This article was written by Greg Council and first published on March 27, 2017 in the IBM ECM Blog and an excerpt is provided below:
Payment by check may be viewed as passé by many, but use of checks is still quite large. According to research by Aite Group, well over 50 percent of total business-to-business (B2B) payments are still made by check. In fact, 2016 AFP research showed that use of checks in B2B scenarios actually increased from levels in 2013.
As the latest Federal Reserve Payments Study revealed, the decline of checks has slowed. In the consumer arena, the ability to deposit checks via a smartphone has helped to maintain and stabilize its use, and in some cases, increased its use in person-to-person payments.
At the same time, checks continue to show the lowest relative levels of Straight-through Processing (STP) when compared to electronic payment mechanisms. This delays deposits and continues to carry significant costs associated with manual processes; compare the $10 cost to the $1 fee for a faster ACH payment. Additionally, according to that same AFP research mentioned above, checks also are the most common payment mechanism for fraud with 71 percent of surveyed companies having experienced check fraud.
So while pundits continue to predict the downfall of check use, many technologies are available to make check payments much less painless and help with achieving STP.
Pushing Paper to the Edge
Ever since Check21 was passed, check processing has increasingly become a digital affair. Whereas once banks were required to ship paper checks to each other, the paper is now scanned and only a digital image remains. Advances in ATMs with multiple check deposits without requiring envelopes further aided this digital transformation. It also increased use of remote deposit and mobile deposit capabilities allowing banks to entirely remove paper handling.
IBM Datacap Insight Edition offers cognitive capture capabilities, which can help move banks closer to total automation, while Parascript can help address the costs of ingesting traditional paper checks.
Data Validation: Manual Processing vs. Full Automation
So where are the continued costs of handling checks vs. fully digitized payment options? One area is with validation. When tellers were directly receiving and viewing checks, issues regarding negotiability were not considered a problem. If there was data missing on the check or if some data was non-compliant, the issue could be dealt with immediately. Now with remote deposit in its various incarnations of ATMs, desktop and mobile, the ability to identify problems before a customer submits a check is not quite as robust, which can result in frustrated customers.
Enter automated check validation. There is technology that allows the analysis of a submitted image to determine if the check can be accepted – often over 20 different areas are analyzed including the quality of the image, which can affect downstream data extraction processes. The analysis can also detect negotiability problems such as stale dates and missing endorsements. Fraud prevention techniques—such as determining if the written and numeric amounts match as well as verifying the check —can also be accommodated, all prior to transaction completion.
Check Validation in Real-time
Imagine a case where checks can be examined in real-time. If a submitted check doesn’t pass image quality requirements, it can be re-submitted immediately. If there are negotiability issues, the system can inform the user of the problem instead of the bank having to contact them a day later with the issue. Ultimately, dealing with these issues upfront can help avoid unnecessary costs for the bank and unwanted frustration for customers.
When moving branch transformation beyond typical transactions, other points of “transaction friction” for service delivery are sure to be scrutinized by banks to make them more efficient and more customer-friendly. Successful examples of this are in place today with regard to loan origination.
Branch Transformation to Service Opportunities-Focus
I believe that any document required as part of a transaction could and should be avoided in its physical form. If ID or proof of income is required, why force the customer to bring it to a branch or fax it? They can simply snap a picture with their phone and submit it remotely. The same quality and validity analysis associated with checks can be applied to any document that is part of a transaction. Documents can automatically be classified, presence and integrity of required data can be performed, and data automatically extracted and sent to the bank’s origination systems. The ultimate goal is to focus customer visits to the branch as service opportunities and not as paperwork tasks. Customers, in most cases, should not be required to step into a branch, but enticed to do so when wider range of service opportunities are available to them.
While much in banking has been automated to slash unnecessary costs and enhance the customer experience, much more can be done when the focus is on the straight-through processing of data that exists in documents.
Other recent articles by Greg Council
- Wanted: Smart Humans | IBM Big Data & Analytics Hub | February 20, 2017
- ‘De-Blackboxing’ AI: Bias in Learning Machines | EntrepriseTech | February 8, 2017
- Leveraging Signature Verification Across Industries | IBM ECM Blog | August 29, 2016
- Verifying Signatures: When Signatures Lie | IBM ECM Blog | August 15, 2016