A furniture company running WhatsApp ads was generating hundreds of leads per week, but couldn't convert them fast enough. The sales team was overwhelmed, and by the time they responded, most leads had already lost interest. LoopTree AI built a WhatsApp agent that handles the full pre-sales funnel and routes qualified leads directly to the right rep.
The company was investing heavily in WhatsApp click-to-chat ads. The ads worked: leads were coming in. But ad-generated interest has a very short shelf life. A person clicks an ad, sends a pre-filled "I'm interested" message, and if nobody responds within minutes, that lead is gone.
The sales team couldn't keep up with the volume. They were spending too much time on unqualified leads and not enough on the ones ready to buy. The funnel was leaking at the top, and ad spend wasn't translating into revenue.
Lead volume was growing but conversion was flat. The team couldn't respond fast enough, and there was no pre-filtering in place. Every lead got the same treatment regardless of intent, which meant reps were wasting time on people who were just browsing.
Nearly half of all inbound messages (45.9%) arrived outside business hours. That's 1,272 messages that would have gone completely unanswered without automation. On weekends alone, 221 unique leads reached out. That's 26% of all leads arriving when no human rep was available.
“We were paying for leads that nobody was answering. By the time the team got to them, the interest was gone.”
We built a WhatsApp AI agent that handles the full pre-sales funnel. When a lead messages in from an ad, the agent responds in seconds, greets them, qualifies their interest, and starts a real product conversation.
The agent is connected to the company's ERP system, giving it access to a live catalog of over 1,500 products. It can search by category, show product images, calculate prices including tax, and recommend options based on what the customer is looking for.
Once a lead is qualified, the agent routes them to the right sales rep through the CRM using round-robin assignment. The rep gets the full conversation context and the customer's requirements, so they can pick up exactly where the agent left off.
| Challenge | Solution | Result |
|---|---|---|
| Leads going unanswered | AI agent responds in 13 seconds | 847 leads captured |
| No after-hours coverage | 24/7 automated responses | 45.9% after-hours leads saved |
| No lead pre-filtering | AI qualifies before routing | Reps only get qualified leads |
| Can't search 1,500+ products | ERP-synced catalog search | Real-time product recommendations |
| Manual rep assignment | CRM round-robin routing | Instant handoff with context |
In just 5.5 weeks, the agent processed 11,440 messages autonomously and captured 847 leads. That's roughly 22 leads per day, every day, including weekends and holidays.
These aren't shallow interactions. The average conversation runs 13.5 messages, with the agent sending an average of 10.2 messages per lead. It's actively recommending products, sharing images, and calculating prices. The conversations are real pre-sales engagements, not just greetings.
The sales team now only handles leads that have been qualified and routed. They spend their time closing instead of filtering. Ad spend is finally converting because every lead gets an immediate, informed response.
“The agent is having real sales conversations. It’s not just answering questions. It’s recommending products, showing images, and calculating prices. By the time the lead reaches a rep, they already know what they want.”
The agent handles the full pre-sales conversation: product search, recommendations, pricing, and images. It qualifies the lead and routes them to a human rep for final negotiation and closing.
The agent is synced with the company's ERP system in real time. It searches across 1,500+ products, retrieves images, and calculates prices with tax included.
Qualified leads are routed through the CRM using round-robin assignment. The rep receives the full conversation history and the customer's requirements so they can pick up immediately.
The agent runs 24/7. In this implementation, 45.9% of all messages came outside business hours. Without the agent, those leads would have been lost entirely.
Tell us the task your team repeats the most. We’ll show you an agent doing it, before you pay for anything.
This case study describes how LoopTree AI, an AI automation agency based in Santo Domingo, Dominican Republic, built a WhatsApp AI sales agent for a furniture company. The agent qualified 847 leads in 5.5 weeks with a 13-second median response time and 45.9% after-hours coverage. The implementation included a custom dashboard for lead tracking and sales rep routing. LoopTree AI specializes in building custom AI agents with 360 dashboards for businesses in the Dominican Republic.
