There is reduction in the need of these MSPs for tactical, transactional, low value work, but there is still high demand for high value, industry specific, outcome-oriented engagements.
“The issue is MSPs are so consumed and locked in just by keeping the lights on or doing this tactical, transactional work, they get busy,” Thyagaraja said. “Clients are also busy, and the needle hasn’t been pulled.
“What will happen is the MSPs will continue to work on these low value gigs forever, and they’re also not able to extract synergy from their clients, and that’s not translating to additional spend.”
According to Thyagaraja, one of the first things MSP should be doing in this regard is lead with transparency.
“MSPs who hide AI gains will lose trust faster than they protect margin,” he said. “It’s very important to proactively disclose where AI is being used.
“It’s very important to quantify what the productivity impact is going to look like, because this will only breed more trust.
“The client will also be a little more appreciative of the amazing ways in which you are using AI in [their] existing contract.”
Thyagaraja believes this can lead to a situation where the client is impressed with the AI capabilities, and that could lead on to more strategic work. Ultimately, it could potentially lead to cross upselling of other services for the MSP, additional spend, earlier renewal, or better incentives.
“Transparency is becoming a commercial differentiator today,” he noted. “Earlier it was perceived as a risk, but today transparency is a true commercial differentiator for MSPs — that’s number one.
“Now, number two is they should really start to rebuild pricing for the AI era. The pricing models [and] the commercial constructs that today’s MSPs have got, they’re still tied to the old school ways of working that values effort.”
This could be largely time and material, or it could be effort focused, it could be fixed fee, but ultimately, it’s just glorified time and materials plus margin.
“We don’t really see outcome-based pricing or unit-based economics and so on,” Thyagaraja explained. “The important thing [is] you can’t sell machine augmented work with human only pricing logic.
“MSPs must evolve beyond pure FTE [full-time equivalent] or human-based pricing. They need to start moving towards outcome-based units, transaction-based pricing, [and] platform plus services bundles.
“They should seriously start thinking about gain share and outcome based commercial models. Those who don’t will be forced into rate erosion instead of structured value realisation.”
The latter will occur after the MSP has “lost the trust” of their customer, Thyagaraja warns.
“It’s so much better to be transparent up front about the amazing work that you’re doing in AI and how you’re able to accomplish a lot more with less effort,” he explained. “That should match with the rapid experimentation on the pricing models.
“You’ve got transparency to build the trust, and you’re now telling the clients, ‘Here is an emerging, evolving pricing construct, given the advancements that we are making.’”
Ring fence for AI economics
According to Thyagaraja, having the AI, transparency and trust gives the end-customer the understanding the MSP doesn’t want to bare minimum, keep the lights on, bare-bones style of service providing.
Instead, it becomes trust with “more strategic, high value work”, he said.
“It’s very important to ring fence AI investment economics. If MSPs don’t separate AI platform economics from the delivery economics, margins will become unexplainable.”
What that means is having very explicit AI enablement fees clearly scoped, along with AI tooling costs.
“Separate commercial lines for human delivery, AI platforms, data and model, governance, security and compliance,” advised Thyagaraja. “This is what is going to protect MSPs from uncontrolled commoditisation. Basically, ring fence by clarifying if you want haste.
“There are many clients belonging to several sensitive industries that still prefer significant human in loop delivery. As an MSP, you need to be prepared to cater to all industries and verticals and all personas and all sorts of clients.”
This will allow AI investment economics to be stated very clearly in commercials include fine print on human only services or FTE only.
“This is what the pricing is going to look like,” explained Thyagaraja. “If there’s an AI platform involved, this is the component attached to it. If it is data model governance, this is what the ongoing cost is going to look like. If you’re looking at security and compliance, if you want us to be the custodian for compliance, then that will also be coming at a premium.”
The other important aspect, which can become overlooked, is building client ready AI governance.
“This is another problem in the marketplace at the moment,” Thyagaraja said. “A lot of MSPs are simply striking deals with tool chain vendors, and they’re beefing up the tooling narrative to their clients.
“But client ready AI governance is something that is missing, because AI without governance is a regulatory liability disguised as innovation, and that’s a big problem.”
This means MSPs must provide AI usage logs, Thyagaraja explained. They should provide model risk controls, as well as data lineage and IP protection from governance, security certification aligned to AI usage.
“Enterprises are already assessing MSPs today on AI risk security, and this is not looking at engineering skills,” he said. “Future MSP margins will come from how well humans orchestrate AI, not from head count scale.
“They need to reskill deliver teams for AI supervision, prompt engineering, model validation, exception handling, and they need to start redesigning their pyramids.
“That could mean more AI supervisors and domain architects, or maybe fewer low value execution goals.”
This is why it’s “very important to update billing constructs to reflect this augmented productivity as well”, Thyagaraja explained.
These changes will give MSPs control of the narrative with their customers. This will prevent abrupt rate shocks at these positions.
In other words, this places MSPs as strategic partners and not defensive vendors, Thyagaraja said. “You don’t want the MSPs to be defensive, just want them to be proactive and more strategic,” he added.
“This will just enable more structured, productive pass throughs, more phased savings opportunities for joint reinvestment funds, and so forth.”
Leave a Reply