ATV Riders is a Bogotá-based company specializing in the sale of ATVs and off-road vehicles. Their paid campaigns on Meta were generating significant inbound volume, but that volume had become a liability.With over 200 leads arriving per day via WhatsApp and no CRM in place, the commercial team was drowning in conversations, losing qualified prospects in the noise, and spending hours on contacts that were never going to buy.The mandate was clear: build a system that handles the volume automatically, filters for intent, and delivers only sales-ready leads to the advisors.
Diagnosing the Problem
Before writing a single automation rule, I ran a discovery process with the commercial team, mapping the entire customer journey from first message to closed sale. These sessions surfaced three critical constraints that shaped every design decision:1. No existing CRM infrastructure. The team managed everything manually via WhatsApp. Any solution had to introduce CRM capabilities without a steep adoption curve.2. Autonomy after delivery. The client needed to be able to edit chatbot content, update the catalog, and manage conversations independently, without relying on a developer for every change.3. High daily volume with low signal. 200+ daily conversations meant the cost of unqualified leads wasn't just lost time. It was commercial capacity actively drained from the people who should be closing deals.
Why Kommo: A Strategic Tool Choice
With these constraints mapped, I evaluated the tooling options. The decision to build on Kommo CRM was deliberate:
Native WhatsApp Business API integration: no middleware, no workarounds, no additional connectivity layer required.
Visual chatbot builder: the commercial team could modify flows, update response text, and manage the catalog through an intuitive interface without technical knowledge.
Built-in CRM pipeline: the same platform that runs the chatbot also tracks every lead through the sales funnel, giving the team a single operational view.
Low learning curve: adoption friction is a real reason automation projects fail. Kommo made it realistic for a non-technical team to own the system after handoff.
The alternative was repeating the integration complexity I had managed in previous projects. Kommo eliminated that layer entirely.
Building the Automation System
Chatbot Architecture
The conversation flow was engineered around a single insight from the discovery sessions: intent segmentation at the first touchpoint is the difference between a productive conversation and wasted time.The chatbot initiates every new conversation with a greeting and presents three interactive WhatsApp buttons, native tap-to-select options that require zero typing from the user:
📋 View Catalog: Triggers an interactive WhatsApp catalog menu, letting users browse ATV models and pricing directly in the chat.
❓ I Have Questions: Routes to an FAQ-style flow covering location, business hours, financing options, and other common pre-purchase questions.
🏍️ I Want to Buy: Captures high-intent leads and flags them for immediate follow-up by a sales advisor.
This three-path structure meant that every conversation was classified before a human ever had to read it.
Automated Re-engagement & Closure Logic
Not every lead converts on first contact. To prevent qualified prospects from going cold and to keep the pipeline clean, I built two automated behaviors into the system:
Re-engagement trigger: If a user initiates a conversation and doesn't respond for 60 minutes, the chatbot sends a follow-up message to bring them back into the flow.
Auto-closure: If there's still no response after the re-engagement window, the conversation is automatically marked as closed and archived. This keeps the pipeline honest with no ghost leads inflating the numbers.
Sales Funnel Design
Inside Kommo, I built a custom pipeline reflecting the actual sales stages identified during the discovery sessions:
New Lead → Catalog Viewed → Question Answered → High Intent → Advisor Assigned → Negotiation → Closed (Won / Lost)
Every lead that enters via WhatsApp is automatically created as a contact in Kommo and placed at the correct stage based on the path they took through the chatbot. Sales advisors open their day with a structured, prioritized pipeline, not a list of 200 unsorted messages.
Content Optimization
To ensure the chatbot's tone was clear, persuasive, and consistent with ATV Riders' brand voice, I used Gemini 3.1 Pro to refine every piece of copy in the flow, from the initial greeting to the catalog descriptions and FAQ responses. This step eliminated the common failure mode of chatbots that are technically functional but conversationally awkward.
Results
Daily inbound volume managed: 200+ leads/day
Lead qualification: Automated at first touchpoint
CRM adoption: Team operational post-handoff, no technical dependency
Pipeline visibility: Full funnel from first message to close
Advisor workload: Reduced to high-intent contacts only
Follow-up coverage: 100% with no lead dropped without a re-engagement attempt
What This Project Demonstrates
Every design decision in this project traces back to the same principle: automation that fits the team, not automation that requires the team to fit it.Tool selection, conversation architecture, funnel design, and post-handoff autonomy were all engineered around the real operational context of a small commercial team managing a large inbound volume.The system didn't just reduce manual work. It gave the team something they didn't have before: clarity on where every lead is, what they want, and what action to take next.
Tech Stack
WhatsApp Business APIKommo CRMMeta Business SuiteGemini 3.1 Pro
Kommo CRM: chatbot + pipeline
Gemini 3.1 Pro: copy optimization
Related projects
Quick Metal Shop & Dakoma Roofing: AI-Powered Ad Creatives
3.92xROAS
How a lean, one-person production pipeline, built entirely on AI avatars, cloned voices, and rapid editing, replaced traditional video shoots and drove measurable results across two brands, two industries, and two ad objectives.
How a brand manual built from the ground up became one of the standout booths at ArquiExpo 2026, and a corporate video carried by a fully original soundtrack, composed with AI.
Quick Metal Shop: Building a Marketing Department From the Ground Up
Establishing a complete full-stack marketing department and digital infrastructure from scratch, including brand manual, CRM automation, paid media, and AI content engines.
Dakoma Roofing: Freelance Rebranding & Digital Automation
Complete brand overhaul, digital channel modernization, AI-powered SEO automation, and premium ad creative production for a B2C roofing and construction services provider.
Building This Portfolio: AI-Assisted Development in Practice
<2mindeploy time
How I designed, built, and deployed a production-grade Next.js portfolio using AI agents as my development team, and what that says about the future of technical work.
