Apart from government agencies, financial institutions endure some of the most onerous regulations when it comes to managing information, as well as the high volumes of paperwork that typical business transactions require. While banks have tried to reduce paperwork costs over the last few years, the burden of managing document-based information is still a significant challenge.
Take, for example, the task of maintaining documentation for regulatory compliance. Understanding the scope of the regulations in question and how those apply to a specific financial institution’s documents is one challenge. Another is how to describe the documents in a manner that allows for easy identification. Yet another is how to enforce the application of those descriptions without harming staff productivity.
Account servicing also poses issues, whether during a specific business transaction, such as with a new mortgage, or over the lifetime of a service such as a checking account. One department’s notion of document descriptions and metadata (search terms for documents) might be significantly different from another’s, yet both often rely on a common set of account documents.
This inconsistency leads to the inability to locate existing documentation or delays in retrieving it. Both can lead to suboptimal customer experiences, either due to requesting the same documentation multiple times or through the inability to quickly handle an inquiry.
What happened and how did the financial industry arrive at this level of disorganization? Clearly, the adoption of new channels and technologies has contributed to a fragmentation in how documents are received and recorded. Another culprit is a misplaced faith in document management technology and relying on it alone to solve the problem.
We have worked with both small and large banking institutions and found a common refrain: an enterprise content management system purchased years ago is not configured or architected to allow retrieval of documents that the bank needs today. In one case, documents for an account were scanned all at once and stored as a single binary file. As a result, the bank’s staff has to sift through potentially hundreds of pages to find the necessary information. In another case, a decision made years ago to record only the account number on related documents did not satisfy current retrieval needs.
While these core problems were different, the result was similar: the customer service representative had to review multiple documents to find the correct one. Multiple causes can lead to the same effect: the lack of efficient access. And significant costs are associated with these problems. One bank was using a staff of 40 attorneys to classify documents and another staff of 60 to apply the metadata!
I have heard multiple counts of organizations investing in document management only to encounter problems down the line due to the lack of a comprehensive plan on how documents should be organized to maximize management and retrieval. There was also a popular belief early last decade that search would solve the “metadata problem.” It is well-known that efforts to enforce proper metadata for documents at the knowledge worker level results in fragmented, inefficient document organization. It’s hard to ask staff to apply precision the way a file clerk or records manager would while they are trying to do their jobs.
Banks must recognize that there is no technological silver bullet; no technology can make all the work go away. A second mistake is to not do the upfront, cross-functional work required to identify document-oriented processes, record the documents involved and then lay out a taxonomy that includes the proper descriptive data used to organize, manage and retrieve document-based information.
Taxonomy and Metadata Plan
It all comes down to taxonomies and metadata. Taxonomy, in the strictest sense, deals with classification. When applied to document-based information, it means identifying and organizing the types of documents that an organization routinely uses to conduct business, whether those documents are used in a supporting role or as the primary deliverable.
In banking, documents play a critical supporting role and it is fairly straightforward to create a taxonomy using cross-functional representatives. Once documents are classified, the next step is to identify all of the relevant descriptors that each department would use to manage and access this information. Again, the best approach is to use a cross-functional team to allow for different departments to include their own specialized descriptors. These descriptors become the metadata that is used in a document or content management system. The greatest challenge is enacting this plan during normal business operations. That’s where technology can be applied, not as the answer, but as an assistant.
Using automated classification technology along with automated data extraction, it is possible to apply a consistent and uniform set of descriptors to each document as it enters an organization. It is important to note that these technologies are not 100%. There are errors and the operationalization of your taxonomy and metadata program must also consider how to handle exceptions. Typically, hard-to-classify documents can be directed to a queue where a subject matter expert can make the correction to be used to update the system. Up to 85% or more of this type of work can be completely automated, leaving the smaller percentage to be handled as an exception. Imagine having 85% of your document-based assets correctly identified and tagged automatically.
With artificial intelligence and machine learning promising a day when humans aren’t required to perform repetitive non-essential tasks, it’s easy to assume that technology alone can solve all our information management problems. However, the answer isn’t a completely manual process nor is it a completely computer-based one; it’s somewhere in-between. The silver lining is that by doing the upfront work to identify and describe all the documents, and leveraging the automation that technology can provide, banks can significantly reduce inefficiencies, improve the customer experience and mitigate compliance costs and risks.
Mr. Council is vice president of marketing and product management for Longmont, Col.-based Parascript. He can be reached at [email protected].