- Matthias Pfeiler
As you know, part 3 to 7 of our IDP blog series are all about key factors to determine the right intelligent document processing (IDP) Service Provider. In our 5th blog, we compared whether an end-to-end IDP solution or a self-service offering makes more sense, and which individual building blocks need to be linked in a self-service offering. Moreover, we looked at how the billing should be done. Per document or per page. In the following list you will find all the terms that are important for the successful introduction of an IDP system in the company:
Today's blog will cover the factors "price/performance ratio", the "type of OCR" and "value added IDP services".
Price is the least relevant of all the criteria. If the goal is to extract a few pieces of information and header data from always the same document layouts, then a template-based approach is the right one.
Template-based means that, for example, for a company that always has the same layout, software rules can be used to set which information is to be extracted.
However, this approach does not work so well for goods items and a large number of different document/invoice layouts, but it is of course about 30% to 70% cheaper.
For more complex projects, an artificial intelligence (AI)-based OCR is the only solution to achieve efficiency gains. Highly complex projects such as customs clearance also require additional services besides document classification and data extraction, such as data normalization, enrichment, aggregation, validation and data integration. The more complex the project and the application, the more expensive the product, of course.
In contrast, of course, is the price/performance ratio, where compared to competing products in an all-in-one solution, the data is delivered as it should be, end-to-end with highest precision and 99%+ accuracy. This is the case in customs clearance, where e.g. at Digicust the product costs a little more, of course, but also actually saves between 3 to 90 minutes per customs case, depending on the complexity.
The reason of this is that complete customs declarations are sent to the subsequent system automatically, e.g. customs software, and not only extracted data. There is no need to annotate data during the intelligent document processing pipeline, nor to correct anything. Still, it is necessary to quickly control small things in the customs software and eventually add missing customs codes or missing information that have not been in the processed documents or in the master data.
Compared to a self-service product, the efficiency gains for simple customs cases are no longer available. The contrary is the case, where customs brokers double their work when applying a self-service product. The main reason is that the document classification and data extraction process are only augmented by applying AI, but it is far away from being automated due to highly complex customs documents. Additionally, a seven-figures investment is needed for 10.000 declarations only, to develop an end-to-end solution that brings the required value in customs clearance.
As described in the first blog of this blog series, Digicust perfectly supports you in your IDP journey. This way you can quickly benefit from huge efficiency gains from day one and use an automated IDP with little or no implementation effort.
Finally, in our 7th and final part, we will talk about processing time. This part is quite extensive and informative. Hence, a separate blog will be dedicated for this interesting topic.
We hope you enjoyed today's blog and wish you a nice day and stay tuned!
Learn about customs clearance, foreign trade, our product updates and our latest achievements.
While you or your staff are processing customs cases, the Digicust master data module is working in the background, automatically populating with data as you process data with Dexter and serving as the basis for data enrichment during customs clearance.Read More
Document processing time can be more or less important depending on the service delivered and the complexity of the document. In contrast, high precision is important at all times, no matter the complexity or size of the processed document.Read More