Lead scoring on Ribbo: Get to know your prospective customers’ intent

Alongside Data Tools, Ribbo has built a full AI-enhanced lead scoring system that automatically classifies incoming conversations based on how prospective a customer is with regards to their inquiries.

Every visitor who chats with a Ribbo bot gets a score from 0 to 100 and a classification: Hot, Warm, or Cold. The scoring engine sends each conversation to an AI parser which scores each dimension independently from 0-100. The final score is then calculated with fixed weights in code; what the AI is good at (reading language) stays in the AI, what code is good at (consistent math) stays in code.

For example, if a visitor lands on your site at 11:00 PM asking about enterprise pricing, mentions they need implementation by next week, and provides their work email, the AI instantly recognizes high purchase intent, urgency, and budget authority. The system pushes their score to 90 (Hot - arbitrary figure*), automatically tags it for urgent follow-up, and fires off a notification so your sales rep sees it first thing in the morning.

The 5 dimensions along which the score is weighted are:

1. Purchase intent - a higher score is assigned for customers who are specific with quantities or a product they want to buy versus those just getting product information
2. Urgency - a higher score is assigned for customers who mention they would like to make a purchase as soon as today versus those with flexible timelines
3. Authority & budget - a higher score is assigned for customers who are decision makers with approved budgets (CFOs, CTOs) as opposed to those who are less senior and doing product research
4. Engagement quality - a higher score is assigned for customers who ask follow up questions and provide detailed requirements or use case needs, as opposed to those giving one-word responses
5. Sentiment - a higher score is assigned to customers who seem enthusiastic based on the words they use as opposed to those who use hesitant or negative language

Once these 5 scores are returned, they are weighted by a cloud function that applies fixed weights in code and summated to give a final lead score.

Behavioral boosts are also applied based on how many times the customer revisits the chat, whether they used WhatsApp channels (high intent signal), sharing contact details (commitment), requesting handoff to a human and more than 10 messages exchanged (deep engagement).

A rules engine is able to filter out bot traffic, internal testing conversations and also assign a direct Hot score when a customer requests for handoff to a human agent.

The full conversation history, contact details, dimension scores, conversation summary and final score are stored in Firestore and siloed to prevent unauthorized access.

Beyond just saving time for sales teams bogged down by endless qualification calls and manual data entry, this automated lead scoring approach means spending less time chasing unqualified leads and more time closing deals with people who are actually ready to buy. But the impact doesn't stop at sales. Customer success teams can quickly spot high-priority accounts needing attention, while marketing gains invaluable clarity on which campaigns are driving genuinely engaged prospects rather than just casual browsers. By automatically turning everyday chat interactions into clear, actionable insights, Ribbo aligns entire organizations around the people who matter most, ensuring no high-potential opportunity ever slips through the cracks.