AI technology is transforming at an exhilarating rate. The impact and ROI it delivers to organizations, less so, stymied by the same old challenges we always face. AI projects have a real people problem.
What it actually looks like
An email from leadership. A new tool, a go-live date, and little else. No explanation of what it means, no answer to the one question everyone’s actually asking: “what’s in it for me?” Then an interminable, entirely forgettable group training session over Zoom, the kind where three people are on mute, eating lunch, and one screen is just a cat.
That’s it. Box checked, training done. No reinforcement, no coaching, no follow-up, no support, no touchpoints, no check-ins to ensure the AI is augmenting and streamlining workflows as it should. Whatever people picked up starts eroding within weeks, because nobody planned for anything past the training date.
Sound familiar? It’s a “launch and leave.”
This is too often the experience of the very same staff who are expected to use these tools every day and deliver the benefits back to the company, benefits that were never properly communicated to them in the first place. Frontline staff were never involved. The why and the vision never got socialized past the leadership room, communication didn’t land and training showed up too late to matter and rarely met people where they actually were.
This isn’t new
I’ve watched this same pattern for years on CRM rollouts, ERP rollouts, and custom software implementations: these transformation efforts very rarely fail because of technology.
McKinsey’s own research backs that up. Eighty-eight percent of organizations now use AI somewhere in the business, but nearly two-thirds are stuck in pilot mode, and only about six percent see real bottom-line impact from it. That means roughly ninety-four percent of the AI hype cycle is currently running on vibes alone. The common thread in why these efforts struggle isn’t technical, it’s human: unclear goals, a “why” that never got communicated, and change that nobody bothered to sustain past launch. (Source: McKinsey, “The state of AI in 2025: Agents, innovation, and transformation,” QuantumBlack, AI by McKinsey, November 2025.)
If you’re mapping out your own AI rollout, our guide to building an enterprise AI strategy walks through the same planning gaps.
Why AI Raises the Stakes
AI raises the stakes. A new platform interface asks people to learn where buttons have moved. AI asks them to trust a judgment call that came from a model instead of their own experience, and that’s a much bigger leap, one that usually gets the exact same one-hour Zoom session and a PDF nobody opens.
People who were never given a reason for the small asks aren’t going to get one handed to them for the harder one.
The actual point
The answer is simple. Change and transformation involve putting people first, always. Focusing on people should be the lens everything else gets built through, not an afterthought once the tech is scoped and the timeline is set.
In practice, that starts with one question, asked before a single line of the rollout plan gets written: who’s explained to the frontline team why this matters to them, and has anyone actually confirmed they heard it. Not sent an email. Confirmed. If the answer is no, that’s the first thing to fix, before the go-live date, before the training deck, before any of it.
AI will keep getting smarter, faster, and more capable. That’s not in question. What determines whether any of it actually pays off is whether the people expected to use it were ever brought along in the first place.
As a Certified Change Management Professional, I’ve evolved this into what we call Success Enablement at Faye, on purpose. Change management still gets heard by a lot of leaders as a delay, a layer of process on top of the “real” work. Success Enablement is the same discipline, aimed at the same outcome, but framed as what it actually is: the work that determines whether the investment succeeds at all.
If you’re leading an AI initiative and want a partner who treats change management as core to the plan, not an afterthought, talk to Faye’s AI consulting and integration services team about Success Enablement for your rollout.
Key Takeaways
- AI change management, not the technology, decides whether an AI rollout pays off. McKinsey’s 2025 survey found 88% of organizations use AI, but only about 6% see real bottom-line (EBIT) impact.
- The most common failure pattern is a “launch and leave” rollout: one training session, then no reinforcement, coaching, or check-ins after go-live.
- AI raises the adoption stakes higher than past software rollouts, because it asks employees to trust a model’s judgment call, not just learn where buttons moved.
- Frontline staff need their “why” confirmed, not just emailed, before a single line of the rollout plan gets written.
- Faye calls this discipline Success Enablement: the same change management rigor, reframed as the work that determines whether an AI investment succeeds at all.
Frequently Asked Questions
Who should own AI change management inside an organization?
It shouldn’t sit solely with IT or L&D. The strongest AI rollouts assign ownership to someone who can speak for both the business case and the frontline experience, often a change management or Success Enablement lead who reports on adoption alongside the technical rollout, not after it.
How can you tell if an AI rollout’s change management is actually working?
Usage data is the first signal, but it’s a lagging one. Watch earlier indicators: are employees still asking questions weeks in, are managers reinforcing the tool in team meetings, and can staff explain the “why” in their own words. When those go quiet, the rollout is sliding into launch and leave.
Does AI change management matter for small teams, or is it only an enterprise problem?
The principle scales down even if the formality doesn’t need to. A 15-person team can suffer a launch and leave just as easily: one Slack announcement, no follow-up, quiet abandonment. Smaller teams often skip change management because it feels unnecessary, which is exactly how adoption stalls without anyone noticing.
AI governance consulting, for a Claude rollout, means turning ad hoc AI use into a deployment with defined identity controls, Read more
Why the smartest people in the room weren't talking about smarter AI — they were talking about cleaner CRMs. London Tech Week Read more
Your AI is live. The chatbot is answering questions. The predictive model is running. Early results? Promising. Here's the question no Read more
Intercom is one of the leading platforms for managing customer communication, support, and engagement through real-time messaging and AI-driven tools. Read more