For large corporates, forecasting accounts receivables is particularly challenging as the forecast figure is consolidated from a wide variety of inputs.
In addition, as accounts receivable (AR) is often largely made up of customer receipts, the timing element is hard to predict as it is out of the forecasting company’s hands. Though payment terms may be contractually mandated, they might not always be adhered to, thus adding a significant element of unpredictability to the process.
Long-term versus short-term AR forecasting
It is useful to break AR forecasting into two parts, long-term forecasting and short-term forecasting. Long-term forecasting of accounts receivable tends to be easier because customers will, generally, pay their bills eventually.
Short-term forecasting, on the other hand, tends to be considerably more challenging. This is because payments may be withheld and delayed for a number of reasons (difficulties in the supply chain, or holding cash for reporting purposes, for example).
Improving short-term AR forecasting
The following steps are built on our experience in helping large, multinational companies to set-up and maintain forecasting processes. Following each of these should help to increase overall forecast accuracy.
- Analyse historic AR data
While a historical view will not perfectly match a forward looking one, a good understanding of the previous AR behaviours can help identify trends that are likely to continue into the future.
A good analysis of AR data should highlight:
- Payment behaviours of the largest customers
- Problem areas within the receivables ledger (e.g. customers who consistently delay beyond payment terms)
- Highly volatile components of the receivables ledger (e.g. certain business units, product lines or customers whose payment times are variable and hard to predict)
- Seasonality (at both a macro and micro level).
Once these areas are highlighted, you will be able to focus attention where it will have the greatest impact. Following the 80/20 rule (i.e. spending 80% of the improvement efforts on the 20% of biggest contributing factors) will both speed up the improvements and make them more effective.
- Break AR into smaller sub-categories
Following on from the analysis of historic data, accounts receivable can be broken down into further sub-categories. The purpose of the sub-categorization exercise is again to shift improvement efforts to the areas that will have the greatest impact.
Sub categories can include (but are not limited to):
- Customer size. The largest customers will have the biggest impact on the accounts receivable figure, when compared with the long tail of smaller clients. Identifying these clients will highlight the areas of materiality where the greatest improvement efforts should be made.
- Payment terms. Dividing receivables into categories based on their payment terms can highlight if any particular duration (e.g. 30/60/90 days) is a problem area. Additionally, where you have significant flows from individual customers, linking payment terms to rules in the forecasting model provides greater insight into when payments are likely to be received, and allows you to benchmark if they are being met.
- Customer credit score. This entails breaking clients into categories based on their credit quality / score and identifying the most problematic areas.
- Payment date. Breaking receivables into the time of week or month that payment is expected (e.g. last week of month) can reveal further insights into accounts receivable behaviour.
Once any sub-categorizations have been made, applying the 80/20 rule can help to focus efforts on the areas that will have the greatest effect.
- Analyse and adjust
As with all elements of cash forecasting, improving AR forecasting accuracy is a continuous process. While the steps above provide a solid foundation for a good forecasting process, it is important to continually monitor and adjust assumptions as part of an ongoing drive for improvements.
The key analysis that will need by be conducted on an ongoing basis is a variance analysis between the forecast and actual figures. This analysis should reveal how any assumptions made in the process have performed, enabling them to be tuned and tweaked for the next forecast version.
Following a process of “monitor, update, iterate” will help to move a forecast from low quality to high quality after a relatively small number of forecast versions. (We discuss this in greater detail in our whitepaper discussing cash forecasting accuracy measurement.)
Use specialised software
As with all elements of the cash forecasting process, the biggest improvements in terms of accuracy and time savings can only be achieved by automating the process with specialised cash flow forecasting software.
Automation removes the administrative burden from the process and reduces the risks of human error. We discuss how automation can aid the entire process in greater detail in this article: Cash forecasting automation: a practical guide.