You want to reach out to 200 e-commerce sites in Lyon to pitch a service. You have the URLs. The next step is figuring out who actually runs each site: the brand you see on the homepage is not always the company that makes the decisions. A store called "Lyon Déco" might actually be published by a single-member SAS (SASU) based in Paris, run by an investor who manages thirty similar companies. Without the real identity, you are writing to no one.
Legal notices (mentions légales) are the public identity card of every French website. They have been mandatory since the LCEN (Loi pour la Confiance dans l'Économie Numérique, article 6-III) and list the publisher, the editor-in-chief, the hosting provider, the RCS registration number, and the VAT number for businesses. Extracting them at scale from a prospect list reveals the real corporate structure behind the sites you visit — information that neither Google nor LinkedIn provides for free.
What every site must display (by law)
For individuals: full name and address. For legal entities: company name, legal form, share capital, registered address, phone number, contact email, RCS registration + city, intra-community VAT number (if applicable), name of the editor-in-chief, and hosting provider identity including their address.
Why legal notices are more useful than WHOIS for prospecting
A WHOIS lookup gives you the administrative owner of a domain name — but 95% of individuals and 60% of businesses mask that data through a registrar privacy service (GoDaddy Privacy, NameCheap WhoisGuard, OVH anonymisation). Legal notices, however, cannot legally be masked: they are a mandatory disclosure obligation on the publisher, not the registrar. The company behind the site must publicly identify itself, under penalty of one year's imprisonment and €75,000 in fines for individuals, and €375,000 for legal entities (LCEN article 6-VI).
For a sales rep prospecting 300 e-commerce sites, legal notices deliver on a single page: the official company name, SIREN registration number, legal director, and direct contact. Three pieces of information that change everything about what comes next: who to address your message to, on what legal basis to check creditworthiness (Kbis extract via Pappers or societe.com), and which email to use for the first professional outreach.
The 4 common cases of brand / publisher misalignment
Across 100 prospecting sites examined, the same patterns consistently appear:
- Trade name ≠ registered company name. The site "Lyon Déco" is published by "SARL Aménagement Habitat." The decision-maker is Pierre Martin, director of the SARL — not a non-existent "Lyon Déco sales team."
- Holding group owning 30 brands. The site is published by a SASU whose director manages 30 other SASUs. Sending a generic email to 30 sites means reaching the same decision-maker 30 times. You waste 30 personalised messages if you miss the pattern.
- Franchise or licensed brand site. "Pizzeria Bella Vita - Toulouse" is an independent franchisee. The decision-maker is not the Bella Vita head office — it's the local franchisee. The legal notice points you to the right contact.
- Sole trader behind a corporate-looking site. The site looks like a "digital agency" with a team and office space, but the legal notice reveals an individual entrepreneur (EI) with a residential address. Adjust your commercial strategy and pricing accordingly.
Manual method: 3 minutes per site
For any given site, the process is almost always the same:
- Site footer → look for a link labelled "Mentions légales," "Mentions," "Legal," or "Informations légales" (sometimes buried under "About" which contains the legal notice)
- Open the page and copy the content into a clean text editor
- Extract: company name, legal form, SIREN/SIRET, RCS + city, share capital, address, director, email, phone, hosting provider
- Cross-check the SIREN on annuaire-entreprises.data.gouv.fr (primary INSEE Sirene source) to confirm the company exists and is active
3 minutes per site is manageable for 10 prospects. It does not scale to 300.
Why automated extraction is challenging
Legal notices are mandatory, but their format is not standardised. Every site has its own page, its own URL path, its own hierarchy, its own vocabulary. Four technical obstacles:
| Obstacle | Difficulty |
|---|---|
| Finding the legal notice URL without a predefined list | Crawling the site to identify the "mentions" link among 50 footer links |
| Legal notice embedded in a PDF | Extracting text from a sometimes poorly structured PDF |
| Notices displayed as images (rare but found on older sites) | OCR with variable accuracy |
| Free-form content structure | "SIREN 123 456 789" can appear as "n° SIREN : 123456789", "RCS Paris B 123 456 789", or "siren 123456789 RCS Paris" |
For 300 sites, manual extraction takes 15 hours. Naive regex extraction fails on 30% of cases (PDFs, atypical structures, missed French keywords). Reliable extraction requires an HTML parser combined with a model that recognises French corporate patterns (SIREN, RCS, SAS, SARL, SCI, EURL, SASU, EI, micro-entrepreneur).
What the outsend legal_mentions module does
Starting from a list of domains (CSV, or a list extracted via the Google Maps scraping module), outsend visits each site, automatically identifies the legal notice page, and extracts the following structured fields:
- Registered company name and trade name (if different)
- Legal form (SARL, SAS, SASU, EURL, SCI, EI, association loi 1901)
- SIREN / SIRET extracted and validated (Luhn algorithm confirms number consistency)
- RCS registration and city of incorporation
- Share capital in euros
- Full registered address
- Contact email (often a generic
contact@, sometimes a named address) - Phone number
- Editor-in-chief identity (often the legal director)
- Hosting provider (useful for assessing a site's technical maturity: OVH/Gandi vs Wix/Shopify)
Export as CSV/XLSX or enrich an existing list directly. The module flags non-compliant sites (missing or incomplete legal notices) — a commercially valuable data point, since the publisher is technically in breach of the LCEN, which can serve as an angle for selling a compliance service.
Concrete use cases
Local B2B outreach — A web agency targeting 500 e-commerce sites in Burgundy extracts the legal notice from each site, identifies the 80 with incomplete pages (a signal of low maturity and a warm prospect for a compliance service), and personalises an email addressed to the named director by first name.
Competitive intelligence — A SaaS provider extracts legal notices from all sites running a competitor's product (detected via the tech-stack module). Cross-referencing by holding group: if 10 different sites are published by the same SASU, that is a strategic account to approach at the group level, not subsidiary by subsidiary.
Purchased file validation — When you receive a third-party prospecting file, validating the legal notices of each company's website exposes stale lists (defunct sites, departed directors, changed company names). 20–30% of files sold on the market contain information that is more than 18 months out of date.
GDPR compliance of the extraction process
Legal notices are public data by mandatory disclosure. Collecting them does not constitute processing of personal data under GDPR for the corporate information (company name, SIREN, registered address). However, the director's name and contact email are personal data as soon as they identify a natural person. The CNIL states that B2B commercial prospecting is permitted on the basis of legitimate interest, provided that the person contacted has been informed and can easily opt out.
In practice: you may collect and use this data for targeted initial outreach, on condition that your first email discloses the data source (the site's legal notices) and includes a simple unsubscribe link. Data retention must be limited to the duration necessary for prospecting — the CNIL recommends 3 years after the last contact for prospects who have not responded.
FAQ
Do WordPress and Shopify sites include legal notices by default?
No. WordPress and Shopify offer themes that may include a "Legal" page, but nothing is generated automatically. A site built in 30 minutes often has an empty legal notice page or one copied from a template. The legal_mentions module detects and flags this.
Does extraction work for non-French sites?
The module is calibrated for the French legal framework (LCEN). It can still extract equivalent fields from European sites (German Impressum, Spanish Aviso legal, etc.) but with lower reliability. For international prospecting, plan for a manual validation step.
How much do competitors charge for this extraction?
Specialist B2B data providers typically charge per enriched record. Legal notice extraction is rarely sold as a self-service product — it is bundled into broader enrichment packages priced at €0.30–1.50 per qualified contact.
What to do when a site has no legal notice?
It is a breach of LCEN article 6-VI and is subject to prosecution. For prospecting purposes, flag these sites as "non-compliant" and either exclude them from standard outreach — or make them the target of a specialised pitch (compliance services). To file a complaint, the competent authority is the DGCCRF.
Can the hosting provider be useful for prospecting?
Yes — it qualifies the technical stack. A site hosted on OVH or Gandi suggests a French tech team or freelancer. A site on Wix/Squarespace suggests a non-technical founder. A site on AWS/GCP suggests a mature engineering team. Tailor your commercial approach to that level of technical maturity.
Are legal notices enough to qualify a prospect?
No. They give you the legal identity but not the need. Combine with: homepage text extraction (market positioning), tech stack detection (technical needs), Google Maps reviews (commercial health) — all available in outsend.
To see how this fits into the bigger picture, explore all OutSend features.
Try outsend for free
All-in-one, built for European B2B. Free alpha access by application.
Request free alpha accessoutsend.xyz — All-in-one scraping and prospecting platform, currently in alpha. Free application at /demander-acces.