Error reduction and risk

When you are at school and university, getting things right 96% of the time will get you a top grade. Work is different, and most trainee accountants soon realise that being right only 96% of the time at work can be disastrous.

Checking what we do and trying to be as reliable as possible are at the core of our training and built into many of the traditional techniques of bookkeeping and accountancy. Double entry bookkeeping requires double entries in part because of the inherent error checking it provided before computers did the job. Bank reconciliations are a chore we do not regard as optional.

But not every job creates the same pressure to be consistently correct. There are jobs in marketing where it is hard to know if you have been right or not. A sales person who made the right choice 96% of the time would be a star. Consequently, there are big differences between functions within organisations in their attitude to errors and their skills in preventing, detecting and correcting them.

Most of the time accountants focus on errors made within their own work but there are some very important exceptions. This article looks at some of the ways that accountants can be useful outside their department because of their attitude towards errors and skill at reducing them.

Management accounts can highlight the cost of errors. For example, instead of just showing a grand total for ‘sales’ we can show initial sales less (1) sales returns, (2) bad debts, (3) money and products given away to appease unhappy customers, and sometimes also (4) money lost by failing to produce an invoice for all the goods and services received by customers that they were willing to pay for, at the price they were willing to pay. Some of these amounts may be quite small compared to initial sales but they are more important because they go straight to the bottom line.

Incomplete and inaccurate invoicing can be very expensive so the main reason for not showing it is that you simply don’t know how much you have lost. However, there are often ways that you can quantify even these losses by making comparisons with other records of sales made. In the telecommunications industry this is called ‘revenue assurance’ and the losses it tries to reduce are called ‘revenue leakage’.

Seeing the money value of errors can help people decide when it is worth making an effort to reduce them. When the telecommunications industry first woke up to revenue leakage it was soon discovered that investing in revenue assurance produced huge percentage returns, making it one of the best investments available.

Reporting financial numbers showing errors can create interest in non-financial numbers that give more insight into the causes of those losses. Sometimes the causes have been hidden by well-intentioned adjustments to reporting to make it fair. The case of Portsmouth housing repairs services is an example.

The role of the department was to maintain council houses in the area, which it did by receiving calls from residents with news of leaking taps, broken windows and other problems needing to be fixed, and then dispatching tradesmen to do the repairs. All this activity was carefully recorded and performance reporting showed how quickly calls were taken and repairs were made. According to this reporting, performance was good, with almost all repairs completed quickly.

However, the real situation was very different, with roughly half the calls to the call centre being complaints about jobs not done correctly or chasing progress. There were several different reasons why delays were not counted within the performance reporting, such as non-availability of parts and tenants not being at home when the tradesman called. These delays were considered outside the control of the department and therefore excluded from performance reporting.

Managers responsible for this work were reassured by the performance reporting and unwilling to see for themselves what was really happening. But once this was overcome and more realistic reporting was in place the true situation was addressed. The stock of parts was managed differently and tenants were given more accurate times for appointments so that they could be in at the right time. As a result, the time taken to make repairs was reduced from about a month on average to just five days.

I recently came across a similar example where the time taken to recruit a new employee was reported according to complicated rules designed to exclude delays that were not directly the fault of the human resources department. Consequently, the time taken to get someone on board was grossly understated and opportunities to speed things up were being overlooked.

The top priority for accountants is usually to minimise their own mistakes and those of their colleagues in the finance department but there are situations where accountants can find themselves going outside this natural territory. For example, if you have been reporting the cost of errors as described above for some time but making improvements has proved difficult then people may start to look to you for suggestions, since you have been monitoring the situation and understand it at least enough to measure the errors.

There is also an important reason why accountants may wish to venture outside their own department. Some errors made within an organisation only come to light when they reach the accounting stage. An error made processing a sales order might not be detected until the customer complains about their invoice. Not all the errors that accountants battle with at the month-end are their own but we are still under pressure to close the books quickly and correctly. That gives accountants a legitimate interest in reducing the flow of errors coming to them even when those errors have arisen outside the finance function. Sometimes that legitimate interest becomes a pressing need.

Some of the accountant’s favourite controls can be very effective outside the finance function. For example, a cleverly designed reconciliation can reveal the location of faults deep within complex computer systems. There are also times when simply having a second person checking some work and writing their initials to show that they have done so is a worthwhile and effective control.

However, we have to be careful to avoid tackling all problems with our favourite solutions. Outside the finance function the goal is more often to prevent errors arising than to detect and correct them. Checks and reconciliations become less important and we need to focus on ways to make it easier to do tasks correctly the first time.

Common sense is not always a reliable guide to which changes will reduce errors. Some years ago I worked on a project that used statistical analysis to uncover the causes of errors in a large company. One type of error I studied intensively was entering incorrect counterparty details on commodity deals. Our initial expectation was that workload and lack of training would be important drivers of errors but the data showed that these made no difference. Measuring and monitoring the errors and reminding staff to try to avoid them also did not work. We discovered a problem with the computer screen design because the field for the counterparty name was too short to show some of the counterparty names the company was using. It seemed likely that this was the cause of the problems but it wasn’t.

Eventually I found one factor that strongly affected the risk of error. Errors were much more common with new counterparties and counterparties who had not traded with the company for several months. Novelty was the problem.

When doing something for the first time people are more vulnerable than usual to counter-intuitive software, over-complicated procedures, misleading words on forms and computer screens, and account code descriptions that are vague or misleading. Many of these problems can be reduced by careful redesign. We have to remember that other people are less interested in accurate record-keeping than we are and that, sometimes, giving people more instructions only makes things worse. Typically, clarity trumps persuasion.

If you need any convincing that there are errors to be reduced you need only consider your experience as a customer. While writing this article I had to call my new bank because something had gone wrong while setting up internet banking. The customer support representatives were delightful to talk to but I shouldn’t have had to. Or how about the internet security company whose software kept going wrong for me because my product key had been stolen, not from me but from them?

Here were two organisations with a powerful reason to impress customers with their trustworthiness by getting things right every time, and yet they failed. Is any organisation immune?

Download pdf article:

• Error reduction and risk
  Business & Management Magazine, Issue 260, December 2017

Further reading into error reduction is available through the resources below.

Online articles

The Library provides access to leading business, finance and management journals. These journals are available to logged-in ICAEW members, ACA students and other entitled users subject to suppliers' terms of use.