Today’s customers have become much more digital-savvy post-pandemic and want to be self-sufficient. Not only do they want to find information about your business on their own through various digital channels, they also want to solve their issues quickly by themselves without having to call customer support.
To address this customer need, many organizations have turned to chatbots, intelligent virtual assistants (IVA), and other AI-based solutions to provide customers with a DIY path to problem-solving. But few organizations have been able to achieve tangible results with those implementations. Why? Because most bots, IVAs, and similar solutions can’t address users’ questions effectively. According to the recent McKinsey study on this topic, more than 70% of the organizations surveyed have deployed AI chatbots and IVRs, yet only 10% of those companies say they have seen real customer adoption.
What’s the root cause of the failure of the existing solutions? IDC estimates that 80% of all data is unstructured, or free-form, and this is where the answers reside. Without putting data into structured formats, search tools have not worked effectively because they are not designed to effectively interpret human language and process text-based support documents such as product information and policy documents. As a result, when a customer tries these tools to find the answers, this often turns into a futile effort, forcing the customer to call a support agent. If the support agent doesn’t have the right tool to find the answers the customer is seeking, the customer’s experience suffers, and so does the business.
To help customers and agents find the right answers and reduce case escalation, users need a tool that’s great at understanding human language (questions) and processing unstructured text-based data (documents, etc.). A natural language-enabled search can be a promising approach.
Natural-language search is an intelligent search solution that allows individuals to phrase questions using their own words – as if they were speaking to a person – and returns all the relevant answers back to the user instantly. When a customer or a support agent has a question about a specific product or service, she can type the question into a search bar or a chatbot and receive a list of results that point to the exact sentences and passages containing the answers; thus there is no need to sift through the result list or an entire document to find the relevant information. It’s like the customer has a designated customer support representative (or the employee has a personal assistant) standing by to point her to the most precise information she needs.
Natural Language Processing technology has improved rapidly such that search engines now can understand domain-specific, more complex questions – even those with confusing phrases and complex intent. Those advanced search tools also possess the ability to detect the same question phrased in different ways and bring the right answers back regardless of how a question is asked. Such a user-centric design approach to search can yield a number of benefits including better search outcomes, enhanced support efficiency, and reduced case escalations.
Calming a case escalation
NLP-powered search allows customers to state their specific questions related to a product or service in their own words and delivers accurate, contextual answers back to the customer. Alternatively, a search function built upon a conventional keyword-based approach will not be able to deliver the most relevant results back to the customer because it’s not equipped to comprehend all the nuances in human language and process the underlying text it searches against.
For example, try typing the question: “Can Apple Watch receive a PSA alert?” in the search bar on Apple’s support site and see if it gives you the right answer. A public service announcement (PSA) is a message in the public interest disseminated by the media without charge to raise public awareness and change behavior.
When customers can’t find the right answers with legacy search tools, they are relegated to calling a support agent, and most customers (especially younger generations) do not like this experience. Many times, this is where case escalation begins. Subsequently, when a customer calls the support agent, the agent often has to look up the right information, which takes even more time away from resolving the customer’s issue. An NLP-powered search tool can also help customer support agents find the right information in a much more timely manner.
In a worse scenario, when customers cannot find the answers on their own and a support agent is not able to get the right information, sometimes the case has to be escalated to a specialist. This is what everybody wants to avoid at all costs. The right tool, such as NLP-powered search, can understand the customer’s question intuitively, find the right answers immediately, and satisfy the customer quickly. This is the customer experience users expect: type the question, get the relevant answer they are seeking, and move on with their lives feeling satisfied. No more fruitless searches or wasted time waiting to get answers from customer support.
When NLP is used efficiently and serves the company well, the benefit presents itself in a pragmatic business case: reduced call escalation and support tickets result in customer retention, a higher NPS (Net Promoter Score), and customer loyalty. Call escalation results in additional costs to the company due to additional call center staff time and potential customer churn.