Understanding paralanguage and how it can be used to reduce non-disclosure is key to the use of tele-interviewing as a sales tool, explains Andrew Gething
Non-disclosure by definition refers to information that the consumer has not volunteered to the life company. In practice the term is used to cover all missing information that should have been collected and volunteered at application stage. This includes information that was deliberately withheld, not understood or not asked for.
Recent studies of non-disclosure comparing tele-interviewing and traditional paper applications have shown that the vast majority of missing information is due to the process of collecting information rather than the applicant wilfully withholding. This is why tele-interviewing has been so successful in reducing non-disclosure. Further studies have shown that the percentage of real material fraud is relatively low and more have indicated that for tele-interviewing these are around 1% for life and critical illness and 4% for income protection.
This article examines how advanced conversation management techniques can be used to reduce the last 1%-4% of material non-disclosure by detecting attempted deception and fraud and reducing this within the tele-interview itself. In effect, to utilise human lie detection techniques.
The techniques rely on the nature of conversation. We modify what we say everyday by the responses we receive. These include the spoken word, as well as many other indicators of communication. Paralanguage are the non-verbal elements of communication used to modify meaning and convey emotion and may be expressed consciously or unconsciously, including a wide spectrum of variables such as intonation, and voice stress variations.
Learning to talk
Paralanguage preceded the spoken language; indeed, humans used it extensively in cave man ancestry before developing speech. It is estimated that 50%-90% of communication is non-verbal, and provided through paralanguage and body language.
Humans rely on paralanguage in many aspects of their professional life. We would never employ someone before interviewing them over the phone or in person. We recognise that our impressions of someone are greatly influenced within the first 30 seconds of meeting them, relying purely on non-verbal indicators and before we have asked any factual questions.
It is therefore important to be aware that the way a word is spoken is as important as what is said. It cannot be controlled by scripts, but only by training the interviewers and their listening skills.
The legal profession and compliance departments drive people to written documents and logical questions and answers, but even the former admits this is inadequate with the ultimate resolution of issues being decided verbally in court where paralanguage and body language can be heard and seen.
While verbal communication is good at communicating facts and figures, paralanguage is better at revealing the emotions, conscious and unconscious, that lie beneath all communications.
Question and answer time
To date, the use of paralanguage has been largely overlooked, with the industry being focused totally on the questions that are asked when it is the responses that are the most important. This is understandable as when using the traditional paper application processes, the industry could only control the questions and had no influence on the answers and responses. This has been perpetuated with the adoption of expert systems that can store and control millions of questions. Now, in the tele-interviewing era, the industry has control of both the questions and the responses.
Once an interviewer is trained on advanced conversation techniques they can detect changes in paralanguage which could be indicators of stress.
Fraud detection software has become a current hot topic on the basis that it can reduce non-disclosure. Just like the polygraph, it can detect a change in stress of an individual. Stress in an interview may be the result of fraud or a sensitive and emotional topic. Stress can be detected by the interviewer as well as by software. Fraud detection software can only detect a limited amount of all the varied sources of paralanguage communicated during an interview but an interviewer can detect multiple types.
During a discussion a trained interviewer can use the detection of changes in paralanguage in several ways during a discussion. The first is to immediately react to the answer and seek clarification, or re-ask the question in another way. This is termed reactive probing.
The second use is to objectively report back to the underwriter the detection of unusual paralanguage, which may indicate fraud. This is effectively a human version of lie detection.
Keeping the peace
The hypothesis is that using paralanguage indicators is a more effective way of identifying when further evidence is required than using traditional medical limits. It is thought that underwriters will assess the risk associated with an interview that reports significant indicators of stress and will seek independent medical evidence. This may be done during the new business process or after a decision check. It is postulated that this will detect fraud at an early stage and avoid confrontational issues at the time of claim, and that as a process it can replace the traditional model of medical limits because it is more focused.
These advanced conversation management and listening techniques are already used in general insurance, mostly on the claims side, with great effect. As this is a listening technique there is no audible or other indicator that it is taking place, just as in normal conversation. There should be no fear of its use or that it will be a detriment to the consumer, as people use it informally in every conversation they have.
Like with a lot of amateur skills, a little knowledge can be dangerous and, if it is to be used in interviewing and tele-claims, the interviewers must be trained and monitored. People who have not been trained will not normally pick up the finer points of paralanguage because their listening skills are not good enough, and therefore its use and the audit of its use can only really be done by those who have been properly trained.
The aim is to reduce the last fragments of material fraud during the application process, so that there is less conflict at the claim stage. It can be argued that clear fraud can be rejected at claim stage and there is no cost implication to the life company. However, there are several reasons why it is still worth reducing fraud at the application stage. Firstly the identification, management and processing of contentious claims takes a lot of time and is expensive, particularly when paying legal fees. Secondly, and probably more importantly, the industry needs to regain the trust of the consumer, so it needs to appear squeaky clean, and remove any ambiguity about paying out claims. The industry needs to go the extra mile to be able to report on excellent claims payout ratios.
Andrew Gething is managing director of MorganAsh
Typical paralanguage includes:
- Tone of voice, intonation, rhythm, speed, volume and pitch.
- Stalling for time, 'umms,' and 'ahs'.
- Repeating questions back to the interviewer.
- Increased rate and sound of breathing.
- Sniffing, swallowing, sighing, shuffling.
- Nervous laughter.
- Silence.
- Mumbles, interruptions.
Tele-underwriting - the basics:
- Tele-underwriting saves money with costs ranging between £20-£35 an interview depending on the types of interview required and methods of transferring applications. This compares with typical costs estimated at £300 an application, and £70.50 for a general practitioner's report.
- Most interviews can be completed within a few days.
- Costs of acquiring new business are reduced.
- Faster turn-around times mean the amount of policies sold increases due to improved customer satisfaction and reduction in NTUs.
- Better quality information enables better analysis of risk and hence more accurate premiums. The result should be improved mortality profit.
- Improved customer satisfaction improves the insurer's public image.
