We live in a self-service world; we all use ATMs to withdraw cash, automated checkout services at the supermarket and even self-service check-in counters at the airport. But ask consumers whether they would prefer to deal with a live agent or an automated system when calling into a contact centre, and most would say they steer clear of automated systems.
Yet when surveyed, on average 85% of consumers respond to an automated payment reminder and 94% to a fraud verification request. And most find the automated option more efficient, believe it to be more secure than contact with a live agent and report high levels of satisfaction with their interaction.
This gap between the perception and reality of auto-resolution illustrates an intriguing paradox, perhaps best explained by the psychological barrier in the use of the technology. Consumers believe they would rather not use automated systems to interact with their bank or telco provider, but when they do, they find it works surprisingly well.
There’s an art and a science to overcoming this barrier through the adoption of best practices in how, when and what companies contact their customers for with automated interactions. Research shows that customer satisfaction can actually be increased through the use of automation. It’s all about understanding the psychology of customer communications.
To begin with, it’s important to note that auto-resolution works better for some applications than others. Successful applications include intervening in suspected fraud attempts – reaching out to card carriers to confirm whether a transaction that has a high risk of fraud is valid – and in debt collections. Fraud requires fast, efficient interactions and in collections, the embarrassment of speaking to a live agent about their financial situation can be overcome if consumers are talking to a virtual agent.
In such sensitive instances, security is a vital concern and all the evidence shows that consumers perceive machine-based interactions to be more secure than human-based interactions. Consumers believe a live agent call is easier to replicate than an automated call and treat the live call with a higher degree of suspicion; a fair assumption, as it doesn’t take much for a fraudster to go through the phonebook and pretend to be a financial institution.
Psychologically, when there’s been a high-profile security breach, such as the HMRC’s recent loss of benefits data, as many as 10% of consumers refuse to authenticate themselves to live agents. Yet auto-resolution services experience no change in the number of consumers willing to go through the same process.
One major financial services company, put a line in the script where if the caller had any concerns about the call, they could disconnect, and call a number or check on a website to see that the call was genuine. The company had a negligible number of disconnects (0.03%) in this scenairo, showing how much consumers trusted the service. Incidentally, when people do disconnect, the system calls them back, explains the situation to them and the next escalation steps if they refuse to interact with the system.
Automated calling is of course more efficient, as the system can make more calls than an agent. In like-for-like scenarios, automated calling achieves 20% more right-party interactions than human agents, because the system doesn’t have the physical resource constraints that human call centres have, and can keep on calling until it gets through. But not only is the system more efficient for the company, consumers also perceive it to be more efficient for them. In a customer satisfaction survey carried out with one company, 82% of its customers felt the calls were more effective and efficient than human agents and 89% thought they were easy to understand.
Right Voice, Right Time
Another important factor in overcoming the psychological barrier is the script. Generally, the more sophisticated you make the scripts, the better response rates you get. But the key is understanding the call from the consumer’s perspective. For example, by offering the simplest route to an outcome, you can guide people down a particular route. When faced with the option “Press 1 to make a payment, press 2 for more options”, most people will proceed to pay.
People don’t want to be defaulting on their debt payments either and most are only in a collections scenario because they think they can get away without paying. So if you use the organisation’s name first in a collections call, customers will quickly hang up. By using the consumer’s name first, and establishing a personal connection just as the best agent will do, can reduce disconnects by over half. Similarily with fraud, if you use the word “fraud”, consumers get suspicious and don’t continue. Instead, the script uses the word “security”, as it personalises the interaction into something the consumer cares about.
Speaking in consumer-friendly language and letting the consumer feel in control of the call – for example by specifying the amount they wish to pay – improves results too. In our survey, 77% felt in control of the call, even though it was not instigated by them and typical promise kept rates rose from 70-75% with live agents to 85-90% with an automated system.
When designing the script, there should be a natural rhythm to these calls, with a quick initial engagement, then a slowdown while the system reassures the consumer about what they need to do to complete the interaction, then a quick resolution. Understanding the best time to call is also about psychology – people are receptive to an auto-resolution request first thing in the morning when they think they can resolve an issue before they leave for work, at lunchtime or in the early evening, but not too early so as to intrude on family time.
Understanding the psychology of consumer communications is a little bit of art and a little bit of science, but it also comes with a long legacy of practical experience. And when carried out effectively, the results speak for themselves.