How financial services companies can bring the value of hidden information to light
By James Adie, VP of Sales, EMEA, Ephesoft
Every organisation that keeps a record of its customers is handling valuable information. Nobody would question that customer data is at the heart of effective marketing; the more we understand customers, the better we can sell to them. Yet much of this information remains in the dark. Ironically, the financial services industry, which for regulatory reasons gathers more information about its customers than almost any other, frequently allows data to remain tucked away in paper and electronic files where it never sees the light of day.
At the heart of this problem is a simple process that traditionally made it difficult to retrieve the hidden data. Most information from customers and suppliers arrives in a “flat” format; i.e. as a document. An invoice may be electronic, but it’s still just a pdf file – and unsearchable. A customer’s bank statement or personal ID is likely to be a scan of a paper copy. What these documents have in common is that they cannot be stored and analysed in the same way as “multi-dimensional” data held in a database. Extracting any useful information from either often entails a human reading the details and manually typing in any data that is to be used in future – whether that’s a supplier name, a customer address or a date of birth.
This is not only inefficient and time-consuming, but it means that swathes of data will be lost. The remainder of the information will simply stay within the document, hidden away inside a bank or insurance company’s electronic filing systems; unusable and occupying valuable storage space. IDC described this in a recent report, “Data, in the absence of meaning and context, is worthless and costly.”
Bringing this information to light does not require rocket science, but artificial intelligence (AI) can help. It begins with changing how information is received into the company. Intelligent document processing (IDP) technology is able to evaluate all documents from various input channels, extract the essential information and automatically classify it for further processing – no matter what format or layout. Even handwritten information can be captured. Manual checking is only required in the occurrence of stains, smeared writing or any other issue that causes ambiguity.
Using artificial intelligence, IDP software learns to identify specific fields on different types of documents, such as the account number on a bank statement or the name and address on a utility bill. It can also be trained to recognise all of the fields on an invoice, such as payment discounts, that need checking against agreed terms.
The more it learns, the lower the error rate, the faster the processing, and the higher the volume of usable data. Fewer people are required to oversee the input of documents; during customer onboarding in the financial services sector, for example, we’ve seen increases in productivity of up to 400%. The average improvements in processing time for documents range from 50% to 80% across the board.
By extracting the data from documents as they are processed and storing this with context, financial services organisations have already begun to transform a stack of flat images into a multi-dimensional treasure trove of accessible data. This is opening up a wealth of new marketing opportunities in addition to maintaining compliance and increasing the customer experience.
Banks and insurance companies are starting to wake up to this type of automation, but many of the fintech disruptors already have. To compete with the newcomers, the financial services sector needs to move fast to bring AI into its document handling processes, not only to save time and money but to make sure that none of that valuable information is lost. Dark data is, quite literally, a waste of space.