work coverage. Define explicitly what is and is not covered when AI outputs require remediation, rework, or rollback. Ambiguity resolves into free labor — close it now.
- Build a governance layer, not a governance deck. If you are selling AI governance, you need technical enforcement: policy controls, audit logging, and defined rollback procedures.
- Repackage around outcomes with defined exception economics. The MSPs who win the margin game are those who can say: “Here is what the outcome costs, here is what an exception costs, and here is how we keep exceptions rare.” That requires simplifying the stack first — automation layered on complexity accelerates cost, not efficiency.
If this trend continues, MSP contracts will split into two explicit rate cards: a lower-priced “standardized platform lane” and a premium “nonstandard/exception lane,” and the exception lane becomes the primary profit engine.