← All articles

CASE STUDY : Search Engine Land Finalist ePPC Improved Atom Mobility’s Lead Quality by making an in house AI Lead Prediction Model

Atom Mobility’s Google Ads campaigns brought in plenty of demo requests. The problem was that most of them weren’t worth a sales call. Atom Mobility makes customizable mobile technology for vehicle sharing, rental and taxi businesses, and for them the real challenge was lead quality, not volume. At ePPC, we took on this challenge by developing a bespoke AI-driven lead prediction model that not only improved campaign performance but also became a finalist in the Search Engine Land Awards 2025.

The Lead Quality Problem

When Atom Mobility first partnered with us, they were seeing a strong flow of demo requests from their Google Ads campaigns. But there was a problem:

Meanwhile, 90% were low-quality or spam leads, and the long sales cycle slowed down campaign learning and optimization. That meant Google Ads could not learn quickly enough what worked, holding back performance growth.

Atom Mobility's lead funnel
From demo request to closed deal
Demo request
Calendly booking
Offline conversion
Deal closed
Offline conversions, the leads sales judged to be good quality, are the signal Google Ads needs to learn from.

Marketing goal

The company has two primary product verticals.

The Old Customer Journey

In this system, demo requests would go through the sales rep. for the assessment, and good quality leads would be sent to Google Ads. The assessment would take weeks or months. Thus, we would not get enough offline conversions within enough time to optimize our campaigns.

The old customer journey
Every lead waited for a manual assessment
Website
CRM
Sales representative
Google Ads

The New Customer Journey and Google Ads Optimizations

To tackle this, our team, including account managers and a data scientist, built a custom AI lead prediction model to assess incoming leads. This model automated the scoring of lead quality and worked as part of an updated customer journey.

Under the new system:

This meant campaigns could learn faster, with more accurate signals about which audiences and keywords were driving genuinely valuable engagement, not just clicks.

The new customer journey
The AI model scores every lead while sales feedback still follows
Website
CRM
Sales representative
AI lead prediction model
Google Ads

What We Optimized and Why It Mattered

With lead quality scoring feeding into campaign intelligence, we also executed a series of tactical improvements:

These combined efforts brought measurable gains across the board.

Real Results That Changed the Game

Once the AI model and strategy changes were in place, Atom Mobility’s performance metrics began to improve significantly:

Demo requests that became offline conversions
Share of all demo requests
Before10%
After22%
Before: June 2023 to May 2024. After: June 2024 to May 2025.
Offline conversions and their cost
Index, before = 100
Offline conversions
Before100
After170
Cost per offline conversion (lower is better)
Before100
After41
Before: June 2023 to May 2024. After: August 2024 to March 2025.

These results helped Atom Mobility not only make smarter decisions about where to invest marketing dollars but also speed up learning cycles that were previously constrained by slow, manual lead assessments.

Industry Recognition: Awards and Accolades

The work didn’t go unnoticed. The campaign was recognized at multiple industry award shows, including:

It is rare to see such a clear combination of AI innovation and tangible business impact in PPC campaigns, but Atom Mobility’s example shows what is possible when performance marketing teams embrace smarter data tools.

Team behind the success

The success of this project was driven by a cross-functional team at ePPC Digital with expertise spanning project leadership, account management, and data science.

Want this thinking applied to your account?