Five in-depth articles. Every legal statement in them has been read at the primary source — the statute itself, the Official Journal, a published authority document. Every figure comes from a measurement of our own. Anything we cannot substantiate is not here.
MLBF · services strategy, as of 08 January 2026
The competent authority has published its market surveillance strategy. What it really says about automated preliminary checks, sampling and the escalation procedure.
§ 2 no. 17 · § 3(3) BFSG
The micro-enterprise exemption in its exact wording, the “or” that is usually dropped, pure B2B, and three further exemptions that are barely mentioned.
Our own measurement · 20 sites · 108 pages
20 sites, 108 pages, 792 consolidated findings: contrast accounts for 54 per cent, four rules for 86 per cent — and one site had none at all.
Our own analysis · 19 sites · 794 findings
84 per cent of findings can be proposed automatically, more than half of those are pure contrast arithmetic — and 792 findings become 284 decisions.
§ 25 TDDDG · 8 sites measured
A fresh browser, one page load, no click: on one of eight sites, 17 advertising and tracking calls and 5 tracking cookies ran before any interaction.
So that it is clear what any statement here rests on — and what it does not.
Primary sources, not summaries. A legal statement appears here only if we have looked up the wording ourselves: the consolidated statute, EUR-Lex, the published documents of the competent authority. Every article carries a source list at the end with the exact provision, a link and the retrieval date. We take nothing from blogs or law-firm pages. That is not affectation: twice in this project a claim adopted that way had to be withdrawn.
Our own measurements, not market estimates. The figures on barriers, remediation effort and cookies come from runs of our own testing tool against real, publicly reachable sites. Date, scope and method are stated in each article. Where a sample is small, we say so — it then shows spread and recurring patterns, not a market average.
No names of the sites we measured. We give the sector and the number, never the domain. A list of names and defect counts reads as a ranking, and a ranking quickly becomes a verdict about someone else’s business. That is not ours to pass: an automated test covers only part of the criteria, and every measurement is a snapshot of a single day.
Findings, not legal verdicts. We describe what was measured. We do not say whether a site breaches a law — not even when it seems obvious. And we do not certify the opposite either: a clean bill of health would be just as much a legal statement about someone else’s business as an accusation.
Enter your address, get the number of findings straight away. We test against EN 301 549 V3.2.1, which points to WCAG 2.1 Level AA — the same standard the market surveillance authority applies.