It pays to be accurate

01 September 2019

This article was featured in the September 2019 issue of the magazine. 

Catherine de Salvo, managing director at Scott Bradbury Ltd, delves into the cost of human error within payroll, explores why relying on software isn’t always a good idea, and provides an introductory tip or two for overcoming the natural barriers to accuracy. There are a couple of fun accuracy activities, too! 

Imagine the scene: a group of people from different organisations, brought together to discuss ways of reducing data error. In the group are three or four payroll professionals. If you were one of them, what examples would you have of things that have gone wrong with your payroll? How about, continuing to pay someone long after they’ve left? Starting a new employee on the wrong salary? Paying part-time staff full-time rates? You undoubtedly have your own horror stories of things that have gone wrong, despite your clever payroll software, which promised to eliminate mistakes.


The error cost multiplier

Accuracy within payroll is crucial. And the good news is that typically, most competent people have an accuracy rate of around 97% so most of the time everything works fine. But that 3% error rate is a killer in terms of wasted time and damaging repercussions. 

Making an error in someone’s name or transposing a couple of digits takes a matter of seconds. Finding it again, so you can put it right, can take hours, days or even months, depending on how quickly the error is picked up. 

It’s surprisingly common for people not to spot that their payslip or bank balance isn’t right – even if they’ve been underpaid – until the repercussions start. It’s the ‘error multiplier’ factor which takes a tiny error rate of even under 3% to seriously expensive levels. A single data input error which takes seconds to make, can waste 20% or more of an employee’s time. 

...typically, most competent people have an accuracy rate of around 97%...

The ripple effect

Take a piece of paper and draw concentric circles on it, dividing the inner ring into four sections and the outer ring into eight sections (see Diagram 1). In the centre circle write down a short description of a payroll error you know about. In the first (yellow) ring, identify four different direct consequences of the error you have identified. What happens? Who has to be informed? In the outer (blue) ring, identify a further two repercussions for each direct consequence. How does it get escalated? What are the knock-on effects?

It’s easy to see how a single error ‘ripples out’ to demonstrate the multiplier effect. One apparently ‘tiny’ error usually leads to several people and other organisations being involved, disruption to work, significant wasted time and damage to trust and reputation.


Causes of mistakes

People don’t make mistakes deliberately, so telling them not to doesn’t work. It’s as helpful as telling people not to get a cold. Sound advice, but useless. People work in good faith intending to be accurate, so what’s going on when a mistake happens?

One of the reasons for mistakes is to due to the way our eyes work. We have no control over the movement of our eyes as we ‘read’ a piece of data. 

Take the bank account number 89792939. As our eyes look at the numbers, they literally jerk backwards and forwards across the numbers. It’s easy to look at the overall shape of the digits, to transpose them, create repeated digits or simply just read it incorrectly. Simply because of the way the human eye works. Where the eye stops is called a ‘fixation’ and the jerky movement is called a ‘saccade’.

But if you use our ‘advantage of threes’ technique, naming each single digit in your head, in a waltz-type pattern, you focus on each element of the data, with a little emphasis on each group of three like this: 897 929 39. Sub-verbalising the digits in this way, means it’s possible to ‘listen’ for mistakes as well as to look for them. 

There are different techniques for working with grouped (or clustered) data like telephone numbers; or punctuated data like sort-codes or salary amounts; or alphanumeric data like postcodes or National Insurance numbers. Learning and practising these techniques with different types of data is proven to reduce human data error by 50–60%. (The 50-60% is the typical error reduction achieved by participants learning these techniques in the training programme Developing an Eye for Accuracy.)

Other causes of error include having too many levels of checking, where increasingly senior people check a percentage of the output; being stressed; memory lapses and not knowing how to manage distractions. Two other major factors, which we’ll explore briefly now are the way our brains work and the kind of mindset we need to adopt when working with payroll information.

The brain sees what it expects to see.

You cna raed tihs sentnece even thoghu most of teh wrods are not splled crroectly because our brians see waht we expcte to see, not waht is actaully tehre. 

Our reading skills get in the way of our accuracy skills, so we tend to look at the overall shape of data. Moreover, when we are familiar with a task and know what to expect, it’s easy for our brains to ‘sort out’ mistakes for us, so we don’t even ‘see’ them. There is nothing wrong with our optic nerve; but our brain automatically compensates for the mistakes in front of our eyes. This is why it can be so hard to believe we’ve made a ‘silly’ error. But when it is pointed out to us, or we examine the data more carefully, we can see it immediately. 


Being present-minded

Adopting an accuracy mindset is essential. 

Related to concentration and attention to detail, present-mindedness goes one stage further. We define it as ‘being fully engaged in the task and all factors affecting it’. This means that accuracy is improved by being aware of the consequences of errors and by taking conscious steps to mitigate the risk. Since people are generally fairly accurate, it’s easy to be lulled into a false sense of complacency – and therefore it’s important to adopt a ‘self-sceptical’ approach where you assume a questioning approach to your work. Actively looking for mistakes before processing data means you are likely to find them before they do any damage and when they are easy and quick to correct.’s easy for our brains to ‘sort out’ mistakes for us, so we don’t even ‘see’ them


Clever software encourages the wrong mindset

Software organisations like to suggest that payroll errors can be reduced or even eliminated by computer power. The automation of processes previously dependent on data input is of course usually more efficient and not prone to human error. But this in itself doesn’t eliminate mistakes completely. If a process is largely automated, the people using it tend to rely on it to get everything correct. If attempts are made to eliminate errors by tightening the rules and procedures, the unintended consequence can be to squeeze intelligence and judgement out of the process, and that can lead to an increase in error. People who are encouraged to depend on the system alone, without applying critical thinking, stop being proactively engaged in the task. Once in a system, mistakes have a nasty habit of recurring and causing exponential damage, whereas a present-minded, thinking human being would spot and query any anomaly.


The blind side

Mistakes are a drain on productivity. They waste time and cause us to be inefficient. And they sometimes have far-reaching damaging consequences. And what’s worse, we are often blind to just how costly they are. We try to combat error with systems instead of developing human critical thinking skills to spot the problems which systems never can. Make no mistake: it pays to develop payroll professionals’ data accuracy skills.

In that vein, here’s another fun activity. Carefully follow the instructions given in Diagram 2. Your challenge is to find the hidden mistake. Did you spot it? If not, email [email protected] for the answer.

Mistakes are nearly always hidden from view – until you spot them. That’s why our talk this year at the CIPP Annual Conference and Exhibition is called ‘The blind side’. Come along to see Hugh Murray, master accuracy trainer at Scott Bradbury, explore the impact of human error on payroll professionals and share more fun activities and skills development ideas for overcoming the barriers to accuracy.