Think Big…we are in the dawn of “Big Data” and companies need to adapt to prosper from the massive amounts of high velocity data. No longer are companies constrained to samples or long runtimes of programs that could take days. Running a complex analysis on millions of line items in seconds is here, and thanks to SAP’s Hana Cloud Platform, FIRSTPRISE is excited to offer the payOnce application to take advantage of the processing power of Hana to make otherwise hidden data visible. Another way of putting it, if you’re looking for that needle in the haystack, payOnce can point you to a very small corner of the haystack to find the needle you are looking for, or more importantly, find the needle you didn’t know even know existed…
payOnce offers 3 data analytics solutions to analyze your data, and many more functions in progress that we will be rolling over the course of the next few months:
- A duplicate detector
- Relative Value Analysis
- Benford’s Law Tests
According to the International Institute of Internal Auditors, between .05% – .1 % of company spend is wasted on duplicate payments. Other estimates are as high as .3% of spend. For an organization with $1 billion in spend each year, this equates to $500,000 to $3,000,000 per year.
- For a company with Profit Margins of 10%, this equates to wasted sales of $5,000,000 to $30,000,000 per year!
- For a company with profit margins of 5%, this equates to wasted sales of $10,000,000 to $60,000,000 per year!
For a company with potentially millions of invoices, analyzing the data can be time consuming, if not impossible. Bringing in recovery specialists can be expensive, as they take up to 50% of what they recover. Our application is able to use multiple algorithms you design with our tools, to not only search, report and status duplicate items, but also items that are not entirely the same, but similar enough to warrant further research. It uses the same-same-different approach to look for entries that not just exactly the same and perfect duplicates, but entries that are almost similar, but may differ from a field or two. For example, one of the top causes of double payments to vendors is the same invoice being processed against 2 different vendor Id’s. Any control to check if the invoice was already entered then would be circumvented. The best controls in place in most ERP systems can prevent the same invoices from being entered twice, but only if they are exact matches, and even then some can slip through.
We also allow you to run the analysis and exclude certain AP Data. For example, invoices with criteria, such as document type, that identifies it as an inter-company document, can be excluded.
Once we group similar invoices together, research can be carried out to prioritize and status the groups of invoices, so they don’t appear on the radar again.
We focus primarily on invoices, but the same function can be applied to other areas of your accounting data, such as :
- Fixed Assets
- Customer Credits
- Insurance payments
With Relative Value Analysis (RVA) payOnce models your invoice data to look for anomalies by comparing the largest invoice by vendor/company code to 2nd largest, to the average, and and to the median. Ranking the results by the ratios, you can quickly identify questionable transactions. For example, decimal-point error, where a $16,000 invoice is recorded as $160,000, would be apparent very easily in the Relative Value Analysis. RVA can be used not just for invoices, but for other transactions: inventory by material by plant, payments, fixed assets, sales by SKU at stores. All could highlight meaningful differences that may lead you to discovering errors, abuse or fraud.
Benford’s Law is a curious phenomenon in mathematics. It states that in a set of naturally occurring numbers, about 30% of those numbers will start with the digit ‘1’, 18% will start with the digit ‘2’, and so on that 4.6% will start with the number 9 . The law also applies to the 2nd digit, and the first 2 digits, 3 digits and so on . Although this law was discovered in 1881 by Simon Newcomb, it was re-discovered by Frank Benford in 1938. In the natural world, the lengths of rivers, populations of towns, heights of all the buildings in the world – all conform to Benford’s Law. And, so does much of your accounting data! It wasn’t until 1999 when it became a critical tool in looking for accounting fraud & errors when Mark Nigrini published a paper about it’s effectiveness in uncovering fraud.
Our Benford’s Law function allows any given data that you come up with – invoices, expense reimbursements, overtime hours for example – and it will analyze your data and compare to the Benford’s Law distribution. By looking at a high level – the first digit, you can drill down to the 2nd 2 digits and the third if needed, to look for anomalies in the data. Often the data can be explained, but often the data can focus on errors, abuse, fraud, or inefficiencies: that needle in the haystack you didn’t know existed.
(Disclaimer: Benford’s Law is not used to find fraud where none exist. Discrepancies do not necessarily indicate fraud or abuse and often they can be easily explained)