Case Study: An Advertising Intelligence Company Found Thousands of Untapped Leads

An anonymous Canadian advertising intelligence company received thousands of untapped leads with full contact information and campaign assets from a human-led lead intelligence engine.

2026-06-03 — Craig Major

An anonymous Canadian advertising intelligence company needed a repeatable way to find and organize untapped leads with complete contact information. Flowgrammer built a lead intelligence engine that surfaced thousands of previously untapped leads and prepared campaign assets for human review and outreach.

This case study is approved for publication with the client anonymous. The result is a lead-discovery and campaign-preparation result. It does not claim a reply rate, meeting rate, conversion rate, or closed revenue because those downstream figures were not supplied for publication.

The client and starting problem

The client operates in advertising intelligence. Its team needed a larger, more usable pool of potential accounts than the lists it was already working from. The approved starting problem was finding untapped leads, resolving contact information, and turning research into campaigns the team could actually review and use.

The baseline was measured by the delivered lead database and campaign assets. No approved before-count, conversion baseline, or revenue baseline is claimed here.

What Flowgrammer built

The system was designed as a lead intelligence workflow rather than a static list purchase. Its major parts were:

1. Signal discovery: find relevant companies and contacts from approved sources and market signals.
2. Record enrichment: assemble available company and contact information into a usable record.
3. Deduplication and organization: reduce repeated records and structure the pool for review.
4. Segmentation: organize leads into campaign-relevant groups rather than one undifferentiated export.
5. Campaign preparation: create campaign assets and supporting context so a human could review the audience and message.
6. Human review path: keep the final decision about fit, message, consent, and outreach with the client’s team.

The output was a working lead engine and campaign-ready material—not a claim that every discovered record was qualified or ready to contact.

What stayed human

The client retained control of:

• the definition of a good account and the priority segments;
• the offer, positioning, and campaign message;
• consent and outreach decisions;
• exceptions, sensitive accounts, and relationship context;
• the decision to pursue, hold, or reject an individual lead;
• downstream sales conversations and revenue attribution.

The system handled discovery, organization, enrichment, and preparation. People created the commercial judgment.

Result and evidence boundary

The delivered system found thousands of untapped leads with full contact information and campaign assets. The claim is supported by the delivered database and campaign materials; the evidence register records the approval and attribution limits.

What we are not claiming:

• a specific number beyond “thousands”;
• a guaranteed accuracy or deliverability rate;
• replies, meetings, pipeline, or closed revenue;
• that every record should be contacted;
• that the system replaced the client’s sales or marketing judgment.

The distinction matters. A lead engine can improve the supply and preparation of opportunities; the client’s offer, consent, message, follow-up, and sales process determine what happens next.

Why this is an AI Automation System

The value was not in a single scraper or enrichment call. It was in coordinating discovery, record handling, segmentation, campaign preparation, and human review around a business outcome. A production system needs ownership and rules at every boundary: source permissions, record quality, duplicate handling, confidence, and outreach approval.

What we would measure next

To extend the baseline, the client could track:

• percentage of records passing a human quality check;
• duplicate and missing-field rates;
• time from signal discovery to review-ready record;
• campaign launch rate by segment;
• reply and meeting rates by campaign;
• qualified pipeline and closed revenue attributed through the client’s CRM.

Those are next measurement opportunities, not results of this case study.

Discuss a similar system

If your team has a repeatable lead, research, or routing problem, an AI Automation System can be scoped around the workflow, evidence, human review, and measurement—not a generic promise of “more leads.” If the opportunity is still unclear, start with an AI Success Audit.

Related reading

• What stayed human in a lead intelligence system—and why
• How to choose an AI automation agency

Continue with Flowgrammer

• Explore the systems behind this work through AI Automation Systems.
• Review the companion US manufacturing outbound system.
• Identify your best-fit audience with the ICP Generator.
• Book an AI Automation Systems conversation.