Metaviz AI
AI Automation SaaS

Calling a cold list back to life

Outbound Database Reactivation Agent

An outbound voice agent that called more than 500 dormant event leads, asked what each business was struggling with, and sent the interested ones to a product demo.

Overview

A marketing company had collected over 500 leads at trade events. Six months later the list was sitting there untouched, because calling all of them by hand was never going to happen.

We built an outbound voice agent that works through the list on its own. It opens by referring to the event the person attended, asks how the business is going, and listens for what they are struggling with. If there is a real problem it can help with, it points them to the product demo and suggests a discovery call after they have watched it.

What the person says is captured as it goes. Trade type, the challenge they described, and how interested they sounded are stored in Supabase, so the sales team can start with the warmest names instead of the whole list.

Outbound Database Reactivation Agent - project overview
- Relevant keywords -
  • Outbound Voice AI
  • Lead Reactivation
  • Retell AI
  • Lead Qualification
  • Supabase
  • Sales Pipeline
  • Demo Booking
  • Call Automation
Challenge & Solution

Solving real problems with smart engineering

A list that had gone cold

Problem

Over 500 leads collected at events were sitting in a database with no follow up after the show.

Solution

The agent works through the whole list and opens every call by referring back to the event they came from.

Manual follow up that could not scale

Problem

Calling 500 people by hand takes weeks of someone’s time, so it never started.

Solution

Dialling and the first conversation are handled by the agent, with no one sitting on the phone.

Six months of silence

Problem

There was no system for reconnecting with a lead once the event was over.

Solution

The agent reopens the conversation and asks what has changed in the business since then.

No way to tell who was worth calling back

Problem

Every name in the list looked the same on paper.

Solution

Trade type, the stated challenge and interest level are stored during the call, so the sales team can sort the list before touching it.

How it works

How the agent works a list

The same seven steps run on every number in the database.

Step 1

Load the database

The event leads are loaded into the calling list.

Step 2

Outbound call

The agent calls each lead with a warm reconnection.

Step 3

Reopen the relationship

It refers to the event and asks how business has been.

Step 4

Find the challenge

It asks what the business is currently struggling with.

Step 5

Capture the answers

Trade type, challenge and interest are written to Supabase.

Step 6

Send to the demo

Interested prospects are pointed to the product demo.

Step 7

Score the lead

The stored answers let the sales team prioritise who to call.

Features

What it does

On the call

  • Opens by referring to the event the lead attended
  • Asks how the business has been since then
  • Draws out the challenge the business is facing
  • Explains the product in the context of that challenge
  • Points interested prospects to the demo
  • Suggests a discovery call after the demo

What it records

  • Trade type and business details
  • The challenge described in the caller’s own words
  • Interest level for prioritisation
  • Call outcome for the sales team to pick up
Gallery

Inside Outbound Database Reactivation Agent

Tech Stack

Built with

Voice

  • Retell AI outbound calling

Conversation

  • GPT-4.1 for dialogue logic

Data

  • Supabase for captured answers and lead scoring

Automation

  • n8n workflows
Results & Value

What actually changed

A list that nobody had time to call turned into a queue of qualified names, sorted by what each business actually said on the phone.

  • A dormant event database turned into a working pipeline
  • Every call captured in a form the sales team could sort
  • Manual dialling removed from the process
  • Prospects arrived at the discovery call having already seen the demo
01Headline outcome
500+Leads calledThe full event database, contacted without manual dialling.
02
40%+
Watched the demoOf the leads the agent qualified as a fit.
03
94%
Correctly qualifiedCalls where the captured trade type and challenge matched the lead.
04
87%
Positive responseLeads who engaged with the reconnection rather than ending the call.
05
200+
Qualified prospectsNames handed to the sales team with context attached.
06
500+
Hours savedTime that manual dialling and first conversations would have taken.

Have a similar project in mind?

Let's talk about how AI can transform your service business.

Talk to us about a system like this