Just like its predecessors, the internet and the cloud, AI is on the classic technology adoption curve. Businesses vary widely in how far they’ve integrated AI into their customer service. And the customers these businesses serve have varying degrees of comfort when interacting with AI-powered experiences.
Because customers are at different stages with AI adoption, businesses need to meet them where they are. Customer support leaders face an inflection point: customers are hopeful that AI can improve their experiences, but many feel it’s not quite there yet.
To better understand these customers’ current and future expectations on AI-powered customer service, we surveyed 4,500 desk workers to find out where AI excels and falls short. Turns out, how often customers use AI at work greatly influences their AI outlook. Our study revealed three personas:
Customer persona | AI adoption rate | AI sentiment |
|---|---|---|
Adopters | Use AI daily | High usage, high expectations, and crave smart automation |
Dabblers | Use AI weekly or monthly | Curious but cautious, evaluating where AI can add value |
Traditionalists | Use AI less than monthly or never | Skeptical, prefer human-first support and transparency |
AI is enabling support teams to do more with less, particularly amid top-down pressure to cut support costs and boost operational efficiency. But it’s still early for AI-powered customer service: Customers choose other service channels. Resolution rates may be inflated. Self-solvable tickets are still coming through.
That’s why humans still play a critical role in building AI-powered interactions that resonate with their customers. Without real human understanding to guide AI, customers wouldn’t have a thoughtful or authentic experience. Understanding the AI adoption bell curve among your customer base gives you additional insight into what they expect.
While 58% of customers believe AI can improve customer service, there’s room for improvement:
80% think AI struggles with solving problems the first time
78% claim AI has trouble understanding issues
71% wish they could solve their problem without needing a human
Let’s take a closer look at each persona to understand their needs and preferences in an AI-powered experience.
research: The state of service expectations
We surveyed 4,500 desk workers to find out what they really want from customer service, AI-powered support, and more.
AI adopters engage more, so they expect more
Your most AI-savvy customers aren’t just conversing with your chatbot — they’re benchmarking its performance and usefulness against other AI tools they use for work, like ChatGPT, Claude, Gemini, etc.
52% of Adopters are very satisfied with AI-powered customer service and are 2X more likely to say so than the average customer. But they’re also more critical of its limitations and believe there are areas for improvement:
In other words, Adopters aren’t blindly optimistic. Their standards are higher than anyone else’s because they understand what AI-powered service could be.
Where AI can shine for Adopters
Adopters are 6X more likely than Traditionalists to agree that businesses using AI care more about their customer experiences than those that don’t. Adopters see AI making the customer experience more seamless across service channels and enabling more proactive, personalized support:
How to earn Adopters’ trust
Break down barriers to customer context
Customer profiles. Historical conversations. Account health. It’s essential that customer data is easily accessible across your tools and systems. An integrated tech stack makes it easier on your team to get the full picture of the customer to provide more effective support.
Make support predictive
Leverage AI to study your customers’ behavior and analyze their sentiment to offer solutions before problems arise. Track trends within your customer conversations, in-app activity, or customer feedback surveys to detect early signals to personalize support.
Expand your self-service options
77% of Adopters crave more robust self-service options, so investigate where you need to invest next — whether that’s in an existing self-service channel or spinning up a new one. For example, analyze your tickets to feed back into your chatbot logic to improve its resolution rates.
Ensure your omnichannel experience is seamless
Centralize all customer conversations in a unified workspace. Aggregating all channels — email, chat, phone, SMS, social media — into one view provides the historical context needed to resolve issues faster.
Dabblers are at a tipping point: Win trust or lose them
Dabblers are the undecided middle of your AI-powered customer experience journey — and the ones you can sway the most.
These “swing” customers are on the fence: 42% are somewhat satisfied with AI-powered customer service, whereas only 24% are very satisfied.
There’s a big opportunity for businesses to win over Dabblers — if they can improve in the areas Dabblers care about most.
Where AI-led support often fails Dabblers
Despite their curiosity and occasional use, Dabblers are quick to lose trust when the AI-powered experience feels clunky or impersonal. Here’s where AI-led support falls short with Dabblers:
The human handoff is slow or unclear
The AI doesn’t sound trustworthy or human enough
Resolution isn’t possible without involving a human
How to earn Dabblers’ trust
To convince Dabblers to turn to AI-powered customer service when they need help, businesses need to check off a few boxes:
Nail the basics
Accuracy, clarity, and human fallback should be flawless. Ensure your AI has a reliable learning source to offer correct and clear solutions to customers. And if it can’t provide an answer, there’s a quick path to a human agent.
Track the service quality of your AI-powered customer service |
|---|
AI is handling more customer support queries, but visibility into AI’s service quality hasn’t kept up. The new metric, AI Experience Impact Score (AXIS), measures three common friction points of AI-led customer interactions:
It’s an easy-to-interpret metric on a scale of 1-5, allowing support leaders to track trends in their AI-led customer interactions and pinpoint where the experience can be improved. Download the AXIS white paper. |
Reinforce trust
Be transparent when AI is used — and when a human is ready to step in. Indicate when a customer is interacting with AI to set expectations, whether that’s labeling your chatbot to distinguish from human agents or proactively communicating with your customers when AI assists with providing solutions.
Use AI as a guide, not a gate
Design support workflows that assist, not block, resolution. For example, if you don’t offer live chat but your chatbot recommends that you contact a human agent, have it provide the correct support email address or embed the contact form within the chat widget. Be deliberate when delegating tasks to AI that it can actually solve, and don’t delay escalation to a human when necessary.
Standardize tone and language
Avoid cold or robotic replies — Dabblers appreciate empathetic support. 55% agree that businesses are empathetic towards their customers when they give their chatbot personality. One quick tip is to train your AI on your support team’s voice and tone guidelines (use ours as inspo if you don’t have one yet).
Traditionalists avoid AI, but still demand excellence
Traditionalists may be your toughest customer to get on the AI train — they’re skeptical but not apathetic. Here’s why Traditionalists avoid AI-powered customer service:
They fundamentally don’t trust AI in service interactions
Traditionalists are 2X more likely than the average customer to doubt AI can improve customer service. This shows deep-rooted belief-based rejection, not just lack of exposure.
They’ve had little to no experience with it
Traditionalists are nearly 3X more likely to say they’ve never used AI-powered customer service. And when they give it a go, they’re more than twice as likely as the average customer to be very dissatisfied with it.
When they do seek service, they prefer traditional channels and peer solutions
52% are more likely to prefer contact forms to get support
20% are more likely to prefer getting help by phone
Traditionalists are 2X more likely to prefer peer-led help, e.g., sources like Google, Reddit, YouTube, over company help.
Traditionalists have baseline expectations despite AI hesitation
Even if Traditionalists don’t use AI today, they will increasingly engage with AI-powered customer service — and they already have some baseline expectations for their experience:
Above all, Traditionalists are concerned about how their data is being collected, used, and stored when interacting with AI-powered customer service.
How to earn Traditionalists’ trust
This is where simplicity shines. Offer Traditionalists transparency, choice, and not-too-much personalization.
Never hide when AI is being used: Label AI assistance clearly and don’t bury handoff options.
Make escalation to humans seamless: Instead of having customers type “agent” immediately upon entering a chat, offer a button that immediately connects them to human help.
Don’t over-rotate on personalization: Overpersonalizing experiences may cause data privacy concerns. Focus on where personalization can help with more effective resolution, like remembering customer communication preferences to reach them more effectively.
Your customers aren’t a monolith, nor should your customer experience be
So, what does this all mean for crafting your AI customer experience? It boils down to this: Your customers aren’t a monolith. You’ve got your tech-savvy Adopters who expect AI to anticipate their needs. Then there are the Dabblers, curious but cautious, needing reassurance and easy escape routes to human support. And let’s not forget the Traditionalists, who prioritize good old-fashioned service and might need some convincing about AI’s benefits, focusing on tangible outcomes like efficiency and problem-solving.
The big takeaway here? How ready your customers are for AI directly impacts what they expect from you. Think about building experiences that cater to these different comfort levels. Maybe that means highlighting your human support options for Dabblers and Traditionalists while showcasing AI’s personalization power to Adopters. It’s about meeting your customers where they are on their AI comfort level.
How do these three AI personas apply across industries? We’ll explore that in the next blog article for software-as-a-service (SaaS), logistics, professional services, and financial services. Don’t miss it!
research: The state of service expectations
We surveyed 4,500 desk workers to find out what they really want from customer service, AI-powered support, and more.
Written by Andrea Lean