British companies have come a long way with utilizing their data to identify customer needs and predict their future behavior and estimate future profitability (the so-called customer lifetime value). Realizing the customer lifetime value is critical for a company so they can focus on the strategic initiatives that contribute most to the company’s future earnings and growth. This is my assessment based on two major Customer Experience (CX) events I attended in London in the end of 2018. However, there is one area in which Denmark is still at the forefront of CX— closing the operational customer feedback loop.
Anglo-Saxon Companies are Ahead on CX Strategy
At the European Conference, CXForum 2018, held by MaritzCX and the UK CX Awards in 2018, there was a fairly clear picture. The largest companies in the UK, US and New Zealand are better at working strategically to improve their customers’ experiences. There is one area where they differ positively compared to most Danish companies—using advanced analytics for exploiting big data to make better strategic decisions.
Advanced Customer Analysis
In the Anglo-Saxon business world, large amounts of data (Big Data), Artificial Intelligence (AI) and Machine Learning have come a long way in identifying the strategic efforts that most effectively improve customer experiences in a way that increases both customer profitability and loyalty. This allows one not only to make a very precise predictive analysis, but also a so-called prescriptive analysis. While predictive helps to predict each customer’s future behavior, a prescriptive analysis shows what the company should do for each customer, even if you do not have any feedback from that customer.
There are some benefits to combining customer feedback data (e.g. from current NPS measurements across key customer touchpoints) with the internal operational data:
- You get individual insights about 100 percent of your customers based on an analysis of the feedback given from some of the customers
- Predictions can even be achieved earlier
- You do not often have to collect customer feedback
Achieve the Utopian “Response Rate” of 100 Percent
If you analyze the unique patterns in the operational data of detractor, passive or promoter customers, you can achieve response rates of only 5-50 percent and still be able to calculate the expected Net Promoter Score (NPS) for each of the remaining 50-95 percent of customers who have not provided any responses. This allows companies to act proactively to all individual customers and close the customer feedback loop even for those customers who have not directly given their own feedback to a survey.
In my experience with CX projects, the implementation of individual customer follow-up is one of the lowest hanging fruits of all. Normally, you cannot make any operational closed customer feedback loops to the non-respondents. However, if you have a response rate of 20 percent and can estimate the NPS of the remaining 80 percent, it gives a ROI that is five times as big!
Finding the Right Data Sources
Some may think that their business does not have all the data required to make such advanced analysis. Fortunately, this is rarely the case. For most businesses, there is a lot of relevant data, not found in their CRM systems, but still attributable to individual customers and thus can be used for data mining. However, some data is so sensitive that you must be aware of compliance with GDPR. As far as data mining of statistical data is concerned, one can easily make an “anonymous” extract to use for the analysis. As long as you have a unique customer ID, you will subsequently be able to transfer the individual results per customer to the company’s CRM system and thus apply in the CX effort.
There are many sources of data you can utilize. Among many others you can use web services, GPS location, point of sales data, customer requests, IoT devices, external databases with accurate weather knowledge, traffic conditions, details of customers’ telephone usage (for telecom companies), financial transactions (for banks), web clickstreams, smartphone apps, online shopping carts, RFID, text messages, chat, phone calls with the company, payments and social media.
Additionally, traditional information already exists in the CRM system such as demographic and company graphical information, customer size, purchase history, credit rating, NPS scores and the related open comments from customers. Companies may even use the notes the employees have written into CRM based on the service recoveries and root-cause analyses from previous follow-ups. So, it’s “just” a question of collecting data and getting the company’s most talented analysts to do a lot of hard work.
Making Minority Languages Available for Text Analytics
A number of the leading Danish companies in the CX arena have experienced that the IT tools that currently exist for text analysis of the Danish language are not so accurate, as they are based on the English language. Yet, there are a couple of Nordic providers who have built up accurate algorithms because they can use Danish text directly instead of first making a “Google translate” to English. Major suppliers of text analytics are starting to reach minority languages such as Danish. Utilizing machine learning, these suppliers’ algorithms are becoming more precise and improving exponentially every day.
When to Collect Data
As most of the access to data is available in real-time, companies also can act as soon as these trigger an “alarm” because text analytics has identified a pattern indicating that there is either a risk of the customer leaving the company, reducing their engagement, or speaking negatively about the company. Thus, you do not have to wait several months for the specific customer’s feedback in a questionnaire—assuming you are lucky enough that the specific customer was among the fraction of customers who will actually participated in your survey.
One need not ask customers for feedback as often as most believe. Personally, I am tired of invitations to irrelevant surveys of irrelevant interactions with some marginal supplier. As long as there is water, gas in the pipe, phone coverage and connection to the network so to speak, most consumers do not engage in frequent relations with providers.
Following compliance with GDPR is the first rule for collecting customer feedback. The second most important thing to consider when making customer metrics is the rule that data collection itself should help to create a better customer experience rather than reduce it through undue disturbances of customers in their busy lives.
There is Hope for the Danish yet!
While Danish companies could seem lacking when compared to Anglo- Saxon companies regarding the use of big data to improve customer experiences, it looks completely different when looking at the ability to involve the front-line in closing the operational loop by conducting service recoveries and root-cause interviews.
So yes, there is still hope for the Danish companies!