scene in an airport
Industries such as airlines collect huge amounts of data on their customers, but do little with it to improve the customer experience PHOTO: chuttersnap on unsplash

In recent years, companies have invested heavily in customer service functions, adding stacks of new technologies to better understand just how their audiences tick. They’ve beefed up data collections and put advanced analytics, machine learning and other smart technologies to work trying to model customer behavior to help them do a better job solving customer problems proactively.

So far, it’s not working — at least not to the extent people expected it to. Customer service agents still can’t successfully answer consumers’ questions more than half the time, according to Help Scout. And Salesforce.com reports that bad customer service costs U.S. businesses $84 billion a year and that 78 percent of customers say they have ended a business relationship because of bad customer service.

Technologies are improving at breakneck speeds and disrupting the way we do business. But organizations are struggling to implement them effectively enough to make a tangible difference in their customer service functions.

What is holding organizations back? And what can be done about it?

Data Pitfalls Holding Teams Back

Teams are having trouble turning data into actionable insight on a number of fronts.

  • They’re struggling with bad data: It’s hard to get analytics systems to deliver business value when the data you’re correlating it with is incomplete, inaccurate, stale or siloed.
  • They can’t organize data in the right buckets: Many customer service teams are trying to apply new analytics to legacy data systems that have been reconfigured multiple times. Customer information systems need to be intuitive and useful. Too often, today’s systems are a hodgepodge of structured, semistructured and unstructured data sources, with confusing naming conventions and inadequate governance models.
  • They’re not buying in: Customer service teams have been resistant to change, at the leadership level and on the ground. If users don’t see the benefits a new system provides, they’re going to find their own ways to solve problems.
  • They’re confused about the overall strategy: The leadership team has to demonstrate its support of analytics programs. If management doesn’t support it, the team won’t either.

Putting Data to Use for Friendlier Skies

Air travel is an example of an industry that isn’t doing as much as it could on the customer service front. Airlines have plenty of data, but they could do a better job putting that data to use for the benefit of customers. The data that many carriers gather is still holed up in silos that don’t communicate with one another, limiting the airlines’ ability to create positive customer connections.

Leveraging existing passenger data could help turn frustrated travelers into happy customers who say nice things about an airline. Instead of charging baggage fees, change fees and airport check-in penalties, airlines could probably generate more revenue — and improve their images — by investing in efforts to improve the customer experience.

Data and Analytics Can Help

Still, there are cases where data and analytics are helping organizations improve their relationships with their customers.

A Fortune 50 software company increased its call center’s customer satisfaction scores and reduced dissatisfied calls by a quarter using software that routes calls to the best available agent based on a personality match. The software, from Mattersight, uses cloud-based analytics and behavioral models to determine if a caller is outgoing, sarcastic, serious or shy.

Rental car company Hertz gathers extensive information on its customers from emails, text messages and web surveys. In one case, a Philadelphia office adjusted staffing levels based on customer usage patterns. By making sure a manager was on hand during peak activity hours, Hertz resolved customer questions and complaints more quickly and effectively, improving each customer’s experience.

What can organizations do to better integrate analytics to improve customer service? Here are a few suggestions:

  • Make sure you can trust your data: To run analytics, you need to set up systems that make sure data is complete, accurate, timely, relevant and consistent, or business users won’t trust it. They will stop using it, even if the organization has invested millions of dollars in data-centric tools and technologies. Organizations need to invest in the right kinds of data and treat data as a corporate asset. In other words, data must be the new currency for enabling digital transformations.
  • Predict what customers are going to do: Using data mining methods that “learn” from the collective experience your company has recorded in its sales transactions, customer interactions and product usage analytics, the team can create a predictive model for customer behavior. The model applies what has been learned to produce a predictive score for each new customer, in real time. With this type of functionality in place, companies can predict which new customers are likely to renew and which are probably going to move on.
  • Be customer-obsessed: In a subscription economy, companies need to focus on the customer more than ever. Driving product adoption to enable business outcomes is critical. Use an analytical platform powered by machine learning that can learn from every interaction with the customer, be it with the product or the company itself. One approach can be to create an adoption index generated using a supervised and semisupervised machine learning model. This is a weighted moving average with declining weights based on the recency of product usage and customer service interactions.

Technology Can’t Do It All

Customer service has a long way to go. To a large extent, the health of the customer service function will always depend on human factors and the organization’s commitment to improving the customer experience. New concepts like advanced analytics and machine learning can help move the needle forward. But without the organization paving the way, innovative technologies can only go so far.