Sponsored By

A number of challenges and uncertainties persist, including data security, accuracy of AI-generated outputs, compliance and impacts on headcount.

Ken Presti

January 25, 2024

6 Min Read
Steps for an AI practice
Monster Ztudio/Shutterstock

MSPs and channel partners of every description are grappling with questions around how to establish strong and successful practices for the implementation of AI. While there are many paths for climbing that mountain, the common denominator is to begin with the challenges and “unknowns” experienced by the customer.

According to a new report from Accenture, business leaders need a more strategic view of generative AI, as opposed to merely focusing on specific tasks and roles. This translates to an openness to redesign both organizational processes as well as the overall employee work experience.

Two-thirds of surveyed executives say they lack the technology and change leadership expertise to drive this process, and three-quarters of organizations do not have comprehensive strategies in place.

This is nothing short of a call-to-arms for channel partners.

"A lot of CXOs said they don’t even have their narrative as relates to AI, and many of them are not incentivized to create these business opportunities,” said Laurie Henneborn, who served as research lead for the Accenture report. “We’re looking at a future where digital fluency, governance, cross-functional collaboration are all necessary at the leadership level; plus, they also need to know how to develop and nurture strong ecosystem partnerships that can accelerate progress.”

Related:2023 Channel Futures MSP 501: Download the Complete List of Winners

While business leaders are drawn to using AI as a means of automation, improved efficiency and better customer experiences, a number of challenges and uncertainties persist, including data security, accuracy of AI-generated outputs, compliance, impacts on headcount, and the very pervasive fear of unintended consequences.

Fears, uncertainties, doubts, technical expertise and strategy development are all areas where MSPs and other channel partners have traditionally shown strong value, and generative AI will be no exception.

Coach the client toward taking a big-picture view of the potential role of AI within their organizations.

“Every organization should draft an AI policy even if you’re not ready to use it, because people are probably already using AI without your knowledge,” said Michael Gray, CTO of Thrive Networks, a Foxboro, Massachusetts-based MSP that is No. 21 on the 2023 Channel Futures MSP 501. “The first aspect is a very simple binary: Do we allow AI usage? We should at least communicate to our employees whether or not it’s OK to use something like ChatGPT for work. The next question is, if you’re willing to use generative AI services, should it be limited to something that has been chosen and refined by the corporate team? Then, how do these decisions impact other corporate policies? In what areas are we going to use AI? Who’s managing the [large language models]? And who’s going to enforce the policies, and how?”

Related:CF20: Cybersecurity Providers Making the Most of AI

Gray also advises that partners consider taking a vertical market approach to developing their AI-related services. A depth of knowledge of how AI can impact specific industries can go a long way toward building the most successful solutions, as well as establishing credibility in the targeted field.

Timothy Guim, CEO of PCH Technologies of Sewell, New Jersey, No. 15 on the MSP 501, agrees that a vertical approach can be effective, but also adds that some MSPs, consultants, and other partners may target specific departments within a much wider range of verticals. This opens the possibility that channel partners may use a matrix approach to building their go-to-market strategies, seeking to address various departmental needs as well as vertical needs.

PCH's Timothy Guim

Guim is starting with a focus on three or four forward-looking business owners who are willing to make the move toward AI.

“We’ve already leveraged some of these tools in our own company, so we’ve been honing the platforms internally," said Guim. "Then we look at what can we automate quickly and what can we scale. If, for example, we do something for the marketing department that works really well, we might be able to replicate that for multiple marketing departments. If we look at departmental things that are the same for all general businesses, we can scale very quickly across our client base.”

Guim added that go-to-market strategies should be designed to both an “all-in” adoption of AI, as well as an incremental approach.

“MSPs are in a position to drive adoption of AI in different industries because business owners know that AI is important, but it’s difficult for businesses to put it in place,” he said. “We can do incremental changes now but, if you really want to leverage the power of AI, you often have to reinvent the business and how it’s going to run.”

While any large-scale changes are going to need CxO approval, there are varying opinions about where partners should begin their approach.

“Usually, our white knight is an IT manager who has internal face time with the leadership people,” said Anthony Oren, CEO of New York’s Nero Consulting, No. 20 on the MSP 501. “Start there, and let them plant the seeds within the company. From there it can trickle down or climb up.”

Nero Consulting's Anthony Oren

“I tend to start with the CEO because I generally know those people,” countered Guim. “But it’s also true that people in Gen Z tend to be more familiar with these tools, and we can often find allies among that group.”

Michael Gray had a similar take on the issue.

“Starting your customer approach lower in the organization is probably a good idea because it’s more likely to give you an early win,” he said. “I think there’s so much fear, uncertainty and doubt at the high end that even if you can get them comfortable with the idea of AI, the next challenge is deciding where to start.”

In some cases, the best place to start is in reviewing the company data that will interface with AI.

“Let’s establish our source of truth,” recommended Gray. “We see a lot of organizations where we have three, four, maybe five sources of truth for the same record. There are still a lot of data-management problems out there. Oftentimes, the reason they’re running into efficiency problems is that they don’t have a good handle on their data. Business Unit A can be working with an entirely different data set than Business Unit B. Breaking down the source of truth will help you establish the remaining steps.”

The final key area for consideration is the often-overlooked facet of user training, without which successful use of AI can be something of a dice game.

“You have to learn how to talk to AI, and to phrase things in ways that are going to be most effective,” explained Oren. “There’s a certain way to prompt it so that you get what you’re looking for. Part of training should go beyond the operational aspect and help people become familiar, and even experimenting with AI. That will help you establish the necessary level of buy-in, which will be immensely important to your overall success.”

About the Author(s)

Ken Presti

Ken Presti is a technology industry veteran specializing in the use of high-value market research, reports and podcasts to help consultants, agents, service providers and vendors to more effectively help their business customers understand and evaluate Information Technologies (IT) strategies. He specializes in combining empirical data with information acquired through industry contacts to fully illustrate technology trends, business model evolution, likely outcomes, and strategies for success.

Ken also has extensive prior experience in news-talk radio, and has been featured on a variety of media outlets, including CBS Radio News and Reuters. He’s also been quoted in a variety of business publications, including Forbes.

Free Newsletters for the Channel
Register for Your Free Newsletter Now

You May Also Like