The system analyzes every account every two hours and classifies every lead. An experienced CallRadius manager reviews its findings and makes every change.

CallRadius, a managed Google Local Services Ads (LSA) service based in Scottsdale, has released details of the patent-pending AI management system behind every account it runs, as Google moves Local Services Ads into Google Ads as Performance Max campaigns. The system runs 25 scheduled monitoring loops across each account, and every account is run by an experienced human manager who reviews the system’s findings and makes every change.

“We set out to build world-class intelligence for Local Services Ads, and then to put it in the hands of an experienced manager instead of letting it run on its own,” said Doug Brown, founder of CallRadius.

Marketing Technology News: MarTech Interview with Mark Listes, CEO @ Pendulum Intelligence

What the system does

According to the company, the CallRadius LSA AI management system:

– Syncs new leads every 30 minutes and classifies each one with AI. A lead is flagged to the manager as likely invalid only when the system is at least 85% confident.
– Re-analyzes each account’s budget every two hours, or 84 times a week.
– Tracks ad position and competitors about twice a week.
– Builds an hour-by-day model of 168 weekly time slots showing when leads arrive and when they turn into booked jobs.
– Compares location results ZIP code by ZIP code and flags weak areas for the manager to review.
– Keeps four figures separate that are often blended together: cost per charged lead, cost per booked job, the client’s monthly spend target and Google’s own budget setting.
– After a budget change, compares 14 full days before with 14 full days after and records the result as an association, not proof of cause.

Marketing Technology News: How MarTech Is Enabling Autonomous Brand Engagement Across Channels?

Managers also consult HELM, the company’s evidence-based AI expert reviewer for Local Services Ads, which went into production service on September 20, 2026. HELM reads a complete evidence snapshot of an account and states when the evidence is too thin to support a conclusion.

The technology is the subject of a pending U.S. provisional patent application (No. 64/063,539) with 67 claims, covering hierarchical budget optimization with an AI-driven lead-quality feedback loop, revenue-informed optimization, multi-location portfolio management and competitive intelligence.

A person makes every change

The system prepares recommendations and the evidence behind them. A manager reviews each one and makes every change to the client’s live Google account, from budgets to lead-quality feedback. The company said the list of rules allowed to change a live budget without a person is empty by design.

“Local Services Ads spend can’t be taken back,” Brown said. “Our system watches every account around the clock and reads every lead. No person could do that by hand. But a person who knows the business makes the call.”