Insights · monthly network reading
Peppol network, September 2026: 9,568,056 participants, +3,105,220 indexed in 30 days
Published
Figures measured · refreshed hourly.
This is a measured reading of the Peppol network, not an estimate. Every table below names the endpoint it came from and the moment it was measured. Nobody self-reports into any of it.
How the growth number is measured
We count participant identifiers published in the Peppol directory and resolvable through an SMP. The 30-day figure is the net change in that index.
It is not a count of companies that joined Peppol. Two things move it: companies register, and our own discovery passes resolve identifiers that were registered earlier but that our index had not yet reached. From outside the network those are indistinguishable, so the figure is an upper bound on new registrations and never a measurement of them.
One identifier is also not one company. An organisation holding a VAT-based and a national-register identifier counts twice.
How the uptime numbers are measured
We probe every access point and SMP host we can resolve from the SML, from three vantage points, and a verdict needs quorum across them. The league percentage is the share of that operator’s probe windows that returned healthy over 30 days.
It measures infrastructure reachability. It is not a delivery-success rate, and it says nothing about whether a given invoice arrived. An operator with no probe-resolvable host is shown as not measured. A blank is honest; a 100% we did not measure would not be.
What these numbers cannot tell you
Invoice volume, message counts and revenue are published by nobody on this network, so they are absent here rather than modelled. Market share below means published participant footprint, and nothing else.
We also do not attribute causes. We can show the size and date of a movement and say which of our own passes ran. What an operator or authority intended by it is not something measurement can reach.
Sources and cadence
Every endpoint used here is free and keyless: see the developer portal, how we measure, and data licensing for the republication terms. Attribution is required, and a link to this page is the simplest way to give it.
This reading is republished hourly against the live snapshot, so the tables never drift from the API behind them. A new post follows each month.
Directory growth
How much did the Peppol directory grow in 30 days?
The index held 9,568,056 participant identifiers on 2026-09-20, up from 6,462,836 on 2026-08-21. That is +3,105,220 identifiers in 30 days, or 48.0% of the opening total. The identifiers belong to 7,566,316 distinct organisations across 122 countries.
| Measure | Value |
|---|---|
| Participant identifiers indexed | 9,568,056 |
| Distinct organisations behind them | 7,566,316 |
| Countries with at least one identifier | 122 |
| Identifiers on 2026-08-21 | 6,462,836 |
| Change to 2026-09-20 | +3,105,220 |
| Change as a share of the opening total | 48.0% |
Source: GET /v1/stats/participants and
GET /v1/stats/participants/history?dimension=total, measured 20 Sept 2026, 17:39 UTC.
Markets
Which Peppol markets moved most in 30 days?
119 markets have enough history to state a 30-day change. France alone gained 3,559,921, which is more than the whole network gained. Registrations cannot produce that shape: a discovery pass resolved a large backlog of France identifiers that were already registered. Per-market changes do not add up to the network change, and are not meant to: a bucket can lose identifiers, and identifiers whose country we cannot resolve sit in neither column.
| Market | 2026-08-21 | 2026-09-20 | Change |
|---|---|---|---|
| France | 2,265,871 | 5,825,792 | +3,559,921 |
| Belgium | 2,116,250 | 2,193,839 | +77,589 |
| Sweden | 159,264 | 172,518 | +13,254 |
| Norway | 383,755 | 390,558 | +6,803 |
| United Arab Emirates | 716 | 4,697 | +3,981 |
Source: GET /v1/stats/participants/history?dimension=country. A market is a
country code carried in the identifier itself, so identifiers whose country cannot be
resolved appear in no row. Full rankings on
Peppol adoption by country.
Reliability
Which Peppol providers are the biggest, and how reliable were they?
15 of the 15 largest operators carry a measured 30-day uptime window. Wolters Kluwer (Flowin) reads highest at 100.00% and Digdir lowest at 93.56%. Le Village Connecté carried the most incidents, 105. The rest have no probe-resolvable host yet and are shown as not measured, never as a perfect score. Uptime is measured against each operator's own hosts from three vantage points; it is not a delivery-success rate.
| # | Provider | Country | Participants | Share | 30d uptime | Incidents |
|---|---|---|---|---|---|---|
| 1 | Tickstar | Sweden | 1,079,396 | 11.5% | 99.99% | 1 |
| 2 | Serensia | France | 991,227 | 10.6% | 99.95% | 9 |
| 3 | Pennylane | France | 962,617 | 10.3% | 99.99% | 0 |
| 4 | Wolters Kluwer (Flowin) | Belgium | 634,885 | 6.8% | 100.00% | 0 |
| 5 | IOPOLE | France | 493,715 | 5.3% | 100.00% | 0 |
| 6 | jefacture (Unifiedpost x ECMA) | France | 455,389 | 4.9% | 100.00% | 0 |
| 7 | Storecove (Datajust B.V.) | Netherlands | 403,437 | 4.3% | 100.00% | 0 |
| 8 | TIIME PDP | France | 403,309 | 4.3% | 98.53% | 24 |
| 9 | Billit | Belgium | 383,682 | 4.1% | 99.88% | 15 |
| 10 | Digdir | Norway | 365,296 | 3.9% | 93.56% | 15 |
| 11 | Qonto | France | 360,692 | 3.8% | 100.00% | 0 |
| 12 | Cecurity.com SA | France | 211,410 | 2.3% | 99.99% | 0 |
| 13 | Generix | France | 178,126 | 1.9% | 99.99% | 0 |
| 14 | Le Village Connecté | France | 177,260 | 1.9% | 97.72% | 105 |
| 15 | SUPER PDP | France | 165,630 | 1.8% | 93.90% | 4 |
Source: GET /v1/providers/public, snapshot 20 Sept 2026, 19:30 UTC. Share is of
the identifiers we attribute to ranked providers, not of the whole index. The full ranking of
all 340 providers is on
Peppol provider market share.
Operators and fleet
How many Peppol providers and hosts are there?
340 providers publish participants on the network. 314 hold a live OpenPeppol certified access point listing and 247 a certified SMP listing. We probe 1,019 access point and SMP hosts: 988 were up and 31 down at the reading taken 20 Sept 2026, 19:35 UTC, with 40 incidents open. Across the last 90 days 96.58% of daily host-verdicts read up.
Source: GET /v1/providers/public and GET /v1/network, measured 20 Sept 2026, 19:35 UTC. Certified listings come from the OpenPeppol certified service provider list, synced daily.
We publish the certified count as a level: the free surface carries no history for it, so no
month-over-month movement is stated.
Keep reading