How to lead successful AI adoption in your customer success team

As a customer success (CS) leader, I found myself facing a challenge that many others are probably navigating too: AI was suddenly everywhere, but my team wasn’t sure how, or even if, they should be using it.

I wanted to help them feel more confident, more supported, and more equipped to work smarter with AI. That’s what set me on this path: researching how to implement AI in a thoughtful way, testing what worked in real team environments, and sharing what I learned along the way.

Why you need an AI strategy now

We’re truly in a transformative time where AI is being migrated into job roles globally. It’s not just about having the tools, but about how we introduce them in ways that actually make work better for people.

Of course, it's not just customer success being impacted by this revolution. One study from the World Economic Forum in 2023 found that 75% of companies plan to integrate AI into job roles within the next five years. And Goldman Sachs reported that AI could alter 300 million jobs globally. That’s massive.

Yet, even with all this momentum, a study of 550 companies showed that 96% faced challenges with long-term AI adoption. Another study of 1,600 professionals found that 41% experienced resistance to using AI tools. What’s driving this hesitation? A major theme: poor change management.

Common barriers to successful AI adoption

When I dug into the research, three main barriers kept coming up:

1. Lack of understanding: Teams weren’t sure how to use AI tools or how to measure their impact.
2. Resistance to change: There was fear, discomfort, and uncertainty about AI.
3. Poor integration: Many companies were layering AI on top of existing processes without aligning it to workflows.

We’re in this interesting paradox – there’s a huge surge of interest in AI, but actual adoption remains pretty low. That’s where thoughtful change management really comes in.

These findings echo what we’re seeing in the field. In the State of Customer Success 2025, only 15.9% of CS teams are automating adoption nudges, while 65.9% still rely on human-led check-ins. And despite generative AI being on the scene for the last few years, more than half of customer success teams globally aren't using these AI tools.

It’s clear that even with the tools available, meaningful change takes thoughtful leadership and team enablement.

The role of change management

Change management isn’t new. If you’ve been through a merger, reorg, or big process change, you know how messy it can get. People need support through change. McKinsey found that only 35% of change initiatives succeed, and when they don’t, the impact is real – burnout, change fatigue, talent loss, and inefficiency.

That got me thinking: how can we be more thoughtful about bringing AI into our workflows?

Starting with the right AI strategy

I didn’t have a background in AI strategy – I definitely wasn’t an expert. So I decided to take the initiative and start learning.

I came across Dr. Andrew Ng, a global leader in AI, and took his online course, AI for Everyone. It was a fantastic intro to AI strategy and helped shape my approach.

One of the biggest takeaways I learned? Focus on tasks within a job role, not the full workflow.

Step 1: Focus on tasks, not workflows

At first, I thought I’d just layer AI onto our existing customer journey. But after learning more, I realized the better approach was to identify specific roles on the team – like CSMs, onboarding specialists, support reps – and map out the tasks they do every day.

For example, our CSMs:

• Handle tons of emails (inbound and outbound)
• Build success plans
• Create executive business reviews (EBRs)
• Drive customer value
• Mitigate risk
• Support renewals

This gave me a clear picture of where AI might help.

Step 2: Listen to your people

The next step is to just listen to the people you support. That was an easy one for me – my team is really open and honest about where they’re getting stuck.

Some of the feedback I heard:

• “It takes me forever to build presentations.”
• “I just got assigned a new account and need to ramp up fast.”
• “There’s so much info scattered everywhere – I don’t know where to start.”

Mapping these pain points to the tasks helped me prioritize where AI could add value.

Step 3: Do diligence mapping

Another best practice I didn’t have prior experience with is something called "diligence mapping," and I’ve never seen it in any other change management models. But it turned out to be incredibly valuable.

Diligence mapping means partnering with someone who deeply understands the technical side of AI at your organization. For me, it was our business systems leader – he knew which tools we had, how they worked, and how to track adoption.

We started meeting weekly. I’d share task maps and pain points, and he’d help me figure out how our tools could address them. That partnership made all the difference.

Step 4: Start with what’s most adoptable, not most impactful

This one might feel counterintuitive.

Traditional change models tell you to start where the pain is greatest. But with AI, the best place to start is where you’re most likely to get adoption. What I personally found is it’s all about micro wins – those short, baby steps that help build momentum.

For us, that was presentations. We had just refreshed our customer books, and we needed high-impact decks focused on value. So we used NotebookLM, a free AI tool that helped us synthesize a large amount of information, like previous calls, emails, and client websites, and then guide CSMs with clear prompts.

That small win sparked excitement. From there, we expanded to:

• Writing faster, more polished emails
• Identifying and addressing risk
• Using tools like Gemini for problem-solving

We also leaned into built-in features like Atlassian’s Ask AI in our Wiki to help surface the right content faster.

Step 5: Lead the change intentionally

This is the part where change management really matters.

Remember those three barriers – lack of understanding, resistance, and poor integration? I made it a specific point to be really thoughtful about those barriers and try to counteract them wherever I could.

Build understanding

I overcommunicated the why. We had working sessions, one-on-ones, and team meetings – all focused on giving people a safe space to learn and try.

Normalize AI use

I framed AI as a helpful tool, not a job threat. I showed I was invested in their development, and I encouraged champions on the team to share their wins.

Leverage the tech partner

That continued relationship with our technical leader helped us stay aligned on what tools made sense and how to use them.

Apply best practices from change management

One change management best practice I leaned on was called the Rule of Seven, which actually comes from marketing research. It essentially means that people need to hear something about seven times before it sticks. 

So we made AI a regular part of our conversations. It wasn’t a scary “new thing” anymore; rather, it became part of how we worked.

Model what good looks like

As a leader, you set the tone. If your team is using AI, you should be using it too. It’s really about modeling a mindset of curiosity and openness, so that becomes the normalized behavior across the team.

Ask questions, stay curious, and be open to learning alongside them. When you lead with that mindset, it gives your team permission to do the same.

Results and takeaways

And what I was able to see is that by following this approach – mapping tasks, listening to pain points, and layering in AI step-by-step – we went from zero adoption to over 80 instances. Now that felt like real traction.

Here are my key takeaways if you're looking to implement AI thoughtfully in your team:

• Start small: Focus on specific tasks and pain points.
• Partner smartly: Do diligence mapping with technical experts.
• Prioritize adoption over impact: Quick wins build trust and momentum.
• Lead the change: Communicate often, normalize use, and model the behavior you want to see.

AI is here to stay, but how we implement it will define whether it helps or hinders. Thoughtful strategy and change management are the keys to making it work, not just for the business, but for the people in it.

This article is based on a session Christine delivered at our AI for Customer Success Summit. You can watch Christine's full presentation – and others like it – OnDemand with a Pro+ membership plan