The majority of companies think it is important to collect customer feedback, according to my recent research into customer feedback management, and they put that feedback to an average of five uses, the top five of which are to improve customer service (75%), to develop customer experience and interaction processes (54%), to identify agent training needs (54%), to improve products (50%) and to create a customer service strategy (49%).
However, the research shows several flaws in the way feedback is collected and used. As I pointed out in a previous blog post, the majority of companies stick with old ways of collecting feedback. Most use surveys in one form or another, sent out through traditional channels: email, post, the Web. Completion rates are fairly low, and the insights are reduced even further because the process of analyzing the results is largely manual, and thus subjective. Only a minority of innovative companies have automated the process and assess non-survey-based feedback from customers (recorded calls, free-form text such as emails, letters, chat and web scripts, mobile phone text messages and social media) using analytics tools. Such tools allow companies to gain insights from far greater volumes of data and to define rules so that the assessment is consistent across all forms of input.
Using these tools also overcomes another significant weakness in the current process. The current practice of relying on surveys means the majority (52%) of companies select who to survey at random, and few (39%) set up a process to survey customers on a regular basis. This leaves the process open to bias and means companies are potentially making decisions based on the views of a distorted group. For example, it is not uncommon for companies to survey customers who complain, or to select only a few representatives from a single channel of communication (callers to the contact center). We recommend companies instead create a rules-based process that selects a wider representation of customers and determines when surveys are sent out, and that they examine how speech, text and social media analytics tools can extract insight from larger, more diverse forms of feedback.
The research also highlights one further weakness. The main uses companies make of feedback are inward-looking: improving processes, products, service and training. The results show that only 34 percent always respond to customers, 46 percent respond sometimes, 16 percent occasionally, and a small percentage (5%) never respond to customers. For those that do respond, the responses are likely to focus on a single issue and be inconsistent, as responsibility is distributed across different lines of business. Such a process is thus unlikely to pick up issues that were created by another line of business; for example, is not untypical for customer to complain about responses from contact center agents when an issue was created by a bad back-office process such as the wrong product being sent out, bills being wrong, or user guides being inadequate. We therefore recommend that feedback be assessed centrally and actions be decided based on consolidated feedback from as large a sample of customers as possible.
Voice of the customer (VOC) is a hot topic of conversation, although analyzing results from several of my benchmark studies suggests it is not interpreted consistently. For me it needs to be a joined-up process: collect feedback from as large and wide selection of customers as possible, use all forms of feedback, not just surveys, automate the analysis process using tools that can gain insight from unstructured forms of feedback, share the information across all lines of business, and above all let customers know what action you take as a result of their feedback – after all, companies cannot expect customers to complete surveys if they don’t see any action thereafter.
Do you have a VOC initiative? If so please tell us more and collaborate with me on this topic.
Richard J. Snow
VP & Research Director