Magalu's AI WhatsApp: What Multichannel Sellers Can Learn from R$ 100 Million in Sales

Magazine Luiza moved R$ 100 million with “Lu’s WhatsApp” in just eight months, since November 2025. The tool, which uses artificial intelligence to conduct the entire sales process within WhatsApp, has already generated over 7 million interactions and 16 million conversations. And 20% of consumers return to buy through the channel. For multichannel sellers operating on marketplaces like Mercado Livre, Shopee, and Amazon, Magalu’s case is not just a curiosity from a large retailer — it’s a sign of where consumer behavior is heading and how technology can increase conversion without eroding real margin.

What made “Lu’s WhatsApp” work

Developed by LuizaLabs with Google technologies, the system allows customers to search, receive recommendations, pay, and track orders without leaving the chat. Conversion is three times higher than the company’s traditional channels. There were 7.7 million unique users and an NPS of 84.5 — a very high score for any sales channel.

The secret is not just the use of AI, but the complete integration of the journey: from discovery to post-sale, all within an environment where consumers already spend hours a day. For the multichannel seller, this reinforces the importance of being where the customer is, but with execution that does not compromise the real margin per order.

Why this matters for multichannel sellers

If a retailer the size of Magalu can generate R$ 100 million with a conversational channel, it’s because Brazilian consumers are increasingly willing to buy through messages. For the seller operating on multiple marketplaces, replicating this strategy requires care: each marketplace has its own fees, commissions, and shipping rules. An order closed via WhatsApp can have a very different real margin than an order on Mercado Livre or Shopee, depending on how you structure shipping costs, payment gateway fees, and time spent on service.

The case shows that AI can scale service without proportionally increasing costs. But for the seller, the real gain appears when you can measure the real profit of each sale originated in this channel — and compare it with other channels. Without a view of profit intelligence, you may be trading volume for margin.

Lessons for the multichannel seller

• Invest in high-conversion channels, but measure the real margin of each. WhatsApp may have higher conversion, but if shipping costs or payment fees are high, the real profit may be lower than on the marketplace.
• Use AI to reduce operational costs. Automating responses and recommendations frees up time to focus on higher-value products.
• Think of post-sale as part of the journey. Customers who return to buy (20% in the case) have higher LTV and dilute acquisition costs.
• Centralize data from all channels. Without a unified DRE, it’s impossible to know if the effort on WhatsApp is actually generating more real profit than traditional marketplaces.

Magalu showed that it’s possible to generate high revenue with a conversational channel. For the multichannel seller, the opportunity is not to copy the model, but to adapt the logic: use technology to sell more, with control over every cost — from shipping to ADS — and focus on real profit per order.