Quick Summary: New research shows AI in customer success is still early-stage, with most teams experimenting. Productivity is the entry point, but leaders must tie pilots to bigger outcomes, focus on high-impact workflows, and prove ROI.
AI adoption is increasing. But scale, structure, and measurable outcomes are still lagging, most teams are still early in their journey, and customer leaders are struggling to translate experimentation into measurable business impact.
That’s according to brand-new benchmark research from Rod Cherkas, CEO of Hello CCO, and author of The Chief Customer Officer Playbook and REACH.
For our latest webinar, Rod joined ChurnZero CCO and CPO Abby Hammer to unpack the findings of Wave 1 of his AI in Post-Sale Benchmark Report, based on a detailed survey of nearly 200 customer success and post-sale leaders.
This research suggests that AI presents less of a technology problem than a leadership and execution challenge. Meanwhile, expectations of AI-powered retention, renewals, and efficiency boosts are rising.
The good news, he points out, is that as a customer leader, you’re not being asked to predict the future of AI. The ask is that you start with available tools, pick meaningful workflows, measure results, tell a clear story, and bring your team along thoughtfully.
You can watch the webinar in full here. Scroll down for more key findings, plus Rod and Abby’s expert AI tips for CS leaders.
AI in customer success teams: where we stand in 2026.
Takeaway 1: Most CS teams are still in the experimentation phase.
Despite the industry buzz around AI, most teams are still early in their journey. Nearly every organization in the study is using AI in some capacity, but adoption is largely limited to pilots and informal experimentation.
About one-quarter of teams are experimenting informally. Roughly 45% are piloting specific use cases. Only a small minority have fully embedded AI into operational workflows.
“There’s a lot of experimenting,” Rod said. “There’s a lot of piloting going on to see what works and what sticks.”
So, if you’re still experimenting and you’re feeling behind… the data suggests you’re probably not. “The research shows we’re not as far behind as we think we are,” says Rod.
Dos and don’ts for leaders
Don’t: Wait for a company-wide mandate. Even if only 22% of teams have a formal AI strategy, you can create direction within your function.
Do: Be intentional about where you are piloting. Design those pilots so you can measure results, instead of letting experimentation remain informal and unseen.
Related: How to get your company started, not stuck, with AI.
Takeaway 2: Productivity is the entry point, not the long term value.
Today’s AI usage in CS is concentrated in productivity tasks. Teams are using AI to summarize meeting notes, draft emails, prepare for customer calls, and research accounts.
These use cases are useful because they lower the barrier to adoption and help teams get comfortable with AI tools. However, they rarely justify investment on their own
“I think of productivity as the on-ramp, not necessarily the destination,” says Abby. While it isn’t a bad place to start, productivity alone is rarely enough to justify long-term investment. Time savings are often invisible to executives.
“It’s very hard to be able to explain to your leadership team what business outcomes are you driving with productivity alone,” says Rod. Ideally, you can make the shift from time saved to business impact.
Dos and don’ts for leaders
Do: Treat productivity as an on-ramp. Build trust by starting with lower-stakes use cases, then deliberately redeploy saved capacity into retention, expansion, or forecasting improvements. Pick one metric that matters to your CEO or CFO. Crucially, connect time saved to business impact.
Don’t: Stop at productivity gains. You can’t assume that time savings will speak for themselves, so beware presenting AI as efficiency without tying it to measurable outcomes.
Takeaway 3: Execution is the hard part.
Leaders see high potential for AI in account prioritization, renewal insights, and identifying expansion opportunities. Yet, execution lags behind aspiration.
“What this tells us is that people recognize that there’s value,” says Rod. “But they may not know how to apply that.” In other words, awareness isn’t the issue; the structure to operationalize the ideas is.
Dos and don’ts for leaders
Do: Avoid trying to solve everything at once. Focus on one painful workflow tied to revenue, capacity, or efficiency. Move that workflow through clear phases: experiment, prioritize, operationalize, and measure. Once you prove it in one area, repeat the process elsewhere.
Don’t: Try to solve every AI opportunity at once and let your ambition outpace your structure. And don’t confuse strategic potential with execution.
Related: When is it time to centralize your customer team’s AI?
Takeaway 4: AI progress is fundamentally a leadership challenge.
Many organizations assume AI progress will be constrained by tools or technology. In reality, thought most companies already have access to AI capabilities.
“It’s not really a tool problem,” says Rod. “It’s more of a leadership challenge.”
Teams that are making progress tend to share a few traits: clear ownership, defined expectations, and visible metrics. Meanwhile, organizations where AI adoption is optional or undefined tend to stall.
“AI hasn’t created new leadership problems,” Abby points out, “but it has really magnified existing ones.”
Dos and don’ts for leaders
Do: Take ownership. That means deciding whether AI usage in your team is optional or expected, and measuring your ROI intentionally, even if early signals are qualitative. Make sure to tell the story of impact to your peers and execs. If you don’t make AI wins visible, they’ll get absorbed into “productivity” and go unnoticed.
Don’t: Jump to blaming your tools for lack of progress—remember, it’s often a leadership issue. Nor should you leave adoption entirely to frontline experimentation.
Takeaway 5: Frontline champions drive much of the innovation.
One of the most encourageing findings from Rod’s research is where the innovation is happening. CSMs and support engineers are building workflows, testing agents, and experimenting on their own. Leaders who encourage this behavior are often the ones who uncover the most practical use cases.
Abby added an important nuance, however: “Bottom-up innovation is great, but without top-down direction, it’s just a lot of energy.”
Dos and don’ts for leaders
Do: Identify your frontline innovators and AI champions and give them space, visibility, and small budgets if needed. Create a recurring forum for sharing what works. Then, prioritize the highest-impact ideas and standardize them so your team moves in the same direction.
Don’t: Suppress experimentation or expect innovation to originate only from leadership. At the same time, beware allowing energy to scatter in different directions without alignment.
Takeaway 6. Executives and boards want proof, not AI hype.
While AI excitement is high across the industry, executives and boards are focused on a simpler question: what measurable outcomes does this drive?
Rod emphasizes that AI investments should be treated like any other business investment. You should be able to clearly explain the cost, the expected return, and the metrics that prove value.
Instead of presenting AI as a broad transformation, he adds, it’s more effective to tell a specific story about a workflow improvement and its business impact. Start small, prove value, and then expand.
“Start with one or two clear stories of how AI improved a workflow or unlocked capacity,” Rod advised. From there, quantify the impact—whether that’s time saved, cost reductions, improved forecasting, or revenue outcomes.
Working with finance leaders early can also help ensure the story resonates when it reaches the executive team or board.
Dos and don’ts for leaders
Do: Start with one or two measurable AI success stories. Quantify the impact in terms executives understand: revenue, cost, risk, or efficiency, and treat AI initiatives like any other ROI-driven investment decision.
Don’t: Lead with “vision slides” about the future of AI, present experimentation without measurable outcomes, or over-promise impact before results are proven.
Related: Why AI is no longer optional: Insights from SaaS Capital’s 2025 B2B Benchmarking Survey
The real shift AI is creating in customer success.
AI isn’t about replacing customer relationships. It’s about reshaping how work gets done. And, despite concerns about data quality or adoption barriers, neither should stop teams from starting.
“The first person who tells me their data is perfect,” Abby says, “I’ll tell them I’m not going outside to watch pigs fly.”
In other words, perfection isn’t required to begin. What matters is momentum. “It’s not going to slow down,” Rod says. “So I encourage you to just start somewhere.”
And as Abby puts it, the real opportunity is freeing humans to focus on what matters most.
“It really is an assistant,” she says, “clearing the space for humans to do the truly unique human things.”
Get more actionable advice on building and scaling AI use cases in ChurnZero’s new AI guide for customer teams.
AI in Customer Success Q&A: What else did attendees want to know?
The live Q&A surfaced the practical concerns leaders are facing. The questions focused less on futuristic scenarios and more on budgeting, data quality, team resistance, and executive expectations.
Here’s a summary of the key themes and how Rod and Abby addressed them.
1. How do we build the business case when AI has a cost?
Several attendees asked about budget pressure, especially when AI tools are priced per user or usage-based. The guidance was clear: do not lead with hype. Lead with measurable impact. Instead of asking for AI budget in the abstract, tie it to a metric that matters to your CEO or CFO.
Rod emphasizes working with your finance partner first. Treat AI like any other SaaS investment: define the cost, estimate the return, and measure against it. Storytelling, backed by numbers, helps secure investment.
2. What if our data quality is not good enough?
Data quality concerns came up multiple times. Some leaders wondered whether they should wait to invest in AI until their data is cleaner.
The answer: do not wait for perfect data. Rod points out that not every use case requires pristine inputs. Start with workflows that do not depend on highly structured data. Abby added that data enrichment itself can become the first AI use case.
The larger point: imperfect data is nothing new. Don’t let it stall your progress.
3. How do we use AI across multiple data sources?
Start with what you already have access to, Rod suggests. Many organizations already provide enterprise licenses for tools like Copilot, Gemini, or ChatGPT, along with AI embedded in platforms like ChurnZero, Salesforce, and Gong.
Before pursuing complex integrations, maximize the tools and data sources already approved and available. Begin there, learn what works, and evolve over time.
4. What if part of my team is resistant to AI?
Rod acknowledges that change management is a real challenge, and that not everyone will move at the same pace. Leaders should identify early adopters and frontline innovators first. Over time, AI fluency is likely to become an expectation in hiring and performance.
Abby encourages leaders to approach resistance with curiosity. Is the concern about job security, prior bad experiences, or fear of losing the human side of CS? AI today functions as an assistant, reducing administrative load and creating space for uniquely human work.
5. How do I talk to my board or leadership team without over-promising?
The guidance: avoid future-state vision slides. Focus on proof. Start with one or two clear, measurable use cases. Share tangible examples of impact.
Rod reinforces the importance of storytelling to give executives confidence that AI is being used deliberately and responsibly. Showing measured progress builds credibility—while overpromising creates risk.
