Q&A
Diagnosing (and Curing) Microsoft Copilot Adoption Roadblocks
If a Copilot lands in your environment and no one uses it, does it make a sound?
AI is everyhere in Microsoft's stack and its mascot is Copilot, a big and growing family of features, capabilities, integrations and solutions. But just because Microsoft has made Copilot more ubiquitous, that doesn't mean enterprises will automatically embrace it with open arms. There's still a lot of trepidation about AI; organizations remain concerned about data privacy, IT pros about job security and end users about the AI learning curve. There are real benefits to using AI in business, but achieving meaningful user buy-in is easier said than done.
Solutions architect and Live! 360 speaker Stacy Deere is an expert at identifying the roadblocks to broader enterprise AI adoption, from their symptoms to strategies for overcoming them. Ahead of her Live! 360 session titled "Copilot for End Users - Getting Started," she answered some questions about how organizations can start taking advantage of AI while minimizing the friction around its adoption.
Redmond: In general, how open or resistant do you think corporate end users are to adopting AI technologies like Copilot?
Deere: Currently, it is a mix due to security concerns. Some are very excited, and they want it now and are ready to jump in with both feet. The others are concerned that data is going to be seen by individuals that shouldn't be seeing it or will be released externally to their organization.
It's up to individuals like us to show how we bring these two things together to ensure an organization can use AI technologies safely and efficiently.
AI adoption tends to hit a wall when end users get discouraged by their initial experiences. How should IT leaders make sure this doesn't happen with their Copilot rollouts?
By ensuring the company goes through a Copilot readiness evaluation and then ensure users are properly trained on prompting and utilizing.
What other roadblocks to AI and Copilot adoption should IT teams be prepared for?
The biggest issue I see is when Copilot is pushed out without proper preparation such as validating permissions. As many companies find out quick, data is being shown that shouldn't be. Then they immediately turn off Copilot. It leaves a bad experience for the end users, and then there's the scare factor for IT and management.
Do you recommend piloting Copilot in small, selective groups first before broader rollout? If so, how should IT leaders pick which users to test drive Copilot first?
Absolutely. It's a great way to keep exposure at a minimum if a security issue is identified and helps develop and enhance training. I would lean toward the heavier users of the apps in Microsoft 365 that could get the most out of it.
What kind of AI training should IT leaders give their end users to make sure Copilot adoption goes as smoothly as possible?
User adoption training to show how to use it, but also they should have a major focus on prompting. Getting the correct information back from what you are asking means ensuring you are asking it properly.
How do can IT leaders support healthy use of AI in their organizations without worrying end users that their skills are becoming redundant or replaceable?
Explain the purpose of AI overall. Some think that it replaces the thought process of individuals. AI is meant to augment what capabilities they do have, not replace them. Demo or explain how AI can take over repetitive tasks, allowing employees to focus on more strategic and creative work.