Boost Your Signal-to-Noise Ratio with Automated Underwriting


October 5, 2021

Signal-to-noise ratio tells you the strength of a desired signal in comparison to the level of “noise,” in this sense meaning anything that disturbs the signal. By reducing or eliminating the source of noise, you effectively increase the strength of the signal. It turns out that this engineering concept is also applicable to matters of human judgment, such as underwriting. Automated underwriting cuts through the noise and lets you concentrate on the signal. 

Noise In The Decision Making System 

In “Noise: A Flaw in Human Judgment,” authors Daniel Kahneman, Olivier Sibony, and Cass R. Sunstein define “noise” as variability in decisions that should, in theory, have identical outcomes. The book makes a strong case that this variability indicates an issue with the decision-making system. 

It’s important to distinguish between bias, consisting of systematic deviations, and noise, amounting to random scatter. An example the authors give is that of a bathroom scale. If your scale is consistently off by a single pound, that’s an example of bias. Once you’re aware of this bias, you can take simple steps to counteract it, either by adjusting your scale or with a bit of quick mental math. 

“Noise” in this example would be a scale that gave inconsistent readings from one day to the next. On Monday morning the scale gives a reading that’s five pounds lighter than it should be. Tuesday it swings in the opposite direction and provides a measurement that’s eight pounds overweight. When you weigh yourself again on Wednesday, the reading is seven pounds underweight. You cannot account for this variability, as it seems impossible to predict. 

According to the authors, these unwanted variations in judgements and decisions lead to “noisy” systems. Systems become noisy as different people apply different standards. In addition, the noise in a system also increases as the same people apply inconsistent standards from one case to the next. Something as simple as whether the day is rainy or sunny can result in different decisions on essentially identical situations. 

How Noise Can Affect The Payment Industry

In the book, the authors highlight how these noisy systems plague various professionals, including doctors, judges, and scientists. Underwriters are by no means immune. The book highlights one case where an insurance company performed a “noise audit” to determine the unpredictable variability of the company’s underwriting decisions. The results were shocking. 

Looking at identical claims examined by multiple underwriters showed noise levels five times higher than expected. Executives at the company indicated that the amount of noise confirmed to exist in their processes could be costing the company hundreds of millions each year. 

Eliminate The Noise With Automated Underwriting

Automated underwriting holds the key to cutting through the noise and maintaining a clear signal. An automated and configurable scorecard like Merchant ScanXpress uses the same checks you’ve already determined are essential, but removes the noise that comes from human variability. Given identical information and an identical configuration, automated underwriting will provide the exact same results. By eliminating the noise, automated risk scoring gives your underwriters more time to closely examine situations that truly need their expertise. 

For more information on cutting through the noise, boosting your signal, and onboarding more merchants, please see The Merchant Acquirer’s Guide to Onboarding More Merchants.

Are you ready to cut out the noise and focus on the right signal? Click here to sign up for a demo, or here for a free trial of Agreement Express’s Merchant ScanXpress.