Why AI has yet to transform most MSPs

In my 30 years in this business, I have seen many changes. Most were incremental technology changes, but a few fundamentally reshaped the industry: the transition from hourly support to the MSP model, the changing security landscape and the impact of the cloud.

Now AI is here.

The biggest difference with AI is the speed at which it has dominated our thinking. In just over a year, AI has gone from a curiosity to a topic that is part of almost every conversation—both inside the MSP and with customers.

At the same time, very few MSPs have seen a meaningful impact on either their operational efficiency or their commercial business model.

Why?

Let’s start with the MSP’s internal business.

On the surface, an MSP appears to be a perfect target for AI and automation. We perform many repeatable tasks that rely on tools, documented processes, and a knowledge base. In addition, there are now hundreds of AI software vendors demonstrating products that promise to automate some part of the MSP.

Yet, as we benchmark efficiency across more than 1,000 MSPs each quarter, the median results still don’t show tangible efficiency gains.

The primary reason is a lack of business maturity.

Roles are not clearly defined. Processes are inconsistent or undocumented. Accountability is unclear. Adding automation to an immature business or a poorly defined process will not produce the promised results.

Let’s look at the support desk as an example.

There are many automation solutions designed for support. But if your support efficiency is low today, you probably have a process problem that needs to be solved before you can maximize the value of automation.

We can measure support efficiency by looking at metrics such as tickets closed during a period or the number of customer seats managed per support resource. In many MSPs, the biggest obstacles to efficiency are not technical. They are basic process issues involving triage, ticket assignment, escalation, and workflow management.

A common example is technicians having too many tickets open at the same time. They touch many tickets throughout the day but close very few. They spend too much time deciding what to work on, managing their queues, and shifting between unrelated problems.

Centralizing ticket assignment and workflow management can dramatically change these results. It allows technicians to spend more of their time “teching” instead of managing and prioritizing their own work.

Add a well-defined escalation process that reduces unnecessary ticket touches, and an MSP can sometimes double ticket throughput.

All of this can happen before adding automation.

Once the role and process are clearly defined, automation can have a significant impact. This is exactly what we see when comparing top-quartile, mature MSPs with the median MSP. The mature MSP gets more value from automation because it is applying technology to a well-designed operating model.

The same principle applies to every other function in the business.

Before automating something, define the role, document the process, establish the desired outcome, and determine how success will be measured. Otherwise, you may automate activity without improving the result.

Every MSP needs to view its automation journey through this lens. Automating a poorly defined role or process will not transform your business. In some cases, it will simply allow a broken process to operate faster.

But there is an even bigger takeaway.

Every MSP customer is beginning its own automation journey. Those customers face the same challenges involving roles, processes, accountability, and measurable outcomes. Buying AI tools will not solve those challenges for them.

If an MSP does not understand how to navigate this process internally, it will not be prepared to help its customers navigate it.

The MSPs that master this inside their own businesses will gain more than efficiency. They will develop the knowledge and experience needed to build the next generation of services for their customers.

AI technology is not the biggest obstacle.

Business maturity is.

Related: AI technology isn’t the issue for MSPs.