Controls for the Credit Department (#3)
/In this episode, we discuss the controls associated with both a manual and computerized credit department. The main control points and related topics noted in the episode are noted below.
These controls can be dealt with in terms of increasing levels of sophistication. At the lowest level of control usage, you don’t really need the credit department at all, because the whole thing is a gigantic control. Very small companies usually exclude it from the order process flow, so they automatically accept orders from all customers, and don’t worry much about bad debts.
Credit Applications
However, after a while a business starts to see a few bad debts. And when that happens, the typical reaction is to issue a credit application if orders are fairly large, and to not bother with a credit application if orders are relatively small. When this happens, employees will just go to the local office supply store to buy a block of standard credit application forms and that’s what they give to customers.
Approval Stamps
Now that’s not a sufficient control, because what you also need is a credit approval stamp. Someone needs to review completed credit applications, decide whether or not to grant credit, and if so, then stamp a copy of the sales order –you may recall the sales order from an earlier episode. The credit department keeps one copy of the stamped sales order, and sends another copy to the shipping department. The shipping manager is not allowed to ship anything unless that credit approval stamp appears on the sales order.
If anyone tries to add a fake approval stamp to the sales order, well, the credit department has kept another copy of the sales order. And, if a question arises about a sales order, then they can pull it out and compare it to the sales order copy being held by the shipping manager. To make sure that this control works, the sales orders held by both the credit department and shipping department should be locked up, which makes it very difficult for someone to illicitly stamp both copies.
And finally, at this very elementary stage of control, you need one policy, which is – don’t ship anything unless there’s a credit approval stamp on the sales order. This policy needs to be drilled into the shipping manager.
Credit Policy
Now, moving up to a bit larger order volume, you need to develop more standardized rules for which credit applications are accepted, and which ones are rejected. This requires a very formalized credit policy, which needs to contain several key items. First, it should outline a collection goal. For example, it could say that bad debts can be no more than 2% of sales. Or, perhaps from an efficiency perspective, it specifies how many collections employees are allowed per thousand customers, or perhaps it lays out a collection target, such as the number of days sales outstanding. Whatever this goal might be, it needs to specify the performance expected of the credit department, and itemize precisely how credit is to be granted, so there’ll no longer be any variation involved in determining who gets credit, or not, and how much credit is granted.
In addition, they need to have a supporting policy, which mandates a review of the main credit policy at least once a year. This supporting policy is intended to compare the existing credit policy to the company’s current financial and operating position – such as its product margins, strategic direction, general economic conditions, the actions of competitors, and so on.
The reason you need this review is that, for example -- a company may have an innovative product with a high profit margin, and since the profit is so large, they can afford to grant easy credit, and if there are bad debts, they can be easily absorbed. But over time, pricing pressure from competitors may shrink those margins, and that means a much tighter credit policy is necessary to meet the changing conditions.
Procedure Training
Another control you need with this increased order volume is regular staff training in credit procedures, with the intent being to standardize credit granting as much as possible, so there’s no variation by person.
Credit Application Analysis
And a final, fairly minor procedural item, though one that’s quite necessary, is to investigate unanswered questions on the credit application. Early on, when credit applications are being used mostly as an afterthought, this tended to be ignored. But now, with higher transaction volume, you want to lock down the process, and so – in addition to the policies and procedures and training, you have to make sure that the credit applications are totally completed, every single time.
Credit Limits
Now, let’s add a level of complexity to the credit function, and add computers to the mix. This happens when transaction volume increases to the point where credit can no longer be handled manually. The first control should be built into the accounting software, which is a field for a credit limit. If you fill this in for every customer, then the computer system will tell you whenever a customer is exceeding its credit limit.
Also, there’s no longer a need for a credit stamp on the sales order – this is partially because there’s no longer a sales order, because it’s been automated. It’s also because there should now be a credit approval flag in the system. To set a credit approval flag, any person with the proper password can enter the accounting software and flag a specific order as being approved for payment. And if it is flagged, then that order should now appear on the daily shipping report, since that report should only list orders ready for shipment if their credit flags have been set.
If the credit flag has not yet been set for an order, then that order won’t even appear on the shipping report, so the shipping manager won’t ship it.
Additional Credit Controls
Besides these basic credit controls, there are a few others to consider. You shouldn’t necessarily install them all, since that could become burdensome, but they are things to think about if you want to create tighter controls for a computerized credit system.
First, you can create a report that itemizes all customers that comprise the top 20% of your sales. And since that’s where most of your credit risk is concentrated, that’s where you might want to conduct a new credit review for every customer, every single year.
Another control is to review credit levels if a customer skips payments. This is a very difficult thing to spot with a manual system, but it’s very easy to notice with a computer system.
Another possibility is to review credit levels for all customers issuing NSF checks, which is not sufficient funds checks, and in this case – again – very easy to spot with a computer report.
Another option is to review credit levels if a customer stops taking early payment discounts.
So there are a few extra controls to consider.
Now, an additional level of sophistication with a computer system is to use it to access credit information over the Internet. I personally use Dun & Bradstreet’s Business Information Reports. And you can also have a credit rating agency notify you by e-mail when a customer’s credit rating drops, or when some other credit-related issue arises.
If you want to use a bit more technology, consider either posting your credit application on your web site, or e-mailing it to customers. Better yet, set up an electronic form on your website, so they can enter credit information directly into your computer system.
If you really want to get fancy, and spend a great deal of money on programming, you can set up automated rules-based decision tables to evaluate incoming customer orders. These tables can be very sophisticated, but on a basic level, what you’re looking for is something that will scan incoming order and if it’s from an existing customer, then it’ll continue with additional processing. If not, it’ll shunt the information off to a real person for manual review. However, if it is from an existing customer, and if there’s been an order within the past few months, the system will continue with its automated review. If not, and the customer has not placed an order in quite some time, then it can be shunted off again to a real person.
And then, finally, for that limited subset of customers falling within the first two criteria, it can automatically flag them as being approved if it fits within a specific credit limit – perhaps anything under a few thousand dollars.
That’s a really simple rules-based decision system. These systems can become very complicated, and should they be – if you have a lot of customers, and you can’t possibly review them all manually. You’ve simply got to find a way to let the computer system help you out, and a decision table is a good way to do that.