Skip to main content

AI drives new generation of customer experience surveys

(Image credit: Image source: Shutterstock/Shutter_M)

Are brands interacting with their customers in the right way to get useful feedback? The answer is: no, not always.  While surveys are often used as the barometer of customer sentiment, many companies face two key challenges: 

  • How to encourage a healthy survey response rate that allows them to accurately interpret the customer journey. 
  • How to uncover the root-cause of customer issues.   

Given that a recent Forbes interview with a survey expert indicates that response rates are down to an all-time low of less than 2%, brands could be forgiven for thinking their customers don’t want to talk to them. However, the rise in the number of people using social media to contact brands would suggest otherwise. It is simply that they have survey fatigue, and the flaws in many survey programs are putting customers off from the start.    

Too many surveys are long and complicated. Research from Stanford University shows that the quality of data deteriorates as respondents spend more time in a survey. And if this wasn’t bad enough, traditional surveys don’t necessarily produce the results that brands are looking for. This is because they cover only the topics that brands think to ask about, collect data only from the people who choose to respond, and rely on customers to have accurate memories about what happened, in some cases a long time after the experience. 

Because of these shortcomings, surveys cannot provide all of the deep customer insights that lead to action. In addition, if the response rate is too low, the survey results can misrepresent the real feelings of the larger customer base, which could lead brands to jump to the wrong conclusions. 

They need a way to use the survey results they do receive, interpret the results correctly, and even more importantly, ensure that after canvassing customer views, these can be interpreted into insight that can be acted upon.    

To help them, brands are turning to technology that will support company-wide action and improve the customer experience (CX). One advance that is beginning to make a positive mark on the incidence of survey fatigue is conversational surveys powered by artificial intelligence.    

To understand how this works, consider the difficulty of constructing a survey in the first place. The questions can aim to anticipate certain responses, but without being open-ended there are bound to be gaps where a customer’s experience with a brand is not explored.   

A conversational survey, as the name suggests, engages the customer in real-time. As the customer provides comments or text-based feedback to a question, a conversational survey will facilitate a two-way responsive interaction. This encourages customers to be more detailed and honest about their experience. This method is already showing strong signs of generating effective user response.   

Take a retail brand as an example. While purchasing an item in store, a customer agrees to participate in a short survey that can be delivered to them on a device of their choosing. The first question is simple: 

‘Did you get everything you needed from your visit into the store today?’ As the customer types their response into the comment box, the chatbot that sits behind the scenes of the conversational survey tool will read and analyse the words and can pick up context as well as both negative and positive sentiment. So, if the customer says: ‘I had difficulty locating the item I wanted’, the system might respond with: ‘Store layout and signage seem to be an issue, what ideas do you have for better accessibility?’. The millennial customer may want a directory on their phone and baby boomers may want more signs, depending on the collective answers to this question, the retail brand will know exactly where to focus its energy and IT investments.

The same process can be applied to multiple customer-centric sectors from hotels and restaurants to entertainment and leisure venues. In each case the conversational survey provides specific and unique feedback to the customer’s experience.     

To enable the process, conversational surveys use a combination of artificial intelligence, natural language processing and machine learning, and combine this with a brand’s own vocabulary to create natural, real-time communication. 

How this differs from the traditional survey methodology is that it makes very few assumptions about the customer’s response in advance. It takes a simple approach with just one or two vital structured questions combined with an open-ended question that allows customers to explain what matters to them the most. It is important to remember that if the research takes longer than around three minutes, customers are likely to start bailing out, so easy to understand, clear questions are essential.   

An open-ended question can also be used to elicit feedback on many topics without forcing the customer into a narrow lane. Brands can consider a question like, “Please let us know why you rated your experience with the score you provided?” to tease out the most information possible. It only takes a few scored questions and an open-ended question, combined with text analytics and correlation analysis, to tease out all the critical drivers of satisfaction, loyalty, profitability, and competitive differentiation.       

In addition, the questions that the customer sees can change based on their previous responses. For example, if the customer is dissatisfied about pricing, the brand can ask follow-up questions about that topic. “Smart branching,” where a follow-up set of questions can be asked in real time based on the consumer’s answer to the open-ended question, can demonstrate to the consumer that the company is actually listening to the feedback and not just looking for scores. 

Conversational surveys have the advantage of appearing very simple because just one question opens the conversation, but they still deliver deep insight into the issues that affect the customer. Customer experience managers will benefit by being able to understand what is driving their customer satisfaction scores, such as their Net Promoter Scores (NPS).    

The technology that sits behind conversational surveys is ground-breaking. Whilst chatbots cannot currently act as a replacement for human interaction, in many areas of customer service, when it comes to encouraging a response as part of a survey, they are increasingly invaluable. As the decision tree of responses that chatbots can manage increases, conversational surveys can become more complex and dive into the root-cause of friction across the customer journey. But even today, they can understand context, detect sentiment, highlight problems and foster an environment that encourages customers to more deeply engage with a brand.    

Susan Ganeshan, Chief Marketing Officer, Clarabridge 

Image Credit: Shutter_M / Shutterstock

Susan Ganeshan
Susan Ganeshan is the Chief Marketing Officer at Clarabridge. Over her 25-year career in services and software companies, Susan has led both product management and product marketing functions, and has developed a maniacal focus on customer success.