Making AI Stick: A Human-First Approach to Adoption

Most companies treat an AI rollout the way they’d treat any other software deployment: buy the licenses, enable the accounts, send the onboarding email. Then they wait for adoption to follow. But what happens when it doesn’t?

Christina Meng has watched this play out across some of the world’s largest enterprises and shared her insights on the [Un]Churned podcast. As the Champion Programs Lead at OpenAI, and before that, a key architect of Slack’s 4,000-member champion community, she has a front-row seat to what separates companies that achieve genuine AI transformation from those that stall after the initial buzz.

Her conclusion? “The organizations that are keeping humans at the center of this massive change management initiative are the ones succeeding.”

That insight has profound implications for Customer Success (CS) leaders navigating their own AI adoption challenges, as well as for the teams responsible for helping customers get there.

The AI Adoption Cliff

There’s a familiar arc to enterprise AI rollouts. An executive champions the initiative, a pilot group gets excited, and usage spikes. Then, a few months in, things plateau. The enthusiasm becomes uneven: some teams experiment aggressively, while others remain stuck at the initial demo.

This is what’s commonly known as the adoption cliff, and Christina believes it’s usually a people problem, not a product problem. “You can ask ChatGPT a question, and it’ll spit out an answer,” she says. “But the real value is moving beyond that.”

The companies that overcome this challenge don’t always have better tools. Instead, they build internal systems that support human behavior change. This is where champion programs make a difference.

What a Champion Program Actually Is (And Isn’t)

There’s a common misconception worth clearing up. A champion program isn’t the same as a customer advocacy program.

Advocates speak at your conference or join a reference call. They’re incredibly valuable, but their impact is largely external.

The champions that Christina nurtures do something different. They’re the internal nodes within a customer’s organization: the people hosting the hands-on workshop, building the internal AI strategy, teaching their peers how to incorporate new workflows into their day-to-day. Think of them as an extension of your Customer Success team operating inside the account.

At OpenAI, Christina structures this across three distinct personas:

Executive Champions are senior leaders, often at the C-suite or VP level, who are directly involved in AI transformation rather than just supporting it from afar. OpenAI brings these leaders together in small, private meetings to discuss common challenges and learn what they need to become stronger change leaders within their organizations.

Champion Leads are usually directors or heads of AI enablement. They design and implement internal AI strategies. These leaders take guidance from executives and turn it into programs that engage the wider workforce. Christina holds regular roundtables for this group, with 25 to 30 people attending. The main benefit isn’t just the OpenAI content, but also the peer-to-peer sharing among companies facing similar challenges.

Internal Champions are the practitioners on the ground: the people leading department-level trainings, hosting internal hackathons, and acting as the go-to resource for their teams. As AI adoption matures, more companies are formalizing this layer, building their own internal champion networks that mirror the structure OpenAI uses externally.

The Adoption Metric That Actually Matters

At first, Christina says, measuring the program’s success was more of an art than a science. They looked at things like attendance, engagement, and vibes.

That’s changed. The OpenAI champion program is now aligned to a single north star: activation and sustained activation.

It’s a small difference in language, but a big one in practice. Getting someone to log in and run a query is activation. Getting them to build new workflows, teach colleagues, and use AI as a default part of how they work is sustained activation. The latter is what drives renewal conversations, expansion, and genuine business transformation.

This new way of thinking is helpful for customer-facing teams. If your health scores only track whether an account is “active,” you might miss a more important sign. The real question is not just “are they using it?” but “are they building real skills and habits around it?”

What This Means for Customer Success Leaders

Whether you’re managing a small book of business or overseeing a scaled CS function, the champion program model surfaces a few durable principles.

First, identify the right people, not just the power users. There’s a considerable difference between someone who loves using a product and someone who’s willing to teach others how to use it. Champions are the latter. The people in your accounts who are already doing informal enablement are your best starting point.

Second, create the conditions for peer learning. One of the most consistent findings from Christina’s roundtables: champions don’t primarily want content from OpenAI. They want to hear how other companies are handling the same challenges. The most valuable thing you can give to your customers may not be your playbook, but access to each other.

Third, connect CS to change management. AI adoption isn’t a one-time onboarding moment. It’s a commitment to changing how people think about their daily workflows. CS teams that position themselves as partners in that journey, rather than reactive support resources, will be the ones that drive long-term impact.

The companies winning at AI transformation have the right mix of human engagement and tools, like customer education platforms and community hubs, that sustain adoption long after initial onboarding.

Go Deeper With [Un]Churned

[Un]Churned is the CS leader’s secret weapon for learning how the best in the industry are tackling the same challenges you are. Every week on the [Un]Churned podcast, host Josh Schachter sits down with practitioners and executives on the front lines of AI transformation, retention, and Customer Success—so you’re never solving these problems alone.

Listen to Christina Meng’s full episode here.

Our [Un]Churned substack goes even further. Subscribe to receive weekly insights and deep dives straight to your inbox from CS leaders and industry thought leaders who are building and learning alongside all of us.