A customer explains an issue to a chatbot, follows several prompts, and eventually reaches a person. The transfer works, but the agent asks them to explain everything again.<\/p>\n
Nothing has technically broken. Each system may have done what it was designed to do. But from the customer’s point of view, they are still doing more work than they should.<\/p>\n
That matters as more organizations invest in AI and automation to improve service. In the latest market study with CMP Research, 49% of customers say they do not see AI actively improving their experience, while only 45% believe their customer service experiences have improved over the past year.<\/p>\n
The question, then, is not simply whether organizations are becoming more efficient. It is whether customers are actually experiencing better service.<\/p>\n
Contact centers need operational measures. Average handle time, containment and cost per contact help leaders understand capacity, performance and cost.<\/p>\n
The problem starts when those measures become the definition of success.<\/p>\n
\u201cThe problem is that efficiency and customer experience are not necessarily the same thing,\u201d <\/em>says Chris Karp, Chief Growth Officer<\/strong> at Afiniti.<\/strong> \u201cA shorter interaction is not inherently better if the customer\u2019s issue remains unresolved, just as an AI interaction is not successful simply because it avoided reaching a human agent.\u201d<\/em><\/p>\n A contained conversation tells you that the customer did not reach a person. It does not necessarily tell you whether they got what they needed. A shorter call tells you how long the interaction lasted, but not whether the underlying issue was resolved.<\/p>\n The live market study makes a similar distinction. AHT, containment rate and cost per contact remain useful measures of operational efficiency, but they mean little when viewed alone. FCR, along with other day-to-day KPIs, can provide a clearer view of whether the operation is delivering the right outcome.<\/p>\n The measures in the first column still matter. They help explain how efficiently service is being delivered. But they need to sit alongside a clearer understanding of what happened to the customer and the business.<\/p>\n Customers do not see separate chat, voice, CRM, routing and workforce systems. They experience one company.<\/p>\n When those systems do not carry the right context between them, the customer can end up doing the work of connecting them. They repeat information, explain what has already happened and work out how to reach the right person.<\/p>\n The market study found that 36% of customers are frequently required to repeat information during customer service interactions. The report identifies this repetition across one or more channels as a consistent customer pain point.<\/p>\n The challenge is not that the technology behind the experience is becoming more complex. That complexity is increasingly unavoidable. The challenge is making sure the customer does not have to understand or manage it.<\/p>\n As Chris Karp puts it in the study, the best experiences combine AI’s speed and scalability with human judgment and empathy without making the customer navigate the boundary between them.<\/p>\n AI has an increasingly important role in customer service. The market study expects self-service to account for 77% of contacts beginning the journey by 2030, compared with 56% today. Leaders also expect containment to rise from 56% to 74%.<\/p>\n But the report is clear that this does not mean every interaction should simply be automated away.<\/p>\n Customers still value speed, empathy, personalization, customized resolutions and accessible support. And 70% of customers still prefer live service.<\/p>\n AI can handle an increasingly broad range of interactions. The more useful distinction is not simply between simple and complex work. It is between situations AI can handle effectively and situations where human judgment, experience or empathy can add more value.<\/p>\n A handoff to a person should not automatically be treated as a failure. The important question is whether the customer reaches the right outcome, and whether the transition happens with the right context.<\/p>\n For many large organizations, customer service already runs across multiple platforms, data sources, automation tools, AI models and workforce technologies.<\/p>\n Replacing everything with one system is rarely a realistic starting point.<\/p>\n The live market study describes Afiniti’s approach as connecting the different layers of the contact center without requiring organizations to replace the technology they already have. The opportunity is to make the existing ecosystem work together more effectively around customer and business outcomes.<\/p>\n That means connecting decisions across customer interactions, automation, intelligence and operational conditions. It also means reducing the inefficiencies that appear between systems when they operate independently.<\/p>\n The technology behind customer service may continue to become more complex. The experience presented to the customer should become simpler.<\/strong><\/p>\n The next time a service metric improves, it is worth asking a few additional questions:<\/p>\n Efficiency remains important. But it should help deliver the right outcome, not replace it.<\/p>\n The full Afiniti and CMP Research market study<\/strong> examines what customers now expect from modern service, how organizations are preparing for a more AI-driven customer journey, and where AI and human support each have a role.<\/p>\nWhat the metric tells you, and what it doesn’t<\/h3>\n
\n\n
\n The contact center records<\/strong><\/th>\n The customer may experience<\/strong><\/th>\n The question still unanswered<\/strong><\/th>\n<\/tr>\n<\/thead>\n\n \n A shorter interaction<\/td>\n The issue may still require another interaction<\/td>\n Did the customer get the help they needed?<\/td>\n<\/tr>\n \n A contained AI interaction<\/td>\n The customer may still need additional support<\/td>\n Did automation provide the right path?<\/td>\n<\/tr>\n \n A completed transfer<\/td>\n The customer may have to repeat information<\/td>\n Did their context move with them?<\/td>\n<\/tr>\n \n A lower cost per contact<\/td>\n The overall journey may still involve more effort<\/td>\n Did the efficiency gain support the right outcome?<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n Customers experience one journey<\/h3>\n
The goal is not to keep every interaction automated<\/h3>\n
Modernization does not have to mean starting again<\/h3>\n
A more useful test of progress<\/h3>\n
\n