Build a List of 500 Targeted Companies in One Hour (Without a €200/Month Subscription)

You're validating a product idea, launching a freelance business, or looking to reach out directly to potential partners. In all these cases, you need a structured list of companies: official name, address, sector, size, and ideally a direct contact. The problem is that the moment you start looking, you run into a Pages Jaunes Pro page asking for a minimum of €33/month with an annual commitment, or a Kompass that doesn't even publish its pricing clearly.

Yet France is one of the countries in the world where company data is the most open. The INSEE Sirene directory is free, public, and contains administrative information on approximately 30 million business establishments. With a bit of method, you can build a list of 500 targeted companies in one hour — no subscription required.

The foundation: INSEE's Sirene directory, free and exhaustive

The Sirene directory is a public file made available by INSEE since 1973. For each French company it contains the SIREN number, official name, address, NAF code (which defines the primary business activity), declared headcount, and date of incorporation. It is freely downloadable on data.gouv.fr and updated daily.

For reasonable volumes (up to a few thousand rows), you don't even need to download the full file. Several web tools let you filter Sirene online: Pappers, societe.com, the official API Entreprise. All free for personal use and limited volumes.

What Sirene does not contain: email addresses, websites, phone numbers (except occasionally in specific cases). For that data, you'll need to cross-reference with other public sources.

Filtering by precise sector using NAF codes

The NAF code is the key to precise targeting. "Hairdressers" is NAF 96.02A. "Plumbers" is 43.22A. "Law firms" is 69.10Z. The full list of 732 NAF codes is published by INSEE.

The advantage: you can combine two or three NAF codes to get a highly targeted list. "Bakeries in Île-de-France" = NAF 10.71B + departments 75/77/78/91/92/93/94/95. Result: a few thousand establishments, deduplicated, with exact addresses.

The downside: for loosely categorised activities (for example, "digital marketing agencies" which can fall under 73.11Z but also 70.22Z or 62.02A depending on how they registered), you'll miss part of your target. Solution: combine multiple codes or supplement with other sources.

Supplementing with Google Maps for contacts and websites

Sirene gives you the postal address. Google Maps adds the phone number, website, sometimes the email, opening hours, customer reviews, and review count. For a targeted lead generation approach, this supplementary data makes all the difference.

Three methods to retrieve this information.

The first, manual, involves searching each company individually on Google Maps and copy-pasting the details. At 30 seconds per listing, 500 companies = 4 hours. Doable but tedious.

The second uses a free Chrome extension for Google Maps scraping. There are plenty of them; they work by batch-opening results and export a CSV. Downsides: variable quality, many include sketchy tracking, and most cap at 50–100 entries per free session.

The third uses a dedicated Google Maps scraping tool. Several options are available: Scrap.io (€49/month, 7-day trial), Smappen (~€30/month), or all-in-one tools in alpha that combine extraction + enrichment + verification in a single workflow.

Cross-referencing and cleaning: the step that separates a useful list from a junk list

When you aggregate Sirene + Google Maps + a third source, you quickly end up with duplicates (the same bakery listed three times under slightly different names), outdated closures (a deregistered company), and conflicting data (two different phone numbers for the same SIRET).

Cleaning takes as long as extraction if done by hand. Three rules to move fast.

First rule: deduplicate on the SIREN, not on the name. Two different names can refer to the same entity. The SIREN is unique.

Second rule: remove deregistered companies. Sirene flags deregistrations, but Google Maps often leaves them online. Cross-referencing with the Sirene deregistration date avoids contacting a closed business.

Third rule: at minimum, check headcount. A "communications agency" declared with 0 employees is a sole trader, not an agency — relevant for some approaches, worth excluding for others.

The sources people overlook that change everything

Beyond Sirene and Google Maps, three lesser-known sources are valuable depending on your targeting.

The INPI (Institut national de la propriété industrielle — France's intellectual property office) publishes registered trademarks. If you're looking for "all companies that registered a trademark in cosmetics in 2025", that's where to look.

The Bulletin officiel des annonces civiles et commerciales (BODACC — France's official gazette for commercial announcements) lists incorporations, amendments, and insolvency filings. Useful for targeting recently created companies (often in the market for initial suppliers or service providers).

Pôle Emploi/France Travail publishes job listings as open data. A company actively recruiting is growing — an actionable signal for sales outreach or targeted unsolicited applications.

The trade-off: what it actually costs in time

For 500 targeted companies with name, address, phone number, and website, using a 100% manual method (Sirene + Google Maps one by one + cleaning), budget 6 to 10 hours.

With a paid Google Maps scraping tool (Scrap.io or equivalent), budget 1 to 2 hours for 500 rows, but add €49/month in subscription costs.

With an all-in-one tool in alpha that combines Sirene extraction + Google Maps + email verification + deduplication, budget 30 minutes to one hour, with no subscription (just apply for alpha access). That's exactly what outsend offers in alpha.

This article is part of a broader series: see the prospecting guide for non-registered businesses.

Try outsend for free

All-in-one, built for European B2B. Free alpha access on application.

Request free alpha access

FAQ — Building a targeted company list

Where can I find a free list of French companies?

The INSEE Sirene directory, freely accessible on data.gouv.fr, contains the name, address, NAF code, and headcount of French business establishments. Free to download, updated daily. For reasonable volumes, web interfaces like Pappers or societe.com allow you to filter online without downloading the full file.

How do I export a list from Sirene?

Three options: download the full dataset as a CSV from data.gouv.fr (several gigabytes — requires a tool capable of handling large files), use the official API Entreprise (free, limited to a few thousand requests per day), or use an interface like Pappers that lets you export filtered subsets as CSV.

What is the difference between Sirene, Pages Jaunes Pro, and Kompass?

Sirene is free, public, and exhaustive, but only contains administrative data (no website, no email, no reviews). Pages Jaunes Pro adds websites, phone numbers, and business categories, for €33 to €78 excl. VAT/month with a commitment. Kompass offers an internationally enriched database with financial data, priced on request (estimated at around €1,490/year minimum). For 95% of use cases in France, Sirene + a supplement is sufficient.

How many companies can I extract from Google Maps for free?

Without a tool, searching manually, you can retrieve as much as you want but the pace is slow. With a free Chrome extension, you're generally capped at 50–100 entries per session. Beyond that, Google blocks your IP. Dedicated tools (Scrap.io, Outscraper, or all-in-one solutions in alpha) handle IP rotation and support volumes of several thousand.

Is scraping Google Maps legal?

Yes, within the framework of scraping publicly accessible data. According to the CNIL (France's data protection authority), collecting publicly accessible data is lawful provided the purpose is legitimate, individuals can opt out, and the site's terms of service do not explicitly prohibit it. For Google Maps, the situation is nuanced: the Terms of Service prohibit automated extraction, but European case law recognises the primacy of the right to information over unilateral contractual restrictions.

How long does it take to build a list of 500 targeted companies?

Using a 100% manual method (Sirene web + Google Maps one by one), budget 6 to 10 hours for 500 enriched records. With a dedicated scraping tool, 1 to 2 hours plus cleaning time. With an all-in-one tool that handles extraction and deduplication, 30 minutes to 1 hour.

Try outsend for free

All-in-one. Far cheaper than every competitor. Alpha access on application.

Request free alpha access