The difference between an e-invoicing programme that pays back in a year and one that never pays back at all is almost never the platform. It is the share of invoice volume that actually arrives electronically and processes without a human touching it — and that share is set by onboarding strategy. Twenty years of market observation produced a consistent picture: convincing trading partners one at a time gets a large organisation to roughly a quarter of its invoice volume after several years, while closing an electronic loop around purchase orders and invoices can convert the PO-based portion of the supplier base within months.
What has changed since that analysis was written is that mandates now do part of the work. What has not changed is that they do the easy part. A mandate compels the channel and the syntax; it does not compel a supplier to quote your purchase order number, use the unit of measure your ERP expects, or bill the entity that placed the order. In mandated markets, the onboarding question shifts from will they send electronically to will they send data you can match — and that is a harder question with fewer levers.
Four onboarding strategies and what each delivers
The e-invoicing analyst Bruno Koch grouped the approaches he had observed into four, with markedly different adoption curves.
| Strategy | Mechanism | Observed outcome |
|---|---|---|
| Classic opt-in | Persuade trading partners individually to join the electronic community, starting with the largest. | Around 25–30% of invoice volume after several years, and stalling there. |
| Pressing | Sustained marketing plus contractual terms that oblige trading partners to move to electronic exchange. | At least 60% after three years for most large companies. |
| Powerplay | Pressing plus financial penalties: electronic declared the default channel, paper surcharged or deducted. | At least 80% after three years. |
| Closed e-order and e-invoice loop | Send purchase orders electronically; the supplier returns the invoice on the same channel. | The PO-based portion — often 40% or more of volume — convertible within months. |
Two caveats before anyone builds a plan on those numbers. They come from observation of European markets in the period before mandates existed, so they describe voluntary adoption dynamics rather than compliance-driven ones. And they measure the electronic share of invoice volume, not the touchless share — a distinction that turns out to matter more than the headline figure.
Why opt-in stalls at a quarter of volume
The classic approach fails for a structural reason rather than an execution reason. A large buyer starts with its largest suppliers because they are the most capable, and that population is finite. Once exhausted, the buyer turns to mid-sized and small suppliers and discovers that it does not have the market power to move them alone — adoption there depends on the maturity of the mass market, not on the buyer’s marketing budget.
The supplier’s incentives explain the rest. For a small business, the driver to adopt was rarely its own business case; it was the strength of the customer relationship. Keying invoices into the web portal of each large customer, one portal per customer, was unpopular enough that many suppliers held on to paper for as long as they were permitted to. Keying into a single independent service provider’s portal that reached several customers at once was better received. The clear favourite, where buyers would accept it, was simply emailing a PDF.
That preference is worth remembering, because it explains why so much of the “electronic” volume achieved under voluntary adoption was digital images of paper — a format that delivers most of its benefit to the sender and very little to the receiver.
Pressing and powerplay, and what mandates did to them
Pressing moved the lever from persuasion to contract: general terms and conditions amended to make electronic exchange an obligation rather than a request. Powerplay added money. In the mature form, suppliers charged consumers roughly €1–3.50 and business customers €5–25 for issuing a paper invoice, while buyers deducted typically €15–25 from the amount paid on a paper invoice received from a supplier unwilling or unable to send electronically.
Both remain instructive, but the deduction mechanism should be approached carefully today. Unilaterally reducing the paid amount on a valid invoice sits uncomfortably with the late payment regime: under Directive 2011/7/EU, a creditor becomes entitled to statutory interest and a fixed recovery sum once the payment period expires on the amount due, and a disputed deduction is a fast route to that outcome. A surcharge agreed in the contract is a different instrument from a deduction applied on receipt.
The larger point is that mandates have nationalised the powerplay. Belgium moved its entire domestic B2B population onto the Peppol network in a single step at the start of 2026; Poland is phasing its KSeF clearance platform through 2026 by turnover band; France requires all businesses to be able to receive from September 2026, with issuance phased by size; Germany has required reception capability since the start of 2025 with issuance phasing through 2028. Italy has run mandatory B2B clearance since 2019. In those markets, no buyer needs to invent a penalty. The state has supplied one, and it is more effective than anything a procurement department could negotiate. Mandate details change frequently, so verify current scope and dates against the national tax authority before planning around them.
What the mandate does not do
Here is where the ROI leaks. Under a mandate the cost side of the business case moves to 100% almost immediately — per-invoice service provider fees, historically in the range of €0.20–0.80, are incurred on every document whether or not it processes automatically. The benefit side does not follow, because automation savings depend on straight-through processing, and straight-through processing depends on data the mandate does not require.
Specifically, a mandate-compliant invoice can still arrive without:
- a purchase order number, or with one in a field your system does not read;
- line-level detail that maps to the order lines, rather than a single summarised line;
- units of measure and quantities consistent with how the goods were ordered and received;
- the correct legal entity and delivery location, where a group has several;
- a buyer reference or cost-centre code needed for routing and approval.
None of these is optional in the sense of being unimportant, and all of them are optional in the sense that EN 16931 permits an invoice to be valid without them. The result is a familiar failure mode in a new costume: compliant XML that lands in the same exception queue that scanned paper used to. Where automation savings of 60–80% and payback inside two years are achievable, they are achievable on the touchless portion, not the compliant portion.
The gap between those two numbers is the real onboarding problem after 2026, and it is why the historic taxonomy still matters. Its object has simply changed.
The closed loop is the one lever mandates don’t pull
No European mandate requires you to send purchase orders electronically. That makes the closed loop the only strategy on the list that remains fully within your control — and it was already the fastest.
The mechanism is a genuine exchange rather than an imposition. Suppliers want orders. If the only way to receive them is electronically, adoption of the channel is rapid, because the supplier is gaining something rather than absorbing a compliance cost. Returning the invoice on the same channel is then marginal effort. In many large organisations at least 40% of invoice volume is already PO-based, and that proportion has been rising.
The reason it produces better ROI than any pressing or powerplay campaign is that it fixes data quality at the source instead of inspecting it on arrival. An invoice generated from an electronic order the supplier received carries the order number, the order lines, the agreed units and prices, and the correct buying entity — because those values were transmitted to the supplier in structured form and can be echoed back. Three-way matching stops being a reconciliation exercise and becomes a confirmation.
The infrastructure for this is already in place wherever a Peppol mandate is. The network was built for public procurement, and the same community that maintains the invoice specifications maintains specifications for orders, order responses, catalogues and despatch advices, with further profiles for transport documents. If your suppliers are reachable for invoices, they are addressable for orders on the same connection.
Segment before you choose a strategy
The four strategies are not alternatives to pick between; they are tools that suit different parts of a supplier base. A workable segmentation runs on two axes: whether the relationship is PO-based, and whether the supplier falls inside a mandate you are subject to.
- PO-based, inside a mandate. Closed loop. The mandate has already solved the channel, so spend the effort on sending orders electronically and on matching tolerances. This segment produces almost all of the automation benefit.
- Non-PO, inside a mandate. Expense, utility, professional services and similar invoices. The channel is solved but there is nothing to match against. The lever here is coding and approval routing rules, not onboarding.
- PO-based, outside a mandate. Cross-border suppliers and jurisdictions still voluntary. The historic playbook applies in full, and the closed loop is still the best of it. Contract terms are the practical instrument.
- Non-PO, outside a mandate. The long tail. Accept that this segment will not deliver ROI, keep the handling cheap, and revisit as cross-border digital reporting arrives from 1 July 2030 under Council Directive (EU) 2025/516.
Sizing each segment by invoice count rather than by spend is the step most programmes skip. Effort follows document volume; savings follow document volume; spend concentration tells you almost nothing about where the processing cost sits.
What to measure
Four metrics separate a programme that is compliant from one that is paying for itself.
- Electronic share and touchless share, reported separately. Under a mandate the first will approach 100% and tell you nothing. The second is the business case.
- Exception rate by cause. Split missing or unreadable order references, line-level mismatches, unit and quantity mismatches, price variances within and outside tolerance, and genuine commercial disputes. Only the first three are onboarding problems.
- Electronic order share of PO-based volume. This is the closed-loop coverage figure, and it is usually far lower than the invoice figure.
- Invoices processed per accounts payable full-time equivalent. Manual accounts payable operations historically handled somewhere between 5,000 and 15,000 invoices per person per year; the best-automated operations reached well above 100,000. The ratio between your figure and those benchmarks is the honest measure of how much automation you have actually bought.
Reviewing his own data, Koch’s conclusion was blunt: “it is surprising that more organizations do not switch to more promising on-boarding methods.” That was written when the alternative to opt-in required a buyer to spend political capital on its suppliers. Mandates have removed that excuse for the channel. They have not removed it for the order.
Frequently asked questions
Does an e-invoicing mandate remove the need for a supplier onboarding programme?
It removes the channel adoption problem within its scope and leaves the data quality problem untouched. A supplier can be fully compliant and still send invoices your system cannot match automatically. Onboarding after a mandate means agreeing order references, line structures, units of measure and entity identifiers — not persuading anyone to go electronic.
Why does opt-in onboarding plateau?
Because a buyer’s leverage is concentrated in a small number of large suppliers, and those are exhausted early. Beyond them, adoption depends on general market maturity rather than on any single buyer’s efforts, which is why voluntary programmes historically settled around a quarter to a third of invoice volume and grew slowly from there.
Can we charge suppliers for sending paper invoices?
Surcharges and deductions were both used during the voluntary era, typically €15–25 deducted by buyers per paper invoice received. Deducting from the amount due on a valid invoice is legally exposed, since it can trigger statutory late payment interest and fixed recovery compensation. A charge agreed in advance in the contract terms is the safer construction, and in mandated markets it is largely redundant.
What makes the closed loop faster than the other approaches?
The supplier gains something immediately. Receiving purchase orders electronically is a benefit to the supplier rather than a cost, so the channel is adopted quickly, and returning the invoice over the same connection is a small additional step. It also produces better data, because the invoice can echo values the supplier received in structured form rather than re-keying them.
How much of our invoice volume is realistically PO-based?
In large organisations the figure has commonly been at least 40% and rising, but it varies enormously by sector — manufacturing and retail run far higher than professional services or public bodies. Measure it before planning, since it sets the ceiling on what a closed-loop strategy can reach.