Sales prospecting tools have built their effectiveness on the same foundation: LinkedIn profiles, job offers, web traffic, technology stack, marketing campaigns or activity on social networks have become the raw materials of modern sales intelligence, but this approach leaves aside a large part of the real economy.
The observation on which Leadbay is based is relatively simple: millions of SMEs, artisans, distributors or local players produce little data that can be used by traditional sales intelligence platforms. In France, the construction, trading, hotel and distribution sectors rely mainly on VSEs and SMEs that are not very digitalized. In the United States, these same companies represent nearly 40% of GDP.
However, modern prospecting tools have been designed around a highly structured internet. They excel at analyzing technology companies or companies with a strong digital presence, but much less at understanding local economic fabrics where data remains fragmented, dispersed in administrative registers, sectoral databases or simply in the memory of sales teams.
Leadbay attempts to circumvent this limitation with proprietary inference models capable of reasoning from weak signals. Where a traditional tool looks for explicit clues (LinkedIn presence, active recruitment, technologies used), the platform seeks to probabilistically reconstruct the activity of a company from very partial data.
The example put forward by the company illustrates this logic, with an independent air conditioning installer, without a significant digital presence, who can nevertheless be qualified thanks to territorial, sectoral or operational correlations. The objective is no longer just to enrich a commercial database, but to reconstruct an economic map invisible to traditional SaaS tools.
This approach gradually brings sales intelligence closer to automated economic intelligence logic. Large groups are no longer just looking for contact lists, but a detailed understanding of their local markets, their distribution networks or their subcontracting ecosystems.
The startup already claims to work with several major accounts including Saint-Gobain, L’Oréal, Nespresso, Gerflor USA, Fayat USA and Deel. According to Leadbay, its clients would triple the size of their addressable market and double the number of new clients signed. The company also claims that 55% of contracts generated via the platform come from companies that would not have been identified with usual tools.
Beyond commercial performance, Leadbay’s technological positioning also reflects a broader evolution in the market for artificial intelligence applied to businesses. After several years dominated by generative tools, part of the ecosystem is now seeking to exploit AI to reconstruct layers of information absent from the traditional web.
The platform is thus based on several levels of analysis called “LIGHT”, “INSTANT” and “DEEP”, intended respectively for massive qualification, real-time processing of priority prospects and in-depth exploration of complex data.
This division also reveals a concern that has become central in the AI economy: the trade-off between analytical depth, execution speed and calculation cost. Modern AI architectures are no longer evaluated only on their precision, but also on their ability to industrialize complex reasoning on a large scale.
To accelerate its development, the French startup announces a fundraising of 3.8 million euros from Y Combinator, Rebel Fund, Progressive VC, Bright Data Ventures, Inovexus, Roosh Ventures and Station F, as well as several business angels including Philippe and Alex Bouaziz from Deel and Edouard Mascré from Pennylane.
Founded by Ludovic Granger and Milan Stankovic, doctor in artificial intelligence and author of more than sixty scientific publications, the company is developing a platform capable of identifying and qualifying companies that are largely invisible to traditional prospecting tools.
Leadbay now plans to open an office in San Francisco in order to accelerate its American development and recruit several engineers as well as commercial profiles. At the same time, the company announces a research partnership with Sorbonne University in order to deepen the scientific work linked to its inference models.
