There is no EU-wide law that forbids collecting a competitor's published prices. The act of reading a price that any customer can see is not itself unlawful. What the law regulates is the method of collection and the reuse of the collection: database rights, the contract terms you accepted, national rules on unauthorised access, and unfair competition law.
In practice, the disputes that reach a court are rarely about the price itself. They are about a login, a terms page somebody clicked through, a server that fell over under load, or a service that reproduced someone's entire catalogue.
This is a plain reading for people who run pricing, not legal advice. Member states differ, and the facts of your setup decide the answer more than any general rule does.
Which rules actually apply?
Six instruments cover almost every argument you will hear:
- Directive 96/9/EC on the legal protection of databases, and the sui generis database right it creates
- Directive 2019/790 (DSM), Articles 3 and 4, on text and data mining
- Ordinary contract law: the terms of use you accepted, if you accepted them
- Directive 2013/40/EU on attacks against information systems, as implemented in national criminal law
- National unfair competition law, for example the UWG in Germany
- The GDPR, but only where personal data is involved. A shelf price is not personal data.
What the database right protects, and what it does not
Article 7 of Directive 96/9/EC protects the maker of a database who made a substantial investment in obtaining, verifying or presenting its contents. Two consequences matter for price collection.
First, the CJEU held in British Horseracing Board v William Hill (C-203/02) that investment in creating the data does not count. A shop that sets its own prices has created that data. The cost of running the shop is not automatically an investment in obtaining a database.
Second, even where the right exists, it bites on extraction or re-utilisation of a substantial part, or on repeated and systematic extraction of insubstantial parts that conflicts with normal exploitation, under Article 7(5). Pulling the 200 SKUs you genuinely compete on is a different fact pattern from mirroring 400,000 listings.
Innoweb v Wegener (C-202/12) is the case to read if your plan is a public comparison site. A dedicated meta search engine that passes a user's query through to someone else's database was held to re-utilise that database. Internal price monitoring is a different use, and it is worth writing that difference down in your own file before anyone asks.
Terms of use: the part people skip
Ryanair v PR Aviation (C-30/14) is the uncomfortable one. The CJEU held that where a database is protected by neither copyright nor the sui generis right, the Directive does not stop the owner from restricting use by contract.
Read that twice. Weak database protection does not mean unrestricted collection. It means the terms page becomes the operative document, provided a contract was actually formed.
So the practical question is whether you formed a contract:
- You created an account and ticked a box: a contract almost certainly exists and its terms bind your collection.
- You use trade or B2B credentials to see prices: same answer, and you have now also raised an access-control question.
- You fetched a public page anonymously with no interaction: the case for contract formation is thin in most member states, though not everywhere and not under every wording.
- Your vendor logs in on your behalf: that is your decision to own, whatever the vendor's own contract says.
Text and data mining under the DSM Directive
Article 4 of Directive 2019/790 permits reproductions and extractions of lawfully accessible works and databases for text and data mining, unless the rightholder has expressly reserved that use in an appropriate machine-readable way.
Two practical points. A bare number is unlikely to be a protected work, so for the price itself you often never need to reach the exception. The page around the price, its photography, its descriptions and its structure, is a different matter, and copying those wholesale is where copyright arguments start.
The machine-readable reservation is where robots.txt earns its place. It is not a statute, and ignoring it is not a criminal act. It is the cleanest available evidence of what the site owner reserved, and a court is entitled to read it that way.
Access, load and the thing that actually causes trouble
Directive 2013/40/EU covers illegal access to information systems and is implemented in national criminal law. Nothing about fetching a public page is illegal access. Getting round a login, a paywall, a hard block or a bot-protection challenge moves you toward it.
Load matters as much as legality. Do the arithmetic before you set a schedule. One request every three seconds is 1,200 an hour and 28,800 a day from a single worker. Ten workers at that rate is 288,000 requests a day against one shop, which is a load most mid-sized retailers will notice and some cannot absorb.
Rules that keep you boring, which is the goal:
- Identify your crawler in the user agent, with a contact address that reaches a human.
- Respect robots.txt and any TDM reservation, and log the version you read on the day you read it.
- Cap concurrency per host, and back off on 429 and 5xx responses instead of retrying harder.
- Fetch only the products you actually compete on, not the whole catalogue because it was easy.
- Do not create accounts or use credentials to reach prices that are not public.
- Store the source URL and the fetch timestamp with every price, so you can show where a number came from.
Unfair competition: the national layer
German law adds a layer that EU-level summaries often miss. The UWG catches targeted obstruction of a competitor under section 4 number 4, and the systematic exploitation of another trader's performance. Overloading a competitor's servers, or building a product that substitutes for theirs by consuming their content, is where those provisions get argued.
Republishing is riskier than reading. Using a competitor's price internally to decide your own is ordinary commercial behaviour that competition authorities expect to see. Publishing their prices to consumers, especially stale or wrong ones, invites both an unfair competition claim and a misleading advertising claim if the number is off.
What this means for a pricing team
The legal exposure here is manageable and mostly procedural. Collect narrowly, fetch politely, keep the receipts: URL, timestamp, and what the page said. Write a short note explaining why you collect what you collect, and review it when your competitor set changes.
Then spend the effort you saved on the harder question, which is whether the two listings you are comparing are actually the same product at the same pack size. In pricing work the more common mistake is a price set below what the market would have paid, not a price set too high, and no amount of legal caution will surface that for you.