Complaints can cost more than you think

Written by Colin Thompson | 27 September 2011

Even the best companies can sometimes get it wrong. They fail to deliver what is promised. When people, processes and technology do not complement each other and work in harmony, service delivery can stall, and customers end up dissatisfied. That puts them at risk of churning and giving bad word of mouth to their social networks. Companies, therefore, need to design their service delivery strategies to "get it right the first time" but, recognising that even the best sometimes fail, also have recovery processes in place to identify and retain valued at-risk customers.

The key customer-save course of action is the complaints-handling process. Customers who complain to service providers and are well treated by the process are less likely to churn than customers who have no cause for complaint. In other words, a well-designed, easy-to-engage and responsive complaints-handling process can help build loyalty.

A health Insurance company, which shall remain anonymous, recently engaged Cavendish to design and implement a complaints-handling process that was compatible with the new international complaints-handling standard, ISO 10002. This was part of the insurer's drive to implement strategic CRM. As with other investments, before committing funds, the client needed to establish a business case. We worked with the client to develop a methodology for measuring the cost of customer complaints. If we were able to show the true cost of the company's current crop of customer complaints, it would provide a justification, or not, for the investment. The revenue losses from unvoiced complaints were estimated at a staggering $16 million.

The methodology is, in effect, a spreadsheet of 10 interlinked pages that uses two forms of data. Factual data imported from customer databases in sales, service and marketing departments is supplemented with expert judgmental data. We worked with a small team of client employees to obtain the data, using our broader knowledge of customer-complaining behaviours to help them generate the judgmental data. The measurement model allows "what-if" scenarios to be considered and for judgmental data to be changed as more information becomes available.

The tool computes three clusters of cost. The first cluster is the cost of handling current complaints. Using the previous 12 months as the measurement period, we computed the number of complaints received; this alone was a challenging task, given that the insurer has four different touch-points at which customers complained. The company incurs real costs in handling these complaints: people costs, technology costs and communication costs; then there are customer recovery costs such as ex gratia payments, gifts, premium waivers and service re-performance. In addition, there are the indirect costs of management time, space and training. The final component of the first cluster of costs estimates the effects of churn and word of mouth that result from the complaints-handling experience.
The second cluster of "costs" computes the organisational benefits, if any, of the current complaints-handling process—any process, product or service improvement; productivity gain; or marketing saving that is a consequence of learning from complaints is given a valuation here.

The third cluster of costs computes the costs of unvoiced complaints—those problems or issues that customers experience but do not report to the insurer to be fixed. The "iceberg effect" means that many customers choose not to complain but simply churn and utter negative word of mouth, with potentially major revenue losses for the client.
The client insures more than 500,000 lives. As a not-for-profit, any surplus is returned to members in the form of improved benefits or reduced premiums. In the course of the previous 12 months, the organisation had received about 75,000 complaints, the vast majority of which had been easy and quick to resolve to the customer's satisfaction, but a small number of which consumed significant time and resource.

The first two clusters of cost aggregated to just under $2 million, an average of about $25 per complaint. However, when the third cluster of costs was computed, the revenue losses from unvoiced complaints were estimated at a staggering $16 million, meaning that the total annual costs of customer complaints totalled $18 million. The insurer's processes were practically invisible to customers, many of whom saw the business as slow, if not reluctant, to respond to complaints. Engaging the process could be time-consuming and troublesome for customers, so it was no wonder that some chose to churn.

So, how much should be spent upgrading the insurer's complaints-handling process? A new process that is easier to engage and more responsive encourages more customers to complain, meaning that costs of handling voiced complaints lift. However, the client could expect compensatory organisational gains in the second cluster of costs and the massive iceberg effect to be reduced. Our sensitivity analysis using the measurement tool indicated that a single investment of $1 million in complaints-handling would reduce the overall complaints-related costs by $5 million each year, delivering massive payback.

Our analysis is a good example of how you can improve customer loyalty while dramatically reducing your costs.

Colin Thompson
colin@cavendish-mr.org.uk
www.cavendish-mr.org.uk