Workers in a call center (` Photo/David A. Lieb)
Associated Press
Customer service has already benefited greatly from artificial intelligence, although many of the company’s customers and employees have also been frustrated. AI has had problems, but better AI, better management and better training will enable better customer service at a lower cost.
Previous articles in this AI and Business series have covered general economic topics and some specific sectors. The customer service function ensures the lowest hanging and fattest fruit in the entire orchard.
“Customer representatives who use an AI tool to control their conversations experienced a nearly 14% increase in productivity,” says an academic study. Less experienced agents saw the largest increase, up 35%, while the most experienced agents, on average, saw no improvement.
Today, companies in the United States employ nearly three million customer service representatives with annual salaries exceeding $40,000. A small boost in productivity will pay off here.
According to an article in the Wall Street Journal, one company’s AI was able to reduce average call time by 36 seconds by simply routing calls to the appropriate department. That might not sound like much until you multiply it by thousands of calls. Customers are very sensitive to wasting time on a call; That is, the time spent waiting to reach someone who can help, or the time it takes that person to retrieve relevant information.
Not only can AI answer questions about how to help a customer, but it can also detect the caller’s mood, such as: B. angry, confused, frustrated or happy. Sentiment analysis can be helpful right now and can also be a tool for developing better practices, both through AI and human customer service agents.
Records of past interactions with a customer are valuable. Older AI products are good at turning an audio recording into a transcript, and newer AI is very good at summarizing those transcripts. The summary will be useful for follow-up calls and to improve future support.
Customer service isn’t just about call centers. Practical work can be recorded on video and then summarized by the AI. And videos are now being recorded by some service technicians. I have seen them for my auto repair and HVAC service. A summary can help both the customer and the next technician work on the project.
The use of AI in customer service has created challenges. Some experienced agents believe that the AI’s suggestions are not as good as theirs. And some companies have given conflicting instructions about when to follow AI and when to use human judgment. It is always bad management to give responsibility to a person without having the authority to complete the task. Clarity about the role of people is crucial for success.
Another challenge arises from the tendency of large language models to hallucinate, which is used in technical jargon for making up falsehoods. Unfortunately, the AI seems pretty confident that it’s right, even when it’s completely wrong. Systems should be designed to take into account the consequences of a hallucination. A call routed to the wrong department is no big deal. If money is transferred to the wrong account, it’s a big deal.
The best practices as we currently understand them start with setting up AI to do a good job. In some cases, fine-tuning the model is appropriate. It involves taking the large language model and reestimating its parameters to achieve a specific goal, such as understanding the jargon of an industry or product. It’s a somewhat expensive task, although much cheaper than developing the model from scratch. Cheaper alternatives, such as a tool called LoRA, seem to work almost as well. Prompt engineering is a cheaper alternative that often produces suitable results. For example, an AI can be instructed to always keep its response consistent with the company’s policy book, its operating manuals, and its troubleshooting guides provided in the prompt.
Data security is a threat that must be addressed. By default, large language models can use the information presented in a prompt. For example, if the AI is asked a question that includes a caller’s name and bank balance, that information goes to the AI company. Most specialized systems install walls to ensure the confidentiality of this information. It’s not difficult to set up and companies should ensure their information is protected.
Another best practice for automated calling systems is an easy way to talk to a human, according to sage advice from helpwise.io.
The academic study on call center productivity mentioned above raises two interesting questions. First, will less experienced agents learn the ropes more quickly with the help of AI, or will they not develop their skills the way today’s experienced agents did? Second, will AI improve enough in the coming years to help more experienced agents? There is no doubt that major language models like ChatGPT, Bard and Claude will improve, but it is not certain that they will ever outperform world-class customer service representatives, especially on the most difficult questions.
AI is here to stay. Implementation may prioritize improving customer service or minimizing costs for a given level of service. This is an important management decision.
The number of customer service representatives, currently about three million in the United States, will certainly decline in the coming years. And in most cases, customer service will improve. Although early job cuts can be achieved through normal fluctuation, layoffs are not far off. The best agents are hired to tackle difficult problems, but other people in this sector should also start thinking about career alternatives.
Follow me up Twitter or LinkedIn. Checkout my website.
When I was 16, I decided I wanted to be an economist, but I also started reading my grandmother’s used copies of Forbes. After degrees including a Ph.D. After graduating from Duke and three years as a professor, I found my calling in the business world. I started as a corporate economist (PG&E, Nerco, First Interstate Bank) and then moved into the consulting industry, helping business leaders connect the dots between economics and business decisions. I wrote “Businomics: From the Headlines to Your Bottom Line – How to Profit in Every Economic Cycle” to help business leaders and small business owners understand how the economy affects their businesses. Side trips on this journey include co-authoring a high school economics curriculum called “Thinking Economics” and earning the CFA designation (although I am not an active charter holder). I served four governors on the Oregon Council of Economic Advisors and am currently board chair of the Cascade Policy Institute. My friends and fans love their monthly economic charts, a 60-second scan of the economy. The latest edition is always available at https://www.conerlyconsulting.com/writing/newsletter/. Also note the free subscription link on this page.
Read moreRead less
Comments are closed.