How Well Does E-Invoicing Work in Practice

Results of the Nationwide Survey on the Implementation Status of E-Invoicing Among 644 Participants

Formally, e-invoicing has arrived for most users. In day-to-day operations, however, the picture looks different: PDF still dominates incoming invoices, just over half say they have more work than before, and about one in four users regularly encounters business transactions that cannot be mapped in the structured data set of the e-invoice. The key findings of the eurofactura survey at a glance.

Contents

1. What the Survey Was About
2. Key Findings at a Glance
3. Who Responded and How to Read the Figures
4. Formally Arrived, Not Yet in Practice
5. Effort Is Rising Instead of Falling
6. Where the EN 16931 Data Model Reaches Its Limits
7. Neither Company Size nor Automation Prevents Problems
8. A Look Ahead
9. What Users Need
10. Interpretation: What the Results Suggest
11. Conclusion: Where Action Is Needed Now
12. Sources

What the Survey Was About

Since January 1, 2025, companies in Germany have been required to be able to receive e-invoices in domestic business-to-business transactions. Only an invoice in a structured electronic format that complies with the European standard EN 16931, such as ZUGFeRD or XRechnung, qualifies as an e-invoice. The standard defines which information an e-invoice can contain and in what structure. Transitional periods apply to issuing e-invoices and are now expiring in stages: from January 1, 2027, companies with prior-year revenue of more than €800,000 must send e-invoices, and from January 1, 2028, all other companies must follow. In addition, the EU’s ViDA initiative (“VAT in the Digital Age”), adopted in March 2025, has already set the next step after the e-invoicing mandate: for cross-border transactions within the EU, invoice data is to be reported electronically and promptly to the tax authorities. For domestic business-to-business transactions, a national VAT reporting system is also planned: the VAT-relevant data of every single invoice is to be transmitted promptly to a central platform of the tax authorities. Mandatory e-invoicing is the technical prerequisite for this.

How do users experience e-invoicing in their day-to-day work? eurofactura e.V. explored this question in a nationwide online survey. From August 1 to September 30, 2026, companies and organizations that receive, process, or send e-invoices were able to participate.

The Survey Focused on Five Questions:

How far has e-invoicing actually been implemented in receiving, processing, and sending?

Can real business transactions be mapped in the e-invoice format – and how do users deal with it when they cannot?

Where in the process do problems arise, how much effort is involved, and are the expected benefits materializing?

How familiar are users with upcoming developments such as ViDA, Digital Reporting, and Peppol?

What support do they need?

How the Survey Was Conducted and Analyzed

The survey was conducted online, and participation could be anonymous. The questionnaire follows the life cycle of an invoice. A mandatory section with seven topic blocks (about five to eight minutes to complete) asked about the company profile, the adoption and implementation of e-invoicing, formats and transmission channels, preparation for the issuing mandate, problems in the process, the ability to map business transactions, and awareness of upcoming developments: ViDA, digital VAT reporting systems, and possible future transmission channels such as Peppol.

Participants could then decide whether to answer further questions. Just over half completed this optional in-depth section (about five to ten minutes): 241 of 471 users (51.2%). Among other things, it asked about company size, revenue and invoice volume, transmission channels, the causes of problems, effort and costs, archiving, how business transactions that cannot be mapped are handled, and support needs.

Most questions offered fixed answer options, some single choice, some multiple choice; where multiple answers were possible, this is noted with the results. The questions were worded neutrally, with counter-options, symmetrical scales, and the option to answer “don’t know.” In addition, there were free-text fields, mainly in the in-depth section: for example, on the biggest implementation problem, on the causes of the problems described, on examples of business transactions that cannot be mapped, and for final comments.

In total, 434 free-text answers from 179 users are available. They were analyzed using qualitative content analysis based on a fixed coding scheme. For quality assurance, two random samples were coded a second time independently, without knowledge of the first coding. For the central categories – mappability of business transactions and requests for support – the two codings agree very well.

The results were first presented at the 13th ZUGFeRD Developer Days on October 16, 2026.

Who Responded and How to Read the Figures

A total of 644 responses were received. Only the answers of participants who stated that they use e-invoicing themselves were analyzed: 471 responses. The answers of 173 pure solution vendors, service providers, and providers from the e-invoicing sector were excluded from the results. Participants with both roles – who receive or send e-invoices themselves and also offer solutions (82 of 471, 17.4%) – were not excluded. Their answers were included in the analysis as user answers, which is one reason for the sample’s high digital affinity (see below). In addition, 241 users completed the optional in-depth section, and 179 respondents provided free-text answers.

Base: all 644 responses received.

Participants fall into four industry groups: services and the public/social sector (150 of 471, 31.8%), construction and skilled crafts (109, 23.1%), industry and manufacturing (106, 22.5%), and trade and logistics (98, 20.8%). Within these groups, four individual industries are represented by more than 50 participants each: industry (89, 18.9%), trade (86, 18.3%), skilled crafts (80, 17.0%), and IT and communications (74, 15.7%). Company size was only asked in the in-depth section: just under half of the companies there have up to 20 employees (116 of 241), and just over one fifth have more than 250 (53 of 241).

Base: 471 users.

Base: 241 participants in the in-depth section (only asked there).

Participants were recruited mainly through three distribution lists: the ZUGFeRD Community with more than 10,000 e-invoicing users, to specifically reach companies that already have implementation experience; a mailing list of more than 10,000 skilled crafts businesses; and another distribution list of more than 10,000 recipients of digital invoices – including, for example, companies that have so far received invoices as PDFs by email. In addition, the Forum elektronische Rechnung Deutschland (FeRD) called for participation in a news post, heise online published a news article about the survey, and participation was promoted on social media. The sample is therefore a convenience sample: those who took part chose to do so themselves. The results are thus not representative and cannot be extrapolated to the German economy as a whole.

An Experienced Group of Participants

Participants have a strong affinity for e-invoicing. 217 of 471 (46.1%) also invoice public-sector clients and therefore have experience with XRechnung. 82 (17.4%) are both users and providers, and 74 (15.7%) come from IT and communications.

This is also shown by a comparison with the e-invoicing survey of the German Confederation of Skilled Crafts (ZDH). For this survey, the ZDH questioned skilled crafts businesses from January 12 to February 27, 2026, recruited through 52 chambers of skilled crafts in all 16 German states; 1,926 businesses took part. Among other things, it asked about receiving, processing, and sending e-invoices, formats, transmission channels, and archiving, as well as effort, costs, and benefits. The ZDH published the results in May 2026. The comparison shows:

Base eurofactura: all users or all recipients of e-invoices; processing based on self-assessment; effort: recipients in the in-depth section. Base ZDH: 1,926 businesses; shares converted to the respective respondents, hence n ≈.

The gap also exists within the same trade: of the 80 skilled crafts businesses in the eurofactura survey, 53 (66.2%) send e-invoices themselves. The participants in the eurofactura survey are therefore more advanced than those in the ZDH survey. Both surveys are based on self-selection; neither is a benchmark for all users.

This has two implications for reading the results. High implementation rates cannot be transferred to the broader market. And problems that occur even among this experienced group of participants carry all the more weight.

Notes on Figures and Quotes

Not every question was addressed to all participants. Questions on processing incoming e-invoices, for example, were only asked of companies that receive e-invoices. The percentages therefore do not always refer to all 471 users but to a reference base: the number of participants who answered the respective question. For the processing questions, this is 375 recipients; for questions from the optional in-depth section, at most 241 users. The base is given for every value: in the text, for example, as “207 of 375 recipients,” in tables and parentheses as “n = 375.”

Open answers show which experiences users have and how varied they are – but not how many users share a particular experience. After all, those who do not write something down have not necessarily not experienced it. How often a topic appears in the open answers is therefore not given as a percentage but in three fixed levels, each representing a certain number of people: “isolated” (one or two people), “several” (three to nine), and “repeatedly” (ten or more). All quotes come from participants who agreed to their publication. They are anonymized and identified only by industry. Obvious typos, spelling mistakes, and punctuation errors were carefully corrected without changing the wording; omissions are marked with […]. The quotes have been translated from German.

Formally Arrived, Not Yet in Practice

Whether e-invoicing has arrived can be answered on two levels. Formally means: a company receives or sends e-invoices. In practice means: e-invoicing shapes day-to-day work – it accounts for most invoices and is processed further without manual work. On the first level, participants are well advanced; on the second, not yet.

375 of 471 users (79.6%) already receive e-invoices in their regular operations. The vast majority of them, however, do so only in part, alongside PDF and paper invoices (298 of 375). This is the permissible norm. The share is more revealing: 207 of 375 recipients (55.2%) still receive mainly PDF or paper, and only 48 (12.8%) mainly e-invoices.

Base: 471 users.

Based on their own assessment, 122 of 375 recipients (32.5%) process incoming e-invoices mainly automatically. 134 (35.7%) work partly automatically, partly manually, and 114 (30.4%) mainly manually – for example, by retyping data or using the PDF view.

Participants are further ahead when it comes to sending. 320 of 471 users (67.9%) send e-invoices themselves, and 197 of them (61.6%) even do so predominantly. Among companies that both receive and send, the outgoing side is usually further ahead than the incoming side (148 of 255, 58.0%). 288 of 471 users (61.1%) consider themselves fully or largely prepared for the issuing mandate from 2027/2028; 46 (9.8%) have not yet started. From 2027, the issuing mandate initially applies to companies with prior-year revenue of more than €800,000. Among participants in the in-depth section, 150 of 240 (62.5%) are above this threshold, compared with 44.0% of skilled crafts businesses in the ZDH survey (n ≈ 1,926).

Base: 471 users; don’t know / no answer: 5 (1.1%).

Base: 471 users.

Formats and Transmission Channels

ZUGFeRD/Factur-X dominates among the formats, followed by XRechnung:

Base: recipients or senders of e-invoices; multiple answers possible, so totals exceed 100%.

E-invoices are almost always transmitted by email: 97.0% of recipients and 95.8% of senders in the in-depth section use this channel (n = 199 and 168, respectively). 13.1% send via the Peppol network.

When it comes to formats and transmission channels, participants in the eurofactura survey hardly differ from the skilled crafts businesses in the ZDH survey. In the ZDH survey, too, ZUGFeRD (75.9%) dominates sending ahead of XRechnung (45.2%, each n ≈ 692), and e-invoices are almost exclusively received (97.7%, n ≈ 1,841) and sent (94.9%, n ≈ 706) by email. Peppol plays practically no role there (receiving 0.3%, sending 1.5%).

In the open answers, participants repeatedly criticize the variety of formats, version changes, or the hybrid format combining PDF and XML. They also repeatedly describe problems with transmission channels, such as customer-specific portals and platforms:

“Because of the different formats and details that many customers want, it is no longer possible to create invoices uniformly. Same field, different customer requirements. Public-sector clients have too many portals that are supposed to be used. […]”(Participant from the trade sector)

“The format is more or less settled. But the transmission channel is terribly chaotic. Some large corporations want their own platform to be used. Public-sector clients as well. […]”(Participant from the skilled crafts sector)

Effort Is Rising Instead of Falling

E-invoicing is supposed to simplify processes. For the majority of participants, this effect has not yet materialized. The in-depth section asked about the change compared with the time before the switch, that is, compared with paper and PDF invoices:

Base: 238 participants in the in-depth section.

The in-depth section also asked about the costs of the switch, both one-time and recurring annual costs. About one third report one-time costs of more than €5,000 (79 of 237), and just under one quarter cannot quantify the costs. In total, 148 participants in the in-depth section specified a cost range for one-time or recurring costs, such as “up to €1,000” or “€5,000 to €25,000.”

Of the 139 among them who also answered the questions on workload and financial impact, 102 (73.4%) see no relief, either financially or in terms of workload. Higher effort predominates most clearly in creating and sending invoices (54 of 167 senders in the in-depth section, 32.3%). In further processing of incoming invoices, higher (27.0%) and lower effort (33.2%) are roughly balanced (n = 196).

For the skilled crafts businesses in the ZDH survey, the balance is even less favorable. Around one in two of them report higher effort than with PDF invoices, both when receiving (49.6%, n ≈ 1,841) and in further processing (51.2%, n ≈ 1,812). In the eurofactura survey, the figures are 24.9% (49 of 197) and 27.0% (53 of 196), respectively. 60.3% of the invoicing skilled crafts businesses rate the benefits of e-invoicing as low or rather low (n ≈ 707). The skilled crafts businesses in the ZDH survey are therefore considerably more burdened than the participants in the eurofactura survey.

The open answers confirm this picture. Participants repeatedly comment on additional effort, costs, or a lack of benefits; only a few, by contrast, report benefits they have experienced.

“E-invoicing is supposedly meant to reduce working time because everything is automated. The reality is the exact opposite, because the workload has increased. […]”(Participant from the trade sector)

But there are other experiences as well. Participants repeatedly report positive experiences or problems that have since been solved:

“XRechnung isn’t a bad thing – once you’ve understood it, it saves time and is easy to use. […]”(Participant from the skilled crafts sector)

Where the Process Gets Stuck

Participants were asked at which points in the process problems occur. This refers to the stages an e-invoice passes through within a company: on the incoming side, receipt, visualization (display in readable form), validation (checking for compliance with the standard), conversion into other formats, and processing in the company’s own system; on the outgoing side, creation and sending; plus archiving. Those who named at least one stage were also asked which of them causes the most effort.

Base middle column: 471 users; multiple answers possible. Base right column: 334 users with at least one problem area; one answer; 13.8% cannot clearly identify the biggest source of effort.

335 of 471 users (71.1%) name at least one point in the process where problems occur; 109 (23.1%) have no significant problems. The most frequently named stages are validation (33.5%), creation (31.4%), and processing (29.7%).

Creation, validation, and processing also form the leading group when it comes to the greatest effort.

What exactly goes wrong at the individual stages was explored in the in-depth section: those who had marked a stage as a problem area could specify the causes (multiple answers possible) and describe them freely. A clear picture emerges for the three main stages:

Base: participants in the in-depth section who named the respective stage as a problem area; multiple answers possible. Because the bases differ, the values can only be compared within a stage.

In processing and creation, the problems each center on one main cause: data that does not flow automatically into the system, and content that cannot be mapped. In validation, two causes are on a par, and both concern how understandable the validation is: error messages are hard to interpret, and it remains unclear whether an invoice meets the requirements. At the other stages, the answers are more widely distributed. In receipt, for example, invoices are technically unreadable or mandatory information is missing (58.9% each, n = 56); in sending, recipients cannot or will not accept the e-invoice, or the transmission channel is unclear (59.4% each, n = 32).

Validation: Often Checked, Rarely Documented

Around two thirds check e-invoices for compliance: 242 of 369 recipients (65.6%) check incoming invoices, and 215 of 311 senders (69.1%) check outgoing invoices. Conversely, just under one third of senders do not check their own invoices or do not know whether they do.

Those who check usually check the content as well: 97 of 236 participants in the in-depth section (41.1%) regularly also check business rules such as mandatory fields and arithmetic consistency. The validation itself, however, is rarely documented. Only 43 of 238 (18.1%) keep the validation report.

Base: incoming validation only recipients (n = 369), outgoing validation only senders (n = 311), other values participants in the in-depth section. Business rules means mandatory fields and arithmetic consistency.

Validation itself becomes a source of conflict. In the open answers, participants repeatedly report contradictory, overly strict, or incomprehensible validation results and rejected invoices.

“Validation […] often too strict. Instead of adding warnings to an e-invoice, some are rejected outright. This leads to an extremely high workload (discussions with the invoice issuer about whether it is correct or not; the invoice issuer cannot issue it any other way because of their software) – usually we then resort, in breach of the rules, to the PDF part of the invoice and process that.”(Participant from the trade sector)

“According to the validator, incoming invoices are not compliant, but the suppliers don’t know any better. What do we do from 2027 on? […]”(Participant from industry)

Where the EN 16931 Data Model Reaches Its Limits

A central question of the survey was whether a company’s own invoice content can be fully mapped in the structured data set of the e-invoice. It was asked of all 471 users. 231 of them (49.0%) answered yes. 178 users (37.8%) report that certain business transactions cannot be mapped or can only be mapped partially; 62 (13.2%) cannot judge. Looking only at the 320 users who create and send e-invoices themselves results in a similar figure: 110 of them (34.4%) encounter such limits.

These are not rare exceptions. Of the 178 users with mapping limits, 118 (66.3%) say that such cases occur regularly or frequently. Relative to all users surveyed, that is just over one in four. Limits are reported in all industry groups.

Above all, typical content of business-to-business transactions is affected:

Base: 178 users with mapping limits; multiple answers possible. The first three form a leading group, not a fixed ranking.

In the open answers, these cases become concrete. Repeatedly mentioned are attachments, proof of services and hours worked as well as delivery notes, industry-specific additional information, advance, partial, and final invoices, corrections and credit notes, line-item structures and collective invoices, as well as special cases related to tax and pricing law. Several participants mention cash discounts, rebates, and bonuses.

“The biggest problem is that the cash discount amount and cash discount date can currently only be transmitted as free text. […]”(Participant from the skilled crafts sector)

Some of those who describe mapping problems in their open answers explicitly attribute the cause to EN 16931 or the format. This is the participants’ assessment, not proof of the cause.

“[…] missing standardized data fields that comprehensively cover typical invoice content. It is standardized and developed for tax authorities, not as a practice-oriented complete solution for companies (focus on B2B).”(Participant from the finance and insurance sector)

“[…] In addition, the standard prescribed by the legislator is often not sufficient to map the diversity of business processes. […]”(Participant from the skilled crafts sector)

When the Data Model Does Not Fit, Users Find Workarounds

How do companies deal with cases that cannot be mapped? The question was asked of participants in the in-depth section who had reported mapping limits:

Base: 97 participants in the in-depth section with mapping limits; multiple answers possible.

61 of the 97 (62.9%) use at least one workaround outside the standard: they forgo the e-invoice in these cases, add a PDF or attachment, or switch to a different format. Where an e-invoice is mandatory, an invoice in any other form legally counts only as an “other invoice” (sonstige Rechnung); the recipient’s input tax deduction may then be at risk. The base is admittedly small. The pattern, however, is independently confirmed: in the open answers, participants repeatedly describe workarounds on their own initiative – from falling back on PDF and entering information in free-text fields to sending the invoice twice, as a PDF and as an e-invoice.

“I always add the working hours on the individual days and what was done to my invoices. Since XRechnung does not provide for this, I always send the invoice both as a PDF and as an XRechnung.”(Participant from the IT sector)

Recipients repeatedly describe the flip side: fields are filled differently depending on the issuer, which makes automated further processing more difficult.

“[…] Mandatory fields are filled in only rudimentarily and all information is supplied as free text (note), which makes automated evaluation impossible.”(Participant from the trade sector)

Neither Company Size nor Automation Prevents Problems

An obvious explanation would be that small companies with few resources or companies that work manually are the ones having difficulties. The data does not support this.

Company size: Companies with more than 250 employees report mapping limits considerably more often (35 of 53, 66.0%) than smaller companies (around 30 to 39%, depending on size category). A plausible explanation is more complex business transactions in large companies. Company size was only recorded in the in-depth section.

Base: participants in the in-depth section who specified company size (n = 236).

Automation: Those who process incoming e-invoices mainly automatically name at least one problem area only slightly less often (80 of 122, 65.6%) than those who work mainly manually (85 of 114, 74.6%). The gap is not pronounced. In terms of effort, however, there is a clear difference: of the recipients in the in-depth section who work automatically, 27 of 64 (42.2%) have more work than before; of those who work manually, 48 of 62 (77.4%).

Base: recipients of e-invoices who provided information on processing.

Automation therefore reduces effort, but it does not eliminate friction.

“[…] We have been delivering only e-invoices since 2025, with little extra effort thanks to automation – but we now have to handle a large part of our revenue via customer-specific portals and therefore largely manually. That makes our workload ‘explode.’”(Participant from the IT sector)

Where Users Locate the Causes of Problems

In the open answers, participants locate problems almost exclusively outside their own company: with suppliers, with software vendors and service providers, with legislators and public administration, and in the standard. Considerably fewer participants name their own company than any of these.

Base: 179 users with free-text answers; only explicit attributions, several per person possible.

Participants repeatedly describe problems that arise with business partners: suppliers, for example, send incorrect or incomplete e-invoices or are unaware of the requirements, while customers impose special requirements such as their own portals or additional fields.

“Often incorrect invoices and no understanding on the part of the issuer. […]”(Participant from the IT sector)

“Software vendors underestimate the task. As a result, companies are left helpless.”(Participant from the IT sector)

“There are no centrally managed validation artifacts determined by the authorities. This leads to the most varied interpretations by those who validate and to a veritable market – or zoo – of solutions. That doesn’t help. The legislator urgently needs to step in here.”(Participant from the IT sector)

This attribution shows how users perceive the situation – not where the causes actually lie.

A Look Ahead

The issuing mandate does not mark the end of the development. In March 2025, the EU adopted the ViDA reform package (“VAT in the Digital Age”). It is primarily intended to curb cross-border VAT fraud, such as carousel fraud. At its core is a digital reporting system (Digital Reporting): from July 1, 2030, invoices for B2B transactions between companies in different EU member states must be issued as e-invoices, and their data must be reported promptly to the tax authorities. For domestic transactions, Germany is planning its own transaction-based reporting system, through which invoice data is to be transmitted promptly to a central platform of the tax authorities; mandatory e-invoicing is the technical prerequisite for this. There is not yet a legally defined start date for this national reporting system (as of October 2026); the system is currently being developed by a joint federal and state working group and must be aligned with the EU system by January 1, 2035, at the latest. Participants are hardly paying attention to these developments:

Base: 471 users; self-assessment of own level of knowledge.

Preparation is correspondingly low. Only 65 of 471 users (13.8%) consider themselves fully or largely prepared for future digital reporting obligations. 232 (49.3%) have not yet started, and 108 (22.9%) cannot judge. By comparison, 61.1% consider themselves prepared for the issuing mandate.

Base: 471 users.

Awareness is also unevenly distributed: it is considerably lower in the skilled crafts, among companies operating only domestically, and among smaller companies. In the open answers, the reporting system and ViDA play practically no role.

“Considering that e-invoices are to be sent to the tax offices in just a few years, very little is currently known about it. […]”(Participant from the IT sector)

The level of information on the provider side is also low. In June 2026, the Verband elektronische Rechnung (VeR), the association of e-invoicing software providers, conducted a sentiment survey of the German economy on e-invoicing and the digital reporting system: Stimmungsbild der deutschen Wirtschaft zu E-Rechnung und digitalem Meldesystem (in German). Its 143 participants come from the provider side, public administration, consulting, and companies; the results are not broken down by role. 99 of them (69.2%) feel rather poorly informed or not informed at all about the planned national reporting system.

Peppol: What Users Want, What Providers Expect

As for the future transmission channel, 335 of 471 users (71.1%) would like email, and 57 (12.1%) Peppol. 55 (11.7%) already use Peppol or have concrete plans to do so.

The provider side expects the opposite. In the VeR survey – an assessment from the solution provider community – 99 of 143 respondents (69.2%) expect Peppol to prevail as the transmission channel, and only 36 (25.2%) expect email to. The providers’ forecast is thus almost the exact opposite of what users want.

Users: eurofactura survey. Provider side: VeR survey, June 2026, 143 questionnaires; forecast of which transmission channels will prevail.

The contrast is put into perspective when knowledge is taken into account. Those who, by their own assessment, know Peppol well mostly want it as a transmission channel (33 of 57, 57.9%) and mostly use it or plan to (40 of 57). Of those who have only heard of it, 10.2% want it (n = 235). Those who have never heard of it do not want it at all (n = 179). The preference for email is therefore particularly widespread where Peppol is little known; those who know Peppol are considerably more open to this transmission channel.

Base: 471 users; self-assessment of knowledge of Peppol.

What Users Need

Participants were asked what would help most with implementing e-invoicing. Up to three answers were possible:

Base: 392 users who answered. Because a maximum of three answers was possible, a low value indicates a lower priority, not a lack of need.

Practical information, better software, and clearer legal requirements form a leading group at 43 to 45%. In the open answers, too, “knowledge, information, and regulation” is the most frequent topic overall. Participants repeatedly describe unclear or missing requirements, a lack of knowledge among business partners and software vendors, and their own uncertainty.

“In my opinion, many companies are not properly trained. Furthermore, some software manufacturers are not yet as far along as they should actually be shortly before the e-invoicing mandate. […]”(Participant from industry)

“You’re pretty much left on your own, you don’t know exactly what is required and how to implement it in the company. […]”(Participant from industry)

The wishes expressed focus above all on uniformity, binding requirements, and better information. Several participants would like an official, free validation tool and clear, binding documentation of which information belongs in which field of the e-invoice.

“Clear, reliable, and understandable documentation of the requirements for filling in the fields, with exceptions and examples, would be very important. I consider a central validation tool provided by the legislator for testing the formats to be indispensable.”(Participant from the services sector)

The provider side reaches a similar conclusion from a different angle: in the VeR survey, a lack of knowledge in companies is considered the biggest challenge (80 of 143, 55.9%). Both sides thus see an information deficit but locate it differently: the provider side sees it in companies, while users, in their open answers, also see it among software vendors and business partners.

Interpretation: What the Results Suggest

The eurofactura survey shows how problems affect day-to-day practice. It cannot prove what causes them. However, hypotheses can be derived from the results. They are consistent with the findings and indicate what should be examined further.

Hypothesis 1: Some Problems Arise Because the Standard Was Not Designed for the Diversity of B2B Transactions

The European standard EN 16931 was developed under Directive 2014/55/EU for electronic invoicing in public procurement. In Germany, it is applied to all business-to-business transactions.

Supporting evidence: Above all, typical B2B content is affected. The limits occur regularly, in all industries, and more often in large companies with diverse business transactions and high invoice volumes. Automation makes no clear difference to whether problems occur. Some of those affected explicitly attribute the cause to the standard or the format in their open answers.

Hypothesis 2: Unclear Requirements and the Lack of a Binding Reference Lead to Inconsistent Interpretations

Supporting evidence: 62 of 471 users (13.2%) cannot judge whether their invoice content can be mapped. Clearer legal requirements are among the top needs. Around one in four participants in the in-depth section with mapping limits resolves cases through bilateral agreements. In the open answers, unclear requirements, contradictory validation tools, and the lack of an official reference are repeatedly mentioned. Equally often, participants describe fields being filled differently depending on the issuer.

Conclusion: Where Action Is Needed Now

Among participants in the eurofactura survey, e-invoicing has formally arrived, but not yet in day-to-day practice. PDF still dominates incoming invoices, automated processing is the exception, and for the majority, effort has increased rather than decreased. Just over one in four users regularly encounters business transactions that cannot be mapped in the structured data set and then resorts to PDFs, attachments, or free-text fields. All of this applies to a group of participants that is more advanced than the skilled crafts businesses in the ZDH survey. With the issuing mandate from 2027, many companies that are not yet as far along with implementation will join them.

Action Is Needed Above All in Three Areas:

Mappability: Typical content of business-to-business transactions – industry-specific additional information, explanations, references to orders and delivery notes, attachments, advance and final invoices – must be mappable in the structured data set without companies having to resort to workarounds outside the standard.

Binding requirements: Users need a uniform reference for which information belongs in which field and when an invoice is considered compliant, including a freely available official validation tool.

Software and information: Solutions must implement the requirements completely and in a practical way. And users as well as their business partners need understandable, practical information, also with regard to ViDA and the planned digital VAT reporting obligations.

eurofactura e.V. Addresses These Areas with Its Own Offerings:

Industry Centers: In Industry Centers, users, representatives of industry associations, and providers are to jointly develop industry-specific implementation recommendations – for example, on how typical business transactions in an industry can be mapped in the data set and which information belongs in which field. This is intended to make bilateral agreements unnecessary, standardize workarounds for mapping problems, and thus reduce workload and manual intervention.

Quality seal for e-invoicing software: An independent quality seal or certification is intended to show users which solutions implement the requirements completely and in a practical way.

Open source: eurofactura promotes open-source e-invoicing solutions that are freely available to all users.

eurofactura Academy: Training courses offered by the eurofactura Academy provide practical knowledge about e-invoicing – for everyone who creates, receives, or processes e-invoices.

In addition, the association contributes the survey results to its exchange with standardization bodies, public administration, and software vendors and advocates for practical experience to be taken into account in the further development of EN 16931, the formats, and the requirements.

Sources

• eurofactura – Verein zur Förderung der E-Rechnung in Europa e. V. (2026): Survey “Stand der E-Rechnung” (State of E-Invoicing). Online survey, August 1–September 30, 2026; 644 responses, of which 471 users analyzed.
• Zentralverband des Deutschen Handwerks (ZDH) (2026): Umfrageergebnisse E-Rechnung. Survey January 12–February 27, 2026; 1,926 businesses. Results overview (PDF, in German)
• Verband elektronische Rechnung (VeR) e. V. (2026): Umfrageergebnisse zum digitalen Meldesystem; 143 questionnaires. Stimmungsbild der deutschen Wirtschaft zu E-Rechnung und digitalem Meldesystem (PDF)
• Directive 2014/55/EU of the European Parliament and of the Council of 16 April 2014 on electronic invoicing in public procurement, OJ L 133, 6.5.2014, p. 1.
• Council Directive (EU) 2025/516 of 11 March 2025 amending Directive 2006/112/EC as regards value added tax rules for the digital age (ViDA).

Conversion of comparative values: The ZDH reports its percentages based on all 1,926 participants, including those who did not answer a question. For the comparison, these values were converted to the businesses that answered the respective question (published share divided by the share of respondents). The number of these businesses is therefore only approximately known and is marked with “≈.” For the VeR survey, the shares were calculated from the published case numbers, for example 99 of 143 questionnaires.