WhatsApp
Studies show that 44% of salespeople give up after just one follow-up, while 80% of sales require five or more touches. The problem is not motivation, it is logistics. When you are sending dozens of quotations weekly, manually tracking which ones need follow-up becomes a full-time job.
We recently helped a B2B company solve this exact problem by turning Gmail labels into an automated follow-up scheduling engine.
The Challenge
Our client, a B2B equipment supplier, was sending 30-40 quotations per week. Each required a follow-up if no response came within 7 days.
The manual process involved calendar reminders, spreadsheets, daily checks, and manually composing follow-ups. The result: at least 15-20% of quotations never received a follow-up because the tracking system was fragile.
The pain points:
- 10-15 hours weekly spent on follow-up administration
- Quotations slipping through the cracks
- Inconsistent follow-up timing
- No visibility for management into pending quotations
The Solution: Gmail Labels as a Scheduling Engine
Instead of building a complex CRM integration or purchasing expensive sales automation software, we developed a solution using tools the team already had: Gmail and Google Workspace.
The core insight: Gmail labels can act as a visual countdown system. By automatically progressing emails through a series of labels (from “7 Days” to “Today”), we created a self-managing follow-up queue.
How It Works
Intelligent Detection When the sales team sends a quotation, the AI analyzes the thread to confirm it needs follow-up. It distinguishes between our quotes (need follow-up), incoming RFQs (no follow-up), invoices, and order confirmations.
Label-Based Countdown Each quotation receives a “7 Days” label. Every morning, labels automatically progress: 7 Days becomes 6 Days, then 5 Days, until it reaches “Today.”
Smart Reply Detection If the customer responds at any point, the system removes follow-up labels and cancels the scheduled email. No embarrassing “just following up” messages after a customer has already replied.
Automated Follow-up Sending When a quotation reaches “Today,” the system sends a follow-up email with proper threading, original attachments preserved, and a professional template.
Activity Logging Every action is logged to a Google Sheet, giving management complete visibility into quotation status.
AI-Powered Email Intelligence
The original rule-based system worked well, but had limitations. It required exact pattern matches: specific subject lines, particular links in the body, attachment presence. Valid quotations that did not match the pattern were missed, while incoming RFQs (requests for quotes) were sometimes incorrectly flagged.
The Solution: LLM Classification
We integrated Google’s Gemini 2.0 Flash model to intelligently classify emails. Instead of rigid pattern matching, the AI analyzes the full thread context to understand intent.
How the AI Classification Works:
- Analyzes the last 5 messages in each email thread
- Identifies who sent what (sales team vs customer)
- Determines if a quotation is waiting for response
- Classifies the email type: quote, invoice, order, RFQ, inquiry, or general
- Returns a confidence score (0-100%) with reasoning
- Suggests appropriate follow-up timing (1-7 days)
Email Type Detection:
| Type | Description | Follow-up? |
|---|---|---|
| Quote | Quotation we sent | Yes |
| Invoice | Invoice for services | No |
| RFQ | Request for quote (incoming) | No |
| Order | Order confirmation | No |
| Inquiry | General question | Maybe |
Multi-Language Support
The AI naturally handles emails in English, Dutch, German, French, and Spanish without any configuration changes. This was critical for the client’s international customer base.
Two-Stage Architecture
For reliability and cost efficiency, we implemented a two-stage approach:
- Rule-based pre-filter (fast, free) - Quick checks for obvious cases
- LLM classification (intelligent, ~$0.20/month) - Deeper analysis when needed
If the AI service is unavailable, the system automatically falls back to rule-based detection. No follow-ups are ever missed.
Cost: Under $3/Year
Using Gemini 2.0 Flash pricing, the AI classification costs approximately:
- 10 emails/day: $0.48/year
- 50 emails/day: $2.40/year
- 200 emails/day: $9.60/year
This is a fraction of the cost of any commercial sales automation tool.
Technical Implementation
The system runs entirely within Google Workspace:
- Google Apps Script handles the automation logic with time-based triggers
- Gemini 2.0 Flash provides intelligent email classification via API
- Hierarchical labels (
FollowUp/7 Days,FollowUp/6 Days, etc.) keep the inbox clean - Google Doc templates let the team update messaging without touching code
- Automatic fallback to rule-based classification if AI is unavailable
- Error handling flags failures for manual review rather than failing silently
The Results
After three months:
- 10+ hours weekly recovered from follow-up administration
- 100% of quotations now receive timely follow-ups
- 40% improvement in quotation response rates
- Zero additional software cost since it runs on existing Google Workspace
How RAD Digital Solutions Can Help
This follow-up system is one example of how targeted process automation transforms daily operations. We specialize in identifying high-impact automation opportunities and building solutions that fit naturally into existing workflows.
Our approach:
- Process Assessment - Analyze workflows to identify automation candidates
- Solution Design - Use tools you already have where possible
- Implementation - Build, test, and deploy the automation
- Training & Handoff - Ensure your team can use and maintain the system
Whether you are dealing with email follow-ups, data entry, report generation, or other repetitive tasks, there is likely an automation opportunity waiting.
Explore our Process Automation services to learn more. Ready to discuss your specific challenges? Contact us for a free consultation.
