Invoice data capture is a must-have solution for many businesses in the finance industry, from online payments, to bookkeeping, to even banks. There are a few viable options to consider when selecting the right data capture solution to meet your requirements.

Exclusive accounting tips and how-tos

Join now and we’ll send you weekly email updates with the latest articles on accounting and bookkeeping.


Invoice elements you need to capture

An invoice is arguably the most common document that businesses typically issue after the transaction happened but before the payment. Learn more about the difference between an invoice and a receipt.

Depending on the country of origin, your invoices may include varying elements that you need to detect or copy in order to digitise the document and prepare it for electronic processing and storage.

The following invoice details are common for all countries:

  • Document type
    You need to know that an invoice is an invoice, and not, for example, a Purchase Order, bank statement, pro forma invoice, or any other document type.
  • Supplier details
    Given that invoices are issued by the company selling goods or services, their details are always put on the invoice.
  • Total amount
    Money owed related to the sale.
  • Date
    All invoices include at least one date so that it’s possible to locate the transaction. If the document doesn’t specify what the date means, you can probably assume that this date indicates when the document was issued or when the sale happened.
  • Product or service description
    All invoices include at least some reference as to what was sold.

And here are other invoice elements which are either optional or country-specific:

  • Invoice number
    A vast majority of invoices include a unique identifier. This number helps you differentiate between documents if there were many transactions of the same kind on the same date. Learn more as to what is an invoice number and how to build one.
  • Tax rates and amounts
    The most common tax name in the EU is VAT, while in the USA a similar tax is referred to as sales tax.
  • Due date
    It’s a date when the payment is expected.
  • Payment method
    This method specifies in what way the payment is expected to happen, often including payment details, such as bank account number.
  • Buyer details
    In the case of VAT invoices issued in the EU, buyer details are required. In other cases, it depends on the country.
  • VAT IDs and other tax IDs
    VAT ID is required on VAT invoices in the EU. In other cases, a VAT ID or a similar number may be required depending on the country.

All other elements vary from country to country depending on the local laws and tradition regarding the way business is done.

For example, in Poland many clients expect business owners to put a stamp and a signature on their invoices, even though it’s not required by law. As a result, many invoices include empty boxes for signatures and company stamps just in case someone wanted to use them.

Invoices often include marketing and sales material too, especially if they’re issued before the sale happens. The invoice’s role in this way is not just to document the transaction but also to convince the buyer to make the purchase.

From an OCR and text analysis point of view, detecting correct invoice elements can be a real challenge.

Invoice data capture – Pros and cons of potential solutions

Capturing all elements from invoices doesn’t have to be a lengthy, resource-hungry process. Here are the potential options to choose from.

  • In-house manual keying team

Manual data entry is often considered as a go-to choice when it comes to digitising paper documents or converting images and PDFs into an editable format.

Manual keying requires an in-house team of data entry specialists. The biggest advantage is probably security: your documents never leave your office. However, there are plenty of downsides to this option, most importantly including the time and error rate.

A typical typing speed for a professional typist is around 50-80 words per minute but given the monotonous aspect of the job, you may expect a high turnover rate for data entry positions.

  • Outsourced manual data entry

There are many third-party providers who offer manual keying for invoices and receipts which are first OCRed. These companies employ data entry specialists whose main responsibility is to copy text from a large number of client documents as quickly as possible.

Examples of such providers are or The biggest disadvantage here is probably the wait time which takes up to several hours depending on the document. Security can be questionable too due to the fact that external data entry employees get access to company documents.

  • Invoice OCR solution

The third option is to rely on an OCR solution that scans and extracts text from images and PDFs. Read our guide to understand how OCR works and what limitations it has.

Generally speaking, a rough OCRed text is a mess that makes it very difficult to locate and extract key information regardless of what you’re looking for. This is especially true for invoices that include a lot of unrelated text, such as billing details from a mobile operator. Your best bet is to use an OCR solution that can analyse OCRed text with algorithms and patterns to only return key info.

In order to do this, the solution can use, for example, the TF-IDF algorithm which allows you to apply weights to keywords depending on their occurrences. Then, knowing the keyword importance, you can add an algorithm based on the simple naive Bayes Theorem to “guess” what category a text belongs to. A sophisticated system of algorithms and patterns can return selected invoice elements with marginal error rate, without the wait time, and without security concerns.

Try an automatic invoice data capture

If you’re looking to test an invoice data capture solution in your organisation, feel free to sign up and give arbitrue a try.

arbitrue processes invoice and receipt files with an OCR and then detects key elements, such as company details, dates, amounts, payment method, and more. To make things even better, for all extracted files you can use our super-fast 1-click export to your accounting records (QuickBooks Online) or spreadsheet files (CSV). Click the button below to sign up for a free trial.

Exclusive accounting tips and how-tos

Join now and we’ll send you weekly email updates with the latest articles on accounting and bookkeeping.