Nexlead replaced the cork bulletin board near the grocery-store exit with a digital screen that interviews shoppers in under a minute and routes a qualified lead to local service providers, with sixty percent of revenue committed to cancer research.
May 8, 2026
Walk into a Metro grocery store almost anywhere in Quebec and you can still find, near the exit, a cork board pinned with business cards. A snow-removal contractor. A piano teacher. A handyman whose number has been folded and tacked back up so many times the paper has gone soft at the edges. The board has been there since the first version of the store. It is, quietly, one of the oldest local-advertising surfaces still standing in North American retail.
It is also almost completely passive. A shopper sees a card, photographs it on their phone, walks out, and never calls. The advertiser, paying nothing and expecting nothing, has no way to know any of this happened.
Cork boards work. They have been working at maybe two percent of their potential for forty years.
Rollad Media, a small company headquartered in Mont-Royal, set out to recover the other ninety-eight, under a brand it called Nexlead.
Benoit Lefebvre's plan was simple. Replace the cork boards with digital displays. Place them in high-traffic public venues, beginning with Metro, which had agreed to host an initial pilot of roughly ten screens in Quebec at no cost. Let local service providers post for free. Charge only for what the network actually delivered: qualified leads, routed to the right provider. And donate sixty percent of revenue to cancer research, in memory of his granddaughter Billy.
A free posting tier is a strong commercial promise. It is also a heavy load. The network cannot earn its keep on volume of advertisers. It earns its keep on the quality of what those advertisers receive. Every shopper who scans a code on a Nexlead screen has to be quickly understood, quickly scored, and quickly handed to whoever is best placed to help them.
Doing that with human interviewers paid by the hour would have consumed more revenue than the network produced. The whole model collapsed without intelligence in the middle. The donation flow collapsed with it.
This is not a story about a business that wanted AI. It is a story about a business that did not exist without it.
Metro had committed to host the screens, but only if the network was operational by the pilot date. The application around them, the interface, the form, the payments, the dashboards, was being built by Designoweb, a generalist agency working from India. Designoweb's scope explicitly excluded the AI layer. The brain of the system did not exist. The launch was three weeks out.
INTO was brought in to build the brain. Lefebvre had already made a clean choice that shaped everything that followed: he would not hire a team. He would pay a specialist to build the engine and operate it in the background, so he could spend his hours signing the next distribution agreement and recruiting the next class of advertisers.
The system that shipped is most easily described as one continuous minute of time.
A shopper notices a Nexlead screen near the produce aisle and scans the QR code. The scan opens a short form that captures their basic contact information and the service they are looking for. Submitting it lands them on a welcome page that asks one question: would they like to chat now, or be contacted later? If they choose later, a unique link is emailed for them to resume on their own time. If they choose now, the same page slides them into a conversation with an assistant who already knows their name and what they came for.
Her name is Billie, after Lefebvre's granddaughter. The interview that follows is short by design, with a target of under two minutes. It adapts in depth: a snow-removal lead does not need the same questioning as a kitchen renovation, and the system knows the difference. Billie asks somewhere between five and ten questions, listens, follows up where the answer is vague, and thanks the shopper at the end.
Within roughly a minute of the conversation closing, the lead has been summarized, scored on quality, and delivered as a structured record to the Nexlead backend, which routes it to the right advertiser. The shopper, by then, is in the parking lot.
A few choices distinguish Billie from a chat widget bolted onto a website.
Billie has a personality. A name, a defined tone, a coherent welcome, and a length budget calibrated to a conversation held one-handed at the front of a store. People will not spend three minutes filling out a form they could have filled out at home. They will spend forty seconds in a conversation that feels like one.
The system never loses the people who change their minds. If a shopper opens the chat and walks away, the form they already submitted still flows through as a basic lead.
The system protects its own economics. It cannot be re-run by a curious or hostile user trying to drive up costs through the back door. The mission depends on the math, and the math depends on Billie being expensive only when she is producing something.
The screens went live across the Metro pilot footprint. The cork board, in some stores, is still on the wall next to them. That is fine.
What changed is what the surface is now able to do. Each interaction has a known beginning, a known duration, a known outcome. The shopper who would have photographed a card and never called now leaves a structured record of what they wanted, scored on how seriously they wanted it. The advertiser receives a lead they can act on within minutes, often before the shopper has finished the drive home.
And the donation flow runs. It runs on the back of Nexlead's revenue, and Nexlead's revenue runs on Billie doing her job in roughly sixty seconds, in a Metro store, in memory of a granddaughter who will never see the network she helped name.
The cork board has been trying to do this for forty years. With a minute of conversation in front of it, properly designed, it can.
With expertise in strategy and product management, Sebastien helps organizations integrate AI in their business operations and services.
Nexlead, an initiative of Rollad Media in Mont-Royal, set out to replace the cork bulletin boards near Quebec grocery-store exits with digital screens that pay local service providers only for qualified leads, donating sixty percent of revenue to cancer research in memory of founder Benoit Lefebvre's granddaughter Billy. The mission depended on unit economics that did not work without intelligence in the middle. INTO designed, built, and now operates Billie, the AI assistant that turns a thirty-second QR scan in front of a Metro screen into a structured, scored, ready-to-route service lead within roughly a minute.
Thirty minutes. We'll tell you what to build, what not to build, and what it would take.