Galo speaks Spanish, for a better experience, please turn on English subtitles.
The problem
The WhatsApp queue that wouldn't stop
Grupo Mavesa's commercial team had a bottleneck every player in the automotive industry knows: the weekend. Customers would message on WhatsApp on a Saturday afternoon asking about a model, a price, a payment plan. And they had to wait until Monday for someone to respond.
That delay built a backlog. The call center started every Monday with dozens of open conversations with no context, no prioritization, and no way to tell which customer was actually ready to buy and which was just price-shopping. The frustration ran both ways: agents burned hours reading old threads, and the customers with real intent cooled off waiting.
For a group running multiple brands and growing lead volumes, the model didn't scale. Every new brand added more queue to the same call center.
The turning point
The day the AI agent sold a car on its own
The shift didn't happen at once. At first, only 5% of leads from the brands connected to the Patagon AI agent were coming through it. Then it climbed to 22%. And one day they sold a vehicle that had been handled, negotiated, and closed entirely by the AI agent on WhatsApp — with no human from the call center involved.
For the team, that was the signal they'd been waiting for. It wasn't a pilot: the AI was executing the commercial job end-to-end, and winning.
"The moment it clicked was when I saw all the time we saved on handling, and the thing we actually wanted: the sale."
The results
From 5% to 50% of leads via the AI agent
Today, 50% of every lead entering the group comes through the AI agent. It started at 5%, climbed to 22%, and keeps growing month after month as new brands join the platform.
Conversion rate moved too. Before Patagon AI, the group's historical range was 20% to 22%. Today, depending on the brand, they're between 30% and 40%. With the latest version of the agent, one brand jumped from 22% to 32% in a single month. That jump doesn't come from pushing more leads into the funnel: it comes from qualified leads reaching agents better filtered. Price-only inquiries stay in the bot conversation. The ones with real intent get routed to the right dealership with the full customer history ready for the agent to close.
The team impact was just as concrete: call-center time spent answering repetitive questions dropped enough to free up the equivalent of one full-time employee. And with the newer brands launched in February, time to first sale dropped from two months (as it had been with the first brand) to a single month.
"Conversion has improved. It used to be 20 to 22%. We're now above 30%. On some brands, we're at 40%."
How they see it now
From vendor to strategic partner
The relationship with Patagon AI stopped being a tool contract and became an operational partnership. Every new brand adds data, every iteration of the agent improves the filter, and the Mavesa team sees the Patagon AI team as part of their commercial operation, not an outside vendor.
"Patagon has become a strategic partner for Grupo Mavesa, giving us the support we need, initiative, experience, and a clear view of how to improve."
"I can sleep easier. I know the system is responding and we're moving forward."
What's next
From 4 to 6 brands, and new business lines
By month-end, Grupo Mavesa will have onboarded two more brands to the AI agent, bringing the total to six brands running on Patagon AI via WhatsApp. The next step is to replicate the model across other business lines in the group, with the same playbook: 24/7 agent, automatic filtering, dealership routing, and conversion metrics that are comparable brand-to-brand.
For Galo Molina, the direction is clear: keep opening new business lines and keep scaling the agent. The partnership with Patagon AI will grow alongside that expansion.
