Original article was published on artificial intelligence
Over the last decade, experts have predicted that AI would make up the bulk of the current workforce. The World Economic Forum estimated that by 2025, humans would only make up 48% of the workforce, whereas digital workers would account for 52%. Although Accenture’s model showed that the risk of not adopting AI would be much greater than the risk of early adoption, it was based on an upbeat economic climate. With good returns on equity, companies have had little reason to galvanize radical change. Why fix something if it isn’t broken?
This has all changed, however. COVID-19 broke things. Economies stalled. Consumer spending dried up. Revenues tanked. Enterprises have started looking at new ways of doing things to simply survive the current situation. In this market, employing AI-powered digital colleagues doesn’t sound nearly as scary when enterprises need radical change to get out of this valley. This is particularly true when looking at deploying AI for customer service roles.
Customer Service Shortage
Even before this pandemic, enterprises were well versed in the difficulty of sustaining good customer service with seasonal variances in load. The demand for customer service has spikes, and the humans servicing calls can struggle to meet demand. Maintaining a good net promoter score (NPS), which is used to gauge the loyalty of a customer, is difficult even in “normal” times. When our world changed with the shelter-in-place orders, businesses were typically faced with two scenarios. The first was that they had to shut down or dramatically change how business was done — meaning they had to field inquiries around how this impacted their customers. Alternatively, many businesses saw a massive spike in the number of customer inquiries as their services were of even more importance due to the outbreak.
Banks, for example, fell into this latter category. During nearly any crisis, banks see an increase in the amount of customer management required, and the current pandemic has been no exception. The customer service departments at banks were flooded with questions and requests as their customers prepared to face uncertain financial times. Banks weren’t the only business to be flooded by customers, however. Travel agencies and airlines suddenly had to field thousands of calls to change travel dates, ask about refunds and so on. In fact, one customer was unable to speak to a service rep even after waiting for 15 hours and 40 minutes.
The current approach to customer service simply isn’t elastic enough to deal with the unexpected peaks and valleys that are baked into the modern world. Good and sustainable NPS scores require that a business is able to scale the number of servicing agents in proportion to the load; otherwise, your business falters when your customers need you the most. The only practical way to provide this level of flexibility is by activating digital colleagues. This is true during the current crisis — and for future ones we’ll inevitably face.
AI-powered employees provide businesses with the option to scale their customer service reps as needed, and these AI-powered reps don’t burn out from having to field one difficult call after another. There have been numerous studies on the psychological toll that dealing with irate or panicked customers has on individuals, and that is likely even more salient during a time when everyone is dealing with extra stress and anxiety. An AI colleague, however, is immune to this stress, which protects humans from having to face it directly while knowing that the job is still getting done.
Because AI is not impacted by these factors, customers won’t see a drop in their level of customer service due to stress. Conversational AI offers an unprecedented level of natural language understanding. With more advanced AI technologies being able to account for context in a conversation, modern conversational AI is able to go past providing answers to frequently asked questions and instead have an interactive conversation that helps the customer actually solve their problem. Additionally, the technology can show empathy by analyzing the emotional state of the user and responding accordingly. Up until now, this level of intelligence was unheard of from a machine, but AI has come a long way.
The simple truth is that an AI colleague is, in many ways, better equipped to be a customer service agent, and that is a good thing for people. As organizations look to reorder and reengage their workforces in the wake of a recovery, they should be looking to move people out of customer service support roles into jobs where they can use their unique human skills like critical thinking and creativity. Furthermore, digital colleagues can handle the volume that a crisis can instigate. No one wants to wait for 15 hours just to talk to a service rep. Digital colleagues can be scaled to move that wait time down to mere minutes.
Before getting started, enterprises must take a closer look at problems that actually need to be solved. It would be a mistake to invest in an AI solution just for the sake of jumping on a popular bandwagon. Doing so will inevitably lead to an unnecessary rise in expenses that will diminish the firm’s working capital while allowing unaddressed problems to persist.
Once your main problems have been identified, enterprises can start to think about the ways that AI can help resolve them. But they must realize that AI is not movie magic — it cannot do everything. If an AI provider claims that it can, that’s a strong red flag to avoid that particular solution. That’s also why it is imperative that businesses carefully evaluate their AI options and choose solutiona that work for them and their predetermined problems — not new problems a provider suggests they address.