ARTIFICIAL INTELLIGENCE: ADVANCING CUSTOMER SURVEYS
Today, more and more successful organizations are letting customer needs direct their business strategies. Guaranteed success comes from listening to and understanding customers well. It’s a simple concept, yet difficult to execute.
A Walker study found that by the end of 2020, customer experience will overtake price and product as the key brand differentiator.
86% of buyers are willing to pay more for a great customer experience
Customers are willing to pay a price premium of up to 13% (and as high as 8%) for luxury and indulgence services if they receive a great customer experience
88% of companies now prioritize customer experience in contact centers
With this in mind, it’s key to analyze customer feedback and surveys to improve an organization’s customer experience,
The Survey Challenge
The average business hears from less than 4% of its dissatisfied customers. For every customer who voices a complaint, there are 26 other unhappy customers staying silent. However, 95% of customers openly share their bad experience with others.
Gaining in-depth knowledge about customers takes effort. You must collect and analyze customer data (direct customer feedback and customer interaction data) with speed and precision. While customer surveys have historically driven business strategies in the service industry, their limited data points are leading to an unwarranted bias. Deriving meaningful analytics and actionable insights to improve CSAT or NPS from these surveys remains a challenge for these reasons:
- LOW RESPONSE RATE – Typically, survey response rates vary between 5-30%. A low response rate can distort the statistics of collected data and, in turn, dilute the results. No longer indicative of the general population, low response-rate surveys offen have issues with survey design, length, demographics, incentives, lack of personalization, etc.
- QUESTION SCORES LACK INSIGHT - Insufficient data points or poorly designed surveys with open ended questions can’t identify the reasons for a customer’s feedback. This makes it difficult to pinpoint actionable areas for improvement.
- BIASED OR INCORRECT INTERPRETATIONS - Generally, minimum samples are required to determine significance. Low response rates impede the ability to conduct a meaningful statistical analysis. This, in turn, can result in a bias that leads to incorrect business decisions. Customer surveys should be more than a business initiative that simply checks a box. They’re a powerful tool for improving customer experience and, ultimately, growing your customer base.
The Future of Customer Surveys
It can take up to 12 positive experiences to make up for one unresolved negative experience. If you resolve a complaint in the customer’s favor, they’ll do business with you again 70% of the time.
Artificial Intelligence (AI) led customer surveys are the new way to tackle this challenge. An AI engine that shadows and predicts customer behavior can be the new avatar of CSAT/NPS surveys.
Here’s how to apply AI led surveys for meaningful results:
UNDERSTAND AND MAP INTERACTIONS AND DATA POINTS
Tag all interaction touch-points, sentiments, expectations, intents, entities, utterances, etc. These capture points will inform a business’s design and give a holistic view of customer interactions.
PROPERLY DESIGN INTERACTION / CUSTOMER EXPERIENCE RATINGS
Look beyond existing survey templates, which can be short and incomprehensive, to entice customers to respond. When you apply a NLP/ML-based automated survey analysis, you capture vital data points that provide deeper insights. AI allows you to identify and rate interest levels, sentiments, engagement, moments of truth, etc. that are relevant to your business, all with measurement and rating scales.
PROVIDE HISTORICAL INTERACTION DATA FOR THE AI ENGINE TO LEARN AND EVOLVE
Call recordings, chat transcripts, etc. can serve as the foundation. In most cases, three month’s data is a good starting point.
COMPARE CUSTOMER SURVEY RESULTS WITH AI RESULTS TO FINE TUNE
Start by applying the AI prediction engine on historical interactions, then compare to ensure the engine is accurate. Use anomalies to train and coach the engine for higher accuracy.
APPLY REAL-TIME AI RESULTS AND INSIGHTS
Various NLP techniques, combined with some of the latest transcription technologies, can apply these models in real-time. This applies to voice/telephone interactions as well. This can be insightful for both the customer as well as the agent.
Insights gained from artificial intelligence can improve customer service in two meaningful ways. First, when AI results are shared with customers up front, it allows for endorsement and feedback. Not only does this garner goodwill from the customer, it provides personalized coaching inputs for the AI model as well.
The sophisticated engine can meaningfully understand various intents, entities and sentiments, and can automatically provide actionable insights and key business decision inputs. When extended across 100% of interactions, it gives a true picture of customer insights. For the best results, AI survey predictions should be shared with the customer on a near real-time basis, while the interaction is still fresh in their minds.
Second, AI enabled insights can be used internally for QA, Data and Analytics and Marketing teams to improve customer experience, retention and growth. They can identify interests, sentiments and behaviors that offen fail to get the attention they deserve.
AI survey insights can also provide agents with feedback and improvement opportunities for interacting with customers. The engine can help coach them towards a much better customer experience.
Organizations with elevated levels of technology adoption and progressive attitudes towards change can use AI enabled customer surveys to redesign their customer strategy to:
- identify moments of truth and opportunities to attract promoters
- convert passive customers into promoters
- detect detractors earlier
- help promoters promote more
A conservative path would start with an internal deployment of AI customer surveys. A customer-facing strategy could be initiated once confidence and success levels are proven. That being said, organizations seeking to reap exponential benefits would be wise to embrace AI enabled customer surveys from the start.
Global Lead for Healthcare Digital Transformation, CX & Solutioning