6 Challenges You Need to Overcome to Run a World-Class Accounts Payable Function
When you’re in the business of optimizing Accounts Payable, there’s some issues you come across over and over again.
Some challenges seem to be universal no matter the size of the company, the industry, or which transactional systems they’re using. And they’re the reason AP departments everywhere struggle to achieve strategic outcomes like increasing DPO, capturing cash discounts, and getting their touchless invoice rate up. If you want to be the best, then these are crucial challenges to overcome.
These problems aren’t the only thing most AP departments have in common, either. Often they spring from the exact same root causes too.
But most departments are too siloed, their data too disparate to easily establish what those root causes actually are. It can take months of user interviews and post-its to get an answer — and often that answer is subjective at best.
Enter process mining. We’ve collapsed those months of toil and drilled into more than a hundred projects’ worth of data to bring you a quick whistle-stop tour of the 6 most common problems we see in Accounts Payable, why they happen, and most importantly, how you can fix them.
1. Payment term discrepancies between invoices and POs
Namely the single most common reason you’re paying too early, tying up unnecessary working capital.
Usually these discrepancies happen because vendors issue invoices with their standard — or worse, outdated — payment terms, rather than the payment terms negotiated and issued on the PO.
Unfortunately, ERP systems aren’t designed to highlight these errors — there’s too much meta-data involved. You can catch them manually, but it’s a finicky and time-consuming job.
Thankfully technology does exist to help you automatically identify these discrepancies and default to the more preferable set of payment terms, delaying payments until they’re actually due.
A study by The Hackett Group found that companies can improve their cash positions by up to $358B by improving their payment behavior, so this is a pretty worthwhile first step.
2. Missing Documents
AKA: "Where on earth is my goods receipt?"
This isn't just one of the main causes of manual rework and lengthier-than-usual cycle times — it also plays a huge part in missing cash discounts. Typically you only notice this when you’re processing the invoice for payment, and by the time you’ve requested a goods receipt and it’s actually come through, that cash discount will be long gone.
This happens for the obvious reason: requesters and receiving fail to complete the goods receipt when they should.
But rather than bombard them with emails, how about setting up a system that automatically requests the goods receipt, and removes payment blocks when it’s finally submitted?
Sure, most businesses capture less than 21% of all early-payment discount offers. But according to Zycus and Hackett’s P2P Benchmark Study, the best AP departments in the game capture at least 80% of their cash discounts. (Some, like Deutsche Telekom, capture as many as 95%).
Tackling missing documents — amongst these other challenges — is a good move if you want to play in the big leagues.
3. Late Invoices
Perhaps unsurprisingly, the biggest cause of late payments. Because how can you hope to pay on time — let alone early enough to capture a discount — if you’re getting invoices after they’re due?
Usually, this is because some vendors are still billing you manually rather than electronically — so it takes longer to process those invoices in the first place, which also lengthens cycle times. Manually updating baseline dates and letting your vendor know about it takes up valuable time.
Ideally, you’d automate that process — up to and including the email — and flag the pattern to the sourcing manager as additional leverage across the negotiation table.
4. Invoices Filled in Wrong
Also known as: “We ordered 10, not 100.”
Say goodbye to your touchless invoice rate — incorrect fields inevitably mean time-consuming manual rework. No wonder only 39% of electronic invoices generally received by businesses can be processed straight-through without human intervention — but it’s a dire state of affairs when the best-in-class are achieving a 75% touchless invoice rate.
The main reason things are going wrong here? Incorrect or out of date master data, or good old-fashioned manual errors. Happens to the best of us, but it’s painstaking to fix, and it will affect overall productivity and cycle times.
The answer? Machine Learning is in fact able to detect discrepancies by matching purchase orders, historical data, and the invoice, and automatically update standard fields like currency or VAT. Voilà, your invoice stays touchless, and you’ve got a shot at that best-in-class touchless invoice rate again.
If the error is in a non-standard field — say price or quantity — consider adding some AI to the mix. You could be getting automated recommendations about what the correct value should be, making the process of fixing it significantly faster.
5. Duplicate Invoices
Nobody wants to pay twice for the same thing, and yet it happens all the time — generally because duplicate invoices end up in different ERPs for different regions, and you’ve got no way to spot them.
It’s a really frustrating way to haemorrhage cash, and once again, not one ERP systems are generally designed to tackle. For this, you need a way to monitor, detect and block duplicate invoices across all of your finance systems, regardless of ERP and system consolidation.
The good news is, that once you do find a way to handle the issue, you can eradicate duplicate payments forever. (That’s what Deutsche Telekom did, and they saved $3M in the bargain. Not bad!)
6. Lengthy Approval Times
Also known as: “I can’t do anything until Janet approves this, and she’s out on vacation until next month.”
The blame may lie with someone else, but you’re still going to have paid full price, and no one is going to come out of it looking good.
It might be time to streamline your approvals process and make sure that the right people have approvals on their core to-do list. Having said that, parallel processing can also do a world of good here.
For example, you don’t really need to put approvals on hold for non-critical fields, like billing address or company code, when your approver’s just looking at the price and quantity — so why not expedite the process and run at the same time?
The good news: solutions for these AP problems exist.
Are some or even all of these problems getting in the way of your KPIs? Don’t despair. While your ERP systems may not be able to fix them, other technologies — like process mining, judicious automation and a dash of artificial intelligence — can.
AP departments the world over are experimenting with possible solutions that don’t require them to rip and replace existing systems — but rather work as an intelligent layer to extract information and value in a way that previously wasn’t possible.
It’s part of a conceptual shift reconnecting Accounts Payable to the broader Finance organization — one that recognizes that AP isn’t just a cost center. Rather than just optimizing to process invoices as fast and as cheaply as possible, AP has the potential to have an impact on far more strategic outcomes — like working capital and cash preservation.
But first, it needs to tackle the issues that are getting in the way of achieving its overarching KPIs.
Find out more about how companies like Becton Dickinson and Deutsche Telekom tackled issues like late payments and duplicate invoices.
