Companies that find a way to use AI to improve both the customer experience and operational efficiency will hold a serious competitive advantage over the next decade. But too many are getting the balance wrong — and paying the price in customer loyalty.
When AI prioritizes cost-cutting over customer experience
There’s a lot of discussion about which industries and job functions AI will disrupt — and customer service is almost always near the top of that list. But much of the conversation around AI in the contact center hasn’t focused on improving the customer experience. Instead, the emphasis has been on reducing costs. The most common headline? “AI will replace X% of customer service jobs by [insert year].”
That narrative makes sense when you consider how most contact centers operate today. While there are exceptions, many organizations still view customer support as a cost center and don’t enable agents to act like actual humans. With rigid scripts, workflows, and little autonomy, agents often end up behaving like robots. And in that context, replacing “human robots” with cheaper actual bots starts to feel logical.
The problem is, this approach — rapidly adopted over the past year — makes customer service worse. Many companies overestimated the capabilities of their AI solutions, and with fewer human agents available, it’s harder than ever to reach a live person. Tactics like yelling “representative” or pressing “0” no longer work reliably, and chatbots often struggle to escalate issues appropriately.
Self-service tools can absolutely help in the right context. But there will always be edge cases and complex issues where live human support is necessary. Critical soft skills like empathy — which set great support teams apart — can’t yet be replicated by AI.
The good news is that some customer-centric organizations are starting to realize the downside of using AI purely as a cost-cutting tool. Klarna, for example, was widely praised for introducing an AI assistant that could do the work of 700 agents — the future model for AI-powered customer service. But over the last year, their customer service deteriorated enough that they realized the value of human support.
Recently, its CEO acknowledged the importance of balancing cost efficiency with customer experience.
Klarna’s investment in AI isn’t changing. But the company realized the need to balance the cost-saving benefits with those benefits that improve the customer experience.
The most powerful customer support AI use cases are those that make contact centers more efficient (resulting in cost savings) while still providing exceptional support (which keeps customers loyal and leads to long-term growth). Below are a few examples.
Dig deeper: AI adoption in CX is rising, but implementation challenges remain
Understanding customer personas and tailoring support accordingly
Customers vary widely in how they prefer to engage with support. Some never want to speak to a human and are happy using self-service tools. Others need a lot of hand-holding and always want to talk to a live agent. Then, there are those who are comfortable starting with a bot or FAQ, but want immediate escalation to a human when things get urgent or complex. The graphic below shows a few sample customer personas for customer support.
A customer-focused application of AI would be to assign customers to one of these personas based on all of their past interactions with support. If you know what kind of persona a customer is, you can get them the right level of support.
For example:
Danny can be routed to self-solve channels and virtual agents.
Julie and Gary can be routed to live agents, with Julie getting sent to the most experienced agents given the urgency of her issues.
Carla can be started off in self-solve, but given the option to easily transfer to a live agent if needed.
This approach still provides some of the cost-saving benefits of AI support, but it ensures the cheaper self-solve options are being provided to the right people and that all customers are getting their preferred type of service.
Dig deeper: Why relying on AI won’t improve the customer experience
Scaling closed-loop service recovery programs
Sometimes a customer will have an unresolved issue even after contacting support or will generally be unhappy about their experience. Feedback about their experience is often provided through a post-contact survey, social media or other means. Customer-centric organizations make sure that they close the loop and follow up with customers that provide such feedback. However, this can be challenging to scale due to resource and time constraints.
A customer-focused application of AI would be to automate customer follow-ups to reach more customers. AI tools can flag customers that are at risk of churn based on the sentiment of their feedback and past behavior, and then auto-generate a message that acknowledges their feedback. The auto-generated message can then provide a phone number for the customer to contact if they still need additional support.
Most customers will be satisfied that their feedback is acknowledged, and those who want to still speak to someone will have that option. The end result is that the customer experience is improved in an efficient way due to the use of AI.
Dig deeper: 6 steps to help improve your customer experience with AI
With the vast amount of data that companies have, AI solutions can be easily deployed to analyze the types of bad experiences that lead to customer service contacts. This knowledge can then be used to automate efforts to resolve a bad experience before a customer has to contact support.
A common example is with airlines when there are flight delays. If a high-value customer misses a connection and has to spend extended time on a layover, the airline can automatically send a meal or hotel voucher to that customer as a proactive way to resolve a pain point. It’s a goodwill gesture that is both good for the customer (by resolving an issue) and the company (by preventing a future contact and freeing up support resources to handle other issues).
Moving forward
There are many more AI use cases that benefit both customers and companies beyond these examples. And with how fast technology is evolving, many more new use cases will develop. It will be tempting for many executives to focus solely on the use cases that deliver short-term cost savings wins. But companies that recognize the value of human support — and invest in AI solutions that help people work more efficiently rather than replace them — will gain a long-term competitive advantage.
Dig deeper: 5 areas where businesses need to improve their customer experienc
Contributing authors are invited to create content for MarTech and are chosen for their expertise and contribution to the martech community. Our contributors work under the oversight of the editorial staff and contributions are checked for quality and relevance to our readers. The opinions they express are their own.