An Obligation from the State, a Contract with a Private Company: Why Do We Need a Digital Postman?

Policy argument about the access relationship—not a claim that every business must buy another subscription.

I support electronic invoicing. Having previously implemented electronic data interchange, or EDI, I see the attraction of information that moves between business systems without somebody retyping it. My objection is not to structured documents or Peppol. It is to treating the allocation of costs, contracts and responsibility as though it were an unavoidable property of the technology.

A digital postman performs useful work: baseline transport itself has commercial value, not just integration or other extras. The dispute is who procures and pays for that work, what a business must agree to, and whether a publicly guaranteed alternative should exist. A required commercial relationship does not automatically imply a paid fee. Those governance choices deserve an explicit argument.

What Slovakia actually asks businesses to do

The Financial Administration’s eFaktúra page says that receiving electronic invoices requires contracting a certified delivery-service provider, and that the service can be used through accounting software or a web or mobile application.2 Its announcement of 3 June 2026 describes a two-stage process: select a provider through the public portal, then confirm activation with that provider, which registers the business in Peppol after agreement to its commercial terms.1

This is not the same as opening a public mailbox and accepting public-service conditions. The official process combines public identity verification with a provider relationship. Equally, it would be wrong to claim that every business must buy a separate, expensive subscription: the same announcement says several solutions offer a free version and describes integration into existing accounting products.1

There is also an important qualification to the word “mandatory.” The Financial Directorate’s FAQ, in the file dated 15 September 2026, describes alternative delivery such as EDI or email with the recipient’s consent, while retaining the structured-invoice requirements; it separately describes the obligation to be capable of using the delivery service.6 The criticism is therefore about the required access relationship and the chosen default architecture, not a claim that every individual invoice must invariably travel through one private channel.

A postman is more than a tollbooth

OpenPeppol describes end users as buyers and sellers exchanging documents through a Peppol-certified service provider.4 Its network work includes messaging, security, addressing and capability lookup.4 These are real functions. Somebody has to operate the connection, maintain compatible software and deal with failures.

That is why “just remove the intermediary” is an incomplete answer. Removing a supplier does not remove its work. A public alternative would have to perform or procure those functions, meet the applicable network requirements and maintain a working route to other participants. It would not be a website that accepts a file and leaves the actual business delivery unresolved.

But the existence of necessary work does not dictate a separate purchasing decision by every regulated business. The state can, as a policy option, procure infrastructure while presenting a public baseline service to its users. That proposal still needs legal, procurement and operational assessment. It is not something I am claiming Slovakia already offers.

A hypothetical firm exposes the choice

Consider a hypothetical small maintenance company. Its owner issues invoices, its external accountant prepares the books, and its customers use several software products. Assume its relevant invoices fall within the electronic-invoicing regime. The company wants dependable delivery and a record it can retrieve later, not a new software project.

An integrated commercial service might be its best choice. The owner could create an invoice inside the existing application, while the provider handles the network connection. If the accountant receives structured records automatically, the subscription might be worth more than it costs. Nothing in my position requires the company to abandon that arrangement.

Now change the circumstances. Suppose the company has little invoice traffic and already has a satisfactory bookkeeping arrangement. A new account creates work: somebody must evaluate terms, assign access, answer renewal notices and check what happens when the accountant changes. Even a zero-price tariff leaves these decisions in place. “Free version available” and “no additional relationship to manage” are different propositions.

The public-policy question is whether this second company should have a reliable baseline route without negotiating an additional commercial service relationship. The example does not establish the answer through a made-up cost estimate. It identifies the work that any honest comparison must count. Compare a new standalone relationship, an extension of existing accounting services and a public baseline. Identity, delegated access, access recovery and export remain in the public version; standard terms could unify conditions, while individual provider selection for the minimum could disappear. Measure those differences rather than crediting the public model with all administrative savings.

What businesses should receive in return

Where government makes a delivery capability necessary, the minimum offer should be legible. I would require a short standard disclosure covering the contracting entity, the certified operator behind any reseller, the functions included, price changes, retention, export and exit. The obligation should not disappear into the phrase “Peppol ready.”

The Financial Administration explicitly allows certified providers to supply services through intermediaries that resell them in their own name.2 This makes the distinction particularly important: the company whose interface a user sees need not be the only organisation involved. That does not imply wrongdoing. It means support, data handling and responsibility need to be explained across the actual service chain.

A useful disclosure would distinguish receipt by the provider, delivery to the business application and the status of any required tax reporting. These are different questions. OpenPeppol separately describes business-document exchange and its work on reporting consumption-tax data to tax administrations.4 A green tick without an explanation is a poor basis for deciding whether an obligation has been met.

The strongest defence of commercial delivery

The best counterargument is not that public provision is impossible. It is that specialist providers already have relevant products and incentives to improve them, while a public application could be slow, inflexible and costly. A competitive market can offer sector-specific features, integration and support that a basic state service should not try to reproduce.

I take that objection seriously. A public baseline could become an expensive second system. If it attracted users through an unlimited subsidy, it might weaken competition without delivering better reliability. Public ownership would not make its software secure, its support responsive or its procurement competent. Those outcomes would have to be demonstrated.

Yet this is an argument for a bounded service and comparative evidence, not for assuming that compulsory market participation is neutral. Government would need to compare the full public cost of each model with businesses’ onboarding, integration, operation and switching costs. Provider revenue alone is not a measure of either social waste or social benefit.

The choice need not be public monopoly versus private monopoly. A limited public channel could coexist with paid automation, richer reporting, reconciliation and contractual service guarantees. Providers could also compete to operate parts of the public channel. Competition would then concern who supplies the best service, not just who captures the compulsory access relationship.

Separate the public purpose from optional convenience

A baseline should, in my proposal, cover creating or importing a compliant invoice, sending and receiving it, readable display, clear status information and usable export. Its retention boundaries should be explicit. It should not silently become a full accounting suite or promise to resolve the commercial truth of an invoice.

That last boundary matters. The Financial Directorate’s FAQ expressly says delivery does not mean the recipient has accepted the invoice’s substance.6 A delivery service cannot decide whether a repair was completed properly or whether a disputed charge is justified. Nor should the state sell the impression that collecting structured data eliminates the need for ordinary business controls.

The same discipline applies to taxation. A public delivery option should not be an excuse to collect every commercially useful field for every administrative purpose. Its design should identify what is needed to deliver the document, what must be reported under the applicable rules and who is allowed to access each part. That is a proposed design requirement, not an allegation that the present system has been proved unlawful.

A decision that can be tested

Before treating the commercial relationship as the only reasonable baseline, government should publish a comparison of the existing model, a state-procured baseline and a directly operated service. Each should have the same functional boundary and the same treatment of support, failure, export and supplier replacement. Otherwise the cheapest-looking option may simply be the one that leaves the most work outside its budget.

Businesses, meanwhile, should ask who operates their connection, which terms they accept, what is included and how they leave. They should not postpone preparation while waiting for a policy alternative that does not yet exist.

A digital postman can be useful and worth paying for. My argument is narrower: when the state makes electronic participation necessary, it should justify why the basic route requires a commercial relationship rather than offer that relationship as a choice for additional value.