Does this sound familiar? A paper invoice arrives in your accounting department and an employee quickly scans it into the computer, using OCR to capture the full text of the document.
This technology is still used in many document management systems, but leaves much to be desired in terms of process efficiency. Before your team processes the electronic invoice for payment, an employee would still need to manually enter data into several indexing fields.
Manual data entry is expensive and tends to introduce errors, but some sort of indexing process is essential. True, OCR software does capture text in a way that allows users to search for specific documents. The problem is that the data isn’t structured. If you search for
a customer or invoice number, for example, the document you want may be hidden in pages of full text search results
that aren’t relevant, leading to inefficient and frustrating processes.
One solution to this problem is intelligent indexing, in which users “train” your document management system to automatically capture the correct metadata for indexing. This approach eliminates 87 percent of manual indexing tasks, enabling accurate document storage and taking your process efficiency to the next level.
How does intelligent indexing work?
The intelligent indexing process starts when you import a document. The document management software automatically classifies it by type, such as a supplier invoice. Next, the system analyzes the document, searching for the relevant data to capture in previously defined index fields, such as the date, sender’s name, recipient name, etc.
The system then provides you with suggestions for how to index the document. All you have to do is confirm the suggestions or improve them by clicking on the right locations in the document. The system then uses your feedback to “learn” and constantly refine the intelligent indexing process.
The first time you use intelligent indexing for a certain document format, you would probably need to make corrections; for example, the system might select the wrong date for the document. But the accuracy improves dramatically after only two or three samples, so it’s useful for companies that receive small amounts of documents of the same type.
Software that supports “crowd learning” further shortens this initial learning process. This type of system learns how to process documents based on feedback from other users within the same company or from other companies. If a user at another corporation has already indexed an invoice from a specific company, those rules allow more accurate data extraction from the start.
Process efficiency benefits
The main benefit of intelligent indexing is increased speed when inputting data from invoices and other documents. After a quick learning curve, it largely replaces manual indexing, allowing you to streamline processes and store documents far more efficiently than traditional paper filing. A secondary benefit is easy document retrieval, due to the accuracy and thoroughness of this indexing.
Improving process efficiency with intelligent indexing could make a big impact on your bottom line. For example, one company used a document management system with intelligent indexing for 14 months, during which it imported approximately 68,000 documents. These were all analyzed to capture data for five index fields — roughly 340,000 entries — with an average accuracy of 87 percent. While the indexing accuracy varies depending on the type of document, 87 percent is achievable.
That means the company was able to automatically capture 295,800 correct entries, rather than manually entering the data. If you estimate a manual entry speed of 10 seconds per field, the company saved 820 employee hours over 14 months.
Despite these advantages, most document management solutions don’t yet offer intelligent indexing. The technology is relatively new, and new uses are still emerging, such as using intelligent indexing to restructure your existing document archives.
Mobile apps are also available that help eliminate paper receipts for employees on the road — just scan the document with your mobile device and automatically extract the index data and archive it for later use. Whether in the office or on the road, intelligent indexing is a great way to reduce manual data entry and improve your process efficiency.