As organizations carve out budgets for AI technologies, partners are racing to capture an expanding business opportunity, often without the required engineering talent and experience. Smaller managed services providers are at risk of losing out to hyperscalers, large language model builders and other large vendors that have mobilized forward-deployed engineering teams.
Most partners have limited ability to integrate AI into enterprise workflows, said The Futurum Group Analyst Tiffani Bova in a recent report.
According to Futurum data, 71% of surveyed partners sell AI software and 62% offer AI consulting services. Yet only 48% of partners who rate themselves as AI-confident report a robust AI business generating significant revenue.
Partners claiming AI expertise often lack the data scientists, model orchestrators and agentic system know-how needed to provide AI solutions, said Bova, who has almost 30 years of channel experience.
“Think about an MSP — they might sell the hardware, implement it and deploy it,” Bova told Channel Dive. “They might be using AI to improve their business, but an AI practice is heavy. It has PhD data scientists, software developers and coders. That tends to be a skill set that the traditional resale, managed service channel solution provider does not employ.”
Small MSPs may not have the resources to hire AI engineers and data scientists for AI conversion work, Bova added.
Vendors usually aren’t helping. Rather than investing in the partner upskilling, many have banked on enterprises using internal talent and global systems integrators to get the work done, Bova said.
The money is in AI services. Operations and workflow integration is the fastest-growing piece of the AI market, projected to reach $92 billion by 2030, per a Futurum forecast published earlier this year.
A new type of partner built around AI, with proprietary methodologies and the ability to orchestrate multiple models, is now winning enterprise work. Bova calls them frontier partners, and the big GSIs are snatching them up.
In January, Accenture acquired Faculty, a UK-based AI-specialist partner. Last year, Capgemini purchased business transformation company WNS for $3.3 billion to build an agentic operations business.
“We have a new set of partners entering our ecosystem, and they behave very differently, similar to what born in the cloud firms did when the cloud showed up,” Bova said. “They didn’t want to just resell. This is a very similar transition point.”
Partners need to decide where they stand in the AI stack, Bova advised.
“It’s okay to just be in the infrastructure layer,” she said. “That’s a good business. You can sell AI-ready technology from the Ciscos and Dells and IBMs of the world. You can be at the hardware layer, the software layer, the services layer or the user layer. But you do have to pick, because each layer requires different skills, different business models, different partnerships, different certifications and different vendor relationships.”
For MSPs looking to help enterprises implement AI, Bova recommends recruiting employees with model-orchestration and agentic-delivery skills, and building models that allow for IP ownership.
She also advises replacing the traditional certify-and-sell model with co-building partnerships, where vendors and partners work together on customer outcomes rather than just product training and resale.
