Conquering the steepest challenges in customer experience

The customer who waits on hold for 40 minutes, only to be transferred two more times. The shopper who hits a dead-end with a chatbot, with no escape hatch to a human agent. The patient who explains her billing issues to three different support reps.

These aren't just annoyingly common customer experiences—often they are relationship killers. And despite new, AI-powered solutions that can prevent many of these scenarios, they continue to plague even large, well-run customer service operations.

According to Freshworks research, top-performing customer support teams can resolve most queries in under two minutes with 97% first-contact resolution. Yet for most organizations, stubborn challenges like reopened tickets, fragmented systems, and poor-quality self-service continue to erode customer loyalty and team morale.

Bestselling author and customer-service researcher Shep Hyken agrees: “It takes a combination of digital experience and human experience to create the best experience,” he says. 

Many companies are looking beyond chatbots to next-generation AI agents to relieve these pain points. But success depends on first understanding what makes these experiences so devilishly difficult to solve—and why technology alone isn’t enough.

Here’s a look at a handful of stubbornly poor customer experiences that challenge service teams: what makes them so complex, and how agentic AI and other smart tactics can help make them a thing of the past.

1. Endless wait times 

Whether it’s sitting on hold, waiting days for an email reply, or watching a chat message go unanswered, slow response time remains the top customer complaint in 2025, according to a Vonage study.

Although AI and automation tools are already showing they can significantly reduce wait times, many companies haven’t adapted their workflows to best leverage them. Hyken believes they are stuck using legacy support solutions such as outmoded ITSM platforms that fail to prioritize and track tickets—or no centralized support platform at all. This leads to resolution times up to 36 hours or longer, according to Freshworks research. 

But when they put AI to work, they get results: Freshworks research has found that when support teams work alongside Freddy AI Copilot, they see a 43% increase in first response time and 35% improvement in resolution times. 

2. Having to explain the problem over and over

If a retail customer has fired off an email asking for help with an overcharge and gotten no response, eventually they pick up the phone. But the agent fielding their frustrated call can’t access the original email—and after escalating the issue to the finance team for a refund, the hoarse-voiced customer may be ready to jump ship from future purchases.

“We hate repeating ourselves,” says Hyken. “Don’t make me start over again, make it omnichannel. Make it seamless.” 

When companies take an omnichannel approach, they allow customers to contact them over a range of channels and integrate those conversations so that customers have the same experience whether they call, chat, or email.

This only works if these systems are connected, yet Forrester has found that many large companies use as many as 25 to 49 disconnected customer data systems across their organization. And Freshworks research shows that 62% of companies rely on more than one channel (phone, email, chat, and social media) to provide customer support.

3. Getting an unhelpful chatbot to get human help

At a certain point in a chat or self-service interaction, frustrated customers may want to talk to a human agent, but find themselves trapped in an endless loop of suggestions, capped with a mind-numbing one: “Does that answer your question?”

Freshworks research has found that first-contact resolution rates improve to 98% when chatbots handle simple issues. But when chatbots are overused or lack escalation triggers, customer experience quickly deteriorates.

Today’s chatbots remain limited in the complexity of support issues they can resolve, topping out between 70% and 80% of routine queries, according to IBM. Often the human handoff remains elusive thanks to too few agents, high call volume, or the limited hours of available human support. 

And customers may not be able to chat with a human for one simple reason: It’s more expensive. The cost of an average human agent interaction with a customer is $1, while a chatbot interaction costs a mere 12 cents (and an IVR interaction 6 cents), according to Waterfield Tech research. 

But the cost of a bad chatbot interaction is much higher, according to Forrester: 73% of customers cancel a purchase and 30% look for an alternate brand.

Freshworks research notes that the best-performing companies rely on a two-party strategy: AI agents to deflect basic queries, and Copilot tools to support agents in handling complex ones. This hybrid model ensures automation is used to enhance, not replace, human support.

4. Unable to find accurate, up-to-date information when you need it

Whether it’s tracking the status of a delivery or troubleshooting a technical problem, most customers (73%) turn to a company’s self-service support, whether it’s a knowledge base or FAQ, according to a Gartner survey.

But it’s a moment rife with pitfalls: Too often, building a self-service portal is often a one-and-done effort, without continuous monitoring or tracking customer feedback to improve the system. Gartner found that, as a result, only 14% of customers report their issue had been fully resolved by self-service.

Manufacturing companies that keep their self-service portal updated show its value; they see CSAT scores of 100%, first-contact resolution near 98%, and resolution times under three minutes, according to Freshworks. Tools like Freddy Copilot support those outcomes by automatically surfacing relevant content, suggesting updates, and even drafting new solution articles.

Agentic AI steps in  

Technology is part of the answer to these problems, but Hyken says too many companies are reluctant to act.

“Many companies are hanging on to legacy programs that they’ve purchased as recently as just a few years ago, thinking they need to reap their investment before they make a new one,” notes Hyken. “But the irony is that tools from just as few years ago were far more expensive than they are today, so it might be cheaper to throw out the bad and start the new.”

Read also: 5 frustrating employee experiences that demand better solutions

Freshworks analysis of companies using AI agents found that they help customers get answers quicker: 83% of these companies are seeing faster resolution times for common support issues. Another key benefit is consistency: AI agents can provide the same support whether via chatbot on an app or website, DMs via social, or a fast reply over WhatsApp.

The promise of agentic AI goes well beyond speed and consistency. Unlike today’s chatbots that can resolve only basic issues and must escalate the more complicated ones to human agents during call center hours, AI agents could proactively complete more complex tasks at any hour, or preserve conversations that could be picked up days later. And experts expect AI agents to pick up frustration signals more quickly and escalate to a human agent before customers even have to request it.

Being a customer is an experience rife with inconsistencies from one company to the next and even within the same company. With a clear understanding of the issues, companies can bring AI to help solve them, and introduce efficiencies and increased empathy, too.