Improving MSP performance: sell the cake, not the recipe

Ask what improving MSP performance really means and you usually get the answer "KPIs" or "a good CEO". Both are true, and neither says anything. What we saw during the last edition of MSP Late Night in Utrecht is that the real dividing line between MSPs that grow and MSPs that stand still sits somewhere else: in whether you are still selling something the client can already look up for themselves.

What you see at nearly every MSP: winning new clients is the number one concern in the MSP report 2026, and at the same time almost everybody has settled into squeezing a bit more out of the existing book. That works as long as the market grows. Right now that market no longer grows the way it did ten years ago. Back then there were still mid-sized companies without an IT partner. Now you have to take clients away from a fellow MSP, and that colleague knows it and defends their position.

The underlying pain sits deeper. You are selling something the other party does not know whether they need. An owner of a mid-sized business wants to know their wifi works and that they will not be hacked. Whether you use Kaseya or Datto for that, or SentinelOne or CrowdStrike, does not interest them. And if you put it in your proposal anyway, you hand them the exact shopping list they can take to three competitors with the same question.

Meanwhile there is a second problem inside the business. Around half of MSPs say they have strategic conversations with clients, but far fewer charge anything for them. They do the work of a CISO and invoice like a managed service desk. That is no longer a positioning issue, that is money leaking away every quarter.

Sell the outcome, not the vendors

What genuinely works for MSP performance is a shift that sounds simple and rarely happens: sell the end result, not the tools. During the conversation someone made the comparison with a chocolate cake. You either have a list of ingredients, or you have the cake. A mid-sized business owner buys the cake. Yet most MSPs put their RMM, their EDR partner and their backup supplier in the proposal, after which the client goes and checks whether CrowdStrike is not better than what you offer.

The second shift sits in the package structure. Bronze, silver, gold is still standard. Basic, plus, premium too. But how do you sell yourself as an expert when the same proposal says that basic security is not really enough? An MSP owner in Utrecht put it bluntly: one package. Otherwise you are not standing behind your own advice. At first that costs you the clients who want to haggle. Over time it brings you clients who take you seriously.

The third shift is the split between hunting and farming. The MSP report 2026 shows that winning new clients is the number one concern in the market. At the same time ninety five percent of revenue leans on existing clients, because that is the path of least resistance. Anyone who actually wants sales to perform appoints one person for new business and another for existing. Otherwise every new piece of legislation becomes a reason to call through your own client book, and that gets called growth.

The fourth shift sits in positioning. Every MSP has access to the same technology. What sets you apart is a vertical you are known in, a page on your website that AI search engines actually find, and a story the business owner recognises themselves in. Not seven verticals. Two or three, and then really living them through sponsorship, cases and network.

What really decides performance at an MSP

There are two hard signals that decide whether an MSP truly performs, and neither of them appears on a standard dashboard. The first is whether the founder still runs every sales conversation personally. That feels like success for a long time, because revenue keeps coming in. But the moment the founder falls ill, goes on holiday, or spends a week on an acquisition, the pipeline stops. That is not growth, that is a bottleneck in a good suit.

The second signal is the split between hunters and farmers. In the conversation we heard from an MSP owner who has deliberately divided it up: his partners handle new clients, he handles existing ones. Why that works: if you give the same person both targets, they will always pick the existing client. That is human. You have a relationship, a phone number, and the cyber security act as a reason to call. New clients take three times the effort and only pay off six months later.

The AI service nobody is selling yet

The sharpest example from the conversation was about an MSP that built an app for a client in the cold storage industry. That client had to measure and print cooling panels, three hours per room, entirely by hand. The MSP vibe coded an application around it in a few days. The process now takes ten minutes. The saving at the client: twenty to thirty thousand euros a month. What the MSP asked in return: two thousand euros a month, on a twenty four month contract. That is forty eight thousand euros of MRR out of one client problem you already knew about.

Most MSPs know their clients' infrastructure better than the clients do. They know where the double data entry sits, which reports are built by hand, which exports a controller runs every Monday morning. That is exactly the place where AI as a service is not about handing out ChatGPT licences, but about automating specific processes away. And the pricing is fair: if you save twenty thousand, two thousand a month is not a discussion.

The trap in this part of the market sits in security. Vibe coding without a programming background produces applications that work on day one and leak on day ninety. MSPs that pick this up without senior engineers with a software background are building technical debt that pays itself back in incidents a few years later. That is why not every MSP can simply step into this, and exactly why the MSPs that can will make the difference on MRR growth over the next two years.

Anyone who still wants to grow as an MSP in 2026 has to stop selling what the client thinks they already know. Sell the chocolate cake, not the list of ingredients. Split your hunters and farmers, come up with an AI service that automates a concrete cost line at your client, and dare to bring your packages back to one. That is not philosophy, that is homework for the coming quarter. And it is exactly the kind of work MSP owners talk through with each other in our performance club, because your fellow CEOs are further along with it than you think.