PDF and scanned invoices don’t allow for automated invoice processing. Stone Age anyone?

As you know, the E-invoicing Platform takes a very broad perspective on the definition of an e-invoice. Firstly, because we keep a keen eye on the e-invoicing behaviour of end-users. Secondly, because we believe that the world is too complex to bring it down to a singular e-invoice approach. And thirdly, because we believe that without the use of the proper legislation, a narrow definition of an e-invoice limits the adoption instead of boosting it.

As for the latter it is kind of strange that the European Union aims to liberalise e-invoicing as from 1 January 2012 by law, and at the same time strives for a single structured data e-invoice data. It makes no sense. Why not embrace the heterogeneity that is embedded in liberalisation?

Sorry, let’s get back on track here. We received this years second edition of Billentis’s newsletter. With results of two surveys in Spain.

Asociación Nacional de Empresas de Internet (ANEI)

The National Association of Internet Companies (ANEI) recently published the results of their E-Invoicing survey:

  1. 75% of all Spanish companies receive invoices in a format other than paper.
  2. Of all companies issuing non paper-based invoices, 71.7% use PDFs and 15% XML files.
  3. - e-mail dominates e-invoicing: 78.3%
    - paper mail: 55%
    - message exchange systems: 16.7%
    - presentment/download via website: 6.7%
    - other forms of shipping: 8.3%


Red.es, a public corporate entity attached to the Ministry of Industry, Energy and Tourism, conducts an annual survey aimed at measuring the progress of the Information Society in Spain. The survey’s 2012 edition shows that:

  1. E-invoicing is done by:
    - 60.4% of companies with more than 250 employees
    - 53% of companies with 50-249 employees
    - 41.1% of companies with 10-49 employees
    - 19% of micro-companies
  2. 15.7% of all Spanish companies issue e-Invoices in a structured format; however, a majority of 93.8% issue E-Invoices. And, as is stated, “without permitting automated processing (e.g. as images, or PDFs).”

E-invoices that do not allow for automated processing. Get real.

Here we take the risk of offending some of you. But honestly, anyone who says that “PDFs and images don’t permit automated processing” has either been living under a rock for the last ten years (and shows to remain to do so) and/or acts a stakeholder in favour of structured e-invoices.

PDFs  and images have never been and aren’t an obstacle for more efficient forms of e-invoicing. In fact, the data format argument should never be used to support adoption. If used towards end-users, it backfires by creating uncertainty, actually slowing down adoption.

Furthermore, there are (and have been for many years) numerous service providers and software vendors that allow for scanning and processing these invoices. In fact, there are some fine mobile apps to OCR your invoices just fine.

Related Posts


  1. The purpose of e-invoicing is to facilitate automation. Do you have any idea how easy/complicated it is to OCR PDF e-invoices when they all have different layout and are in different languages. There is a reason why most scanning bureaus interpret roughly only 15 data fields from paper invoices. And the best of these bureaus (those who have SLA’s for accuracy of data) have real people validating that the interpreted data corresponds to the scanned image. Talk about automation…
    This post reminds me from the early days of Internet. Those who could not code well enough in html created their webpages as GIF images. Yeah, who needs structured data or standards anyways…
    Adoption without automation is futile.

  2. e-mails with or without PDFs are not to be counted as e-invoices. Promoting PDFs is seriously delaying migration to e-invoicing. In Finland 64% of enterprises used structured invoices in the spring already. Soon close to 100% - one key reason being that receivers increasingly return both paper and e-mail invoices.

  3. Scanning and OCR have indeed improved a lot in the last 10 years, but I think the main advantage of using a scanning and OCR solution is that it does not disrupt your supplier processes.
    Suppliers simply cannot support all the different electronic data formats, standards and versions that are used by customers. This has and will always be a big drain on supplier adoption and it explains why no company can come even close to 100% adoption.
    I know companies that have several people on their payroll doing nothing but PO flip for their customers (where are we with automation, really?). Not to mention the teams of IT people to implement and maintain all the different integrations with their customers. As the number of companies requesting ‘true’ integration it will become inevitable that suppliers will charge back their increased costs to customers (especially if they also have to pay fees).
    Scanning and OCR solutions come at no added cost for suppliers and can drive true supplier adoption. So I think PDF and image solutions absolutely deserve serious consideration as a long-term solution to get rid of the paper invoicing.

  4. Hello Tom,

    Thank you for your comment. It is good to hear from your experience that massive adoption of e-invoiving is less an “either/or” and more a “connect and drive from current very day practice”

    Nevertheless, I also fancy Bo’s point of view that massive adoption could get a boost by mandating it….

    Interesting to see where those can connect….


  5. Wow. This beggars belief.

    Here we are, all working in the same field, trying to remove data entry from invoice processing.

    And on each side are the vested interests. You all know who you are.

    On one side you shout “No, you can’t use paraffin and a lantern. You all need to remain in the dark until such times as we get electricity cables laid to your door. And at this end there will need to be investment in big electricity plants, and at that other end you’ll need to need to wire your home and invest in fixtures, and fittings, and light bulbs.”

    We need to get people used to having light and not stumbling around in the dark, then market demand will drive infrastructure and investment.

    In the short to medium term, does it really matter how we get people thinking about data automation? Does it really matter that it isn’t in the “purest form”?

    Narrow-mindedness slows change and just leaves room for others to come in to the market. Thanks for holding the door open, but don’t be a Kodak.

  6. I will second what Tom said, and note that Intelligent Data Capture technology can take PDFs and capture, then reconcile against internal sources (such as a PO database), both header and line-item detail, often with no need for personnel to interfere. No conversion to an alternate format, no placing any cost or operational burden on the supplier to adopt, extremely efficient and fast. Right now, one of the world’s largest companies is processing 40% of their invoices, nearly 500,000 annually, via PDF, with much of it going through hands-free. Essentially, the technology presents a quick, comprehensive way to account for the “long tail” of invoices that cannot be processed via the e-invoicing model. The Walt Disney Company recently received an award for taking this approach in their AP operations.