One of the biggest challenges of realizing ROI from AI is stickiness — making it easy to use, embedded in the natural flow of how people work, and trusted by its users. 64% of CEOs say success with generative AI will depend more on people’s adoption than the technology itself, according to the IBM Institute for Business Value.

We met this challenge head on at IBM with our global consulting workforce. One year ago, IBM Consulting rolled out our AI-powered delivery platform IBM Consulting Advantage. It acts as a workbench for our consultants to use AI agents, assistants and applications to augment their daily work supporting clients.

We’ve gone from 0 to over 85,000 users and growing. We have over 2,000 assistants and agents and more than 60 industry and domain specific applications, including solutions that support and integrate technology from strategic partners like Adobe, AWS, Microsoft, Salesforce and SAP. Plus, we’re seeing up to 50% productivity gained across various consulting tasks.[1]

This adoption is a direct result of making the platform “sticky,” which matters because it means we can deliver better and faster client results at a lower cost, and we can help clients achieve stickiness internally too.

Here are four keys to stickiness based on lessons learned from our own transformation plus our experience collaborating with thousands of enterprises, from Riyadh Air to Dun & Bradstreet, to scale AI responsibly and sustainably.

Embed AI into employees’ natural flow of work.

Successful adoption depends on a user experience designed to help employees get work done, within the processes and tools they use every day.

For example, consultants use repeatable methods to execute work consistently, like a cloud migration factory method with specific steps to accelerate how consultants migrate a client’s applications to the cloud. We’ve curated bundles of capabilities in IBM Consulting Advantage and embedded them within the specific steps in the method where they can provide the most added value.

We’ve also developed plug-ins between IBM Consulting Advantage and popular integrated development environments like Microsoft VScode and IntelliJ. This means our consultants who are building and testing code can use the power of AI within the familiar applications they use every day.

Cultivate a growth mindset.

Almost every process can be improved with the right ideas, talent and resources. Cultivate a growth mindset within your teams and encourage them to continually innovate and find new applications for AI.

We’re constantly adding new capabilities to IBM Consulting Advantage based on the challenges we see clients facing today and anticipate in the future.

For example, we're seeing increasing need for smaller, curated foundation models that can help with complex industry-specific tasks, such as achieving compliance with industry standards or modernizing code within the constraints of a highly regulated industry. These models need to be customizable with client data, easily orchestrated with agents and small enough to be easily hosted and deployed via on-premise cloud.

We’re building one-of-a-kind industry specialized datasets that can be used in combination with the novel InstructLab approach to train targeted foundation models for industry tasks. We’re starting with the banking industry, in close collaboration with the Banking Industry Architecture Network (BIAN). We’ve also begun to engage clients on applications of this solution, making them part of the innovation process.

Establish trust.

Employees need targeted training to use AI most effectively. They need to be empowered to ask questions about AI’s outputs, source data and more, so they can feel confident in what AI is producing for them. This is especially important as AI agents, which can come up with a plan and execute it autonomously, become more prevalent. And organizations need to establish feedback loops with employees to understand what’s working with these tools, what’s not, or other ways they could be applied.

In IBM Consulting, in addition to providing targeted training on AI for consultants, we solicit feedback via open comment boards, slack channels and regular user interviews and research. That helps inform future platform capabilities.

Make employees creators.

Right now, many employees still see generative AI as something that’s happening to them, not as something that works for them. So how do we make people feel like they’re part of the solution?

Part of the key is an employee culture that embraces change. Find the technology pain points that inspire people to push back or find workarounds, and improve that experience. Establish psychological safety nets so people don’t fear automating themselves out of a job but instead see the tools as an opportunity for their career growth. Lastly, focus on making your people creators, because the people closest to the work often have the best ideas about where and how AI can provide value.

With IBM Consulting Advantage, any consultant can create and train an AI assistant to solve a problem, and that assistant can then be shared and up-voted by peers who want to recommend it for others to use.

Looking ahead into 2025, those business leaders who can make AI “sticky” in their organizations and combine the powerful technology with human expertise and behavior change will be the best positioned to see their AI investments bear fruit.