Google Cloud Next 2026: Expanding AI Agent Adoption Requires Culture Shift

During Google Cloud Next 2026, he stressed the importance of AI adoption not only as a technical undertaking but also as a cultural one. “You're not going to just activate a license for Gemini Enterprise for someone and then send them a link to some on-demand training and expect that they're just going to start using it,” Clishe said. 

Organizations should focus on these four areas to ensure they get the most value out of their AI initiatives: 

• Executive sponsorship 
• Modernizing outcomes 
• Generating momentum 
• Continuous feedback 

“When these things are present, success usually follows,” Clishe said. 

READ MORE: Strategic partnerships turn AI ambition into action. 

Empowering a Workforce to Trust and Rely on Agentic AI 

Starting with executive sponsorship, Clishe said when senior leadership shows personal interest and enthusiasm, that can help the teams’ confidence that AI adoption is not just for optics. 

“When those goals are defined from the top down, and users and staff feel like they have agency in helping to create and craft those goals, we've seen that lead to success,” he said. 

Next, on modernizing outcomes, Clishe explained how that was tied to creating valid use cases fit for an organization. That may align with use cases already shared by Google Cloud so far, but it can also be unique to an organization, crowdsourced from the employees using the tools themselves. “Your individual users, as they get excited and start using the product, they'll build their own agents, and they'll come up with their own use cases, and they'll begin to share those,” Clishe said.

CDW offers an AI use case ideation workshop that can help organizations home in on what would work for them and create a framework on how to score various use cases. Some use cases may benefit the entire organization, while others may be department-specific. And part of brainstorming on those use cases includes thinking about data availability. 

“You have the AI agent, you know what you want it to do, you know it can do it, but upstream, the data just isn't ready. Maybe it's in different systems or it's not in the right format, so there could be a lot of friction there,” Clishe said. 

Understanding data availability can help guide organizations toward use cases that are more straightforward and would provide quicker wins. 

“You're not teaching a tool, you're teaching a new way to solve an old problem. What you're really doing is creating new business processes from the beginning,” he added. 

Then, Clishe said, generating momentum is likely where organizations will spend most of their energy. To keep interest going, they can build a champions network to encourage enthusiastic, tech-savvy users to continue with advanced training, get early access to features and provide more actionable feedback. 

DISCOVER: Common missteps to avoid when launching your enterprise AI agent. 

“Users are going to struggle every time they prompt the tool or ask for a solution,” Clishe said. “They're going to need to iterate on it, and they're going to need a way to provide feedback. Be patient with them, but also encourage that culture of curiosity, because users can provide feedback when something isn't going well.” 

Organizations should also share success stories widely and encourage willing users to share their built agents internally. 

Finally, continuous feedback is straightforward. Clishe added, “This is not a one-time install. This is something that needs that feedback loop. You're going to need to continue to iterate on it, really, for the duration of the lifecycle of the product.” 

Measuring key performance indicators is key, and reporting on them to teams so they can see the progress also helps. 

Closing out, Clishe shared that CDW has various adoption services to help organizations with their projects, such as running hackathons, high-scale training and smaller, white-glove training for champion users.

Click the banner below to keep up with the latest BizTech event coverage. 

An Accelerated Timeline for Macy’s AI Agent 

During a separate session, Google Cloud leaders shared a success story about Macy’s newly launched customer experience AI agent that shoppers can use on the retailer’s website. 

The “Ask Macy’s” button sits close to the search bar. If shoppers want a more personalized, multifaceted experience, they can use the AI agent to find clothing options within a budget range for an event, for example, or home decor options to refresh a bedroom for spring. 

Archana Kannan, senior director of product management at Google Cloud, walked through a demonstration of the AI agent on stage. She said that the outfit she wore for the talk was chosen with the help of the agent, by starting with a prompt that included her height, the time and location of her event, and specifying that she wants a look that can transition from day to night. 

“Instead of pointing me to a generic help center page, the agent understands the exact context of my order and comes up with a logical resolution,” she said. “It presents options of either picking it up from the store or shipping it to my home. This is where digital and physical commerce comes together as one continuous experience.” 

EXPLORE: Understand the basics of agentic AI and application modernization.

The agent could also verify real-time inventory to ensure that the shoe size she needed was available at the location where she would pick up her order. “In one natural interaction, we have completed discovery, purchase, service and store fulfillment,” Kannan said. “This is fluid commerce, and this is the standard that your customers are going to expect.” 

After the demonstration, Macy’s Senior Vice President of Customer Experience Chad Westfall answered a few questions from Darshan Kantak, Google Cloud’s vice president of product for applied AI, about the Ask Macy’s rollout. 

Westfall noted that using Gemini Enterprise for Customer Experience sped up the process significantly, from planning to deployment. 

“The team worked incredibly quickly, and Google consistently committed the right technical resources to ensure the implementation stayed on track,” he added. 

Since the launch, Westfall said customer sentiment toward the AI agent has been positive. One interesting statistic has been that order sizes have increased for shoppers that engage with Ask Macy’s compared with the average cart. 

“As we scale, we're keeping a close eye on how we're looking at those metrics. The initial engagement is really great from a signal perspective, but we look forward to gathering more data over time to see how this impacts our core KPIs. We are specifically looking at things like conversion rates, units per transaction and, of course, revenue. But ultimately, we're really excited to see how these AI-driven interactions actually help our long-term customer value and ultimately work,” Westfall said. 

The AI agent also reduces the disconnect between what’s available online and the in-store inventory. 

“When an AI agent can instantly verify inventory and secure a product for pickup right in the flow of that conversation, it transforms a digital interaction to a physical interaction and brings foot traffic into the store,” Westfall said. 

GET THE DETAILS: How can small businesses bridge the AI readiness gap?

As the retailer looks toward the future, Westfall said, the AI agent is helping move the company from “a reactive service to a truly proactive discovery for our customers.” 

“We're looking at the ways that the agent can anticipate the needs of our customers, before they even come online to search; for example, proactively reaching out to a Star Rewards member to help them understand about a new collection that perfectly matches their unique style profile, or seamlessly preparing a fitting room based on an ongoing chat that that customer's having,” Westfall said. “We aren't just keeping up with retail trends, we are actually redefining the future of how customers experience Macy's.” 

Google Cloud’s Kantak closed the session by highlighting how AI agents can be built with speed to keep businesses agile. 

“Building an enterprise agent with that kind of depth used to take months or even years. We've compressed that timeline down to days,” he said.