Skip to main content
Case studies
Case study

Case Study: How CDL Schools USA Runs on AI Operations

Updated 2026-09-14

Why this case study matters: most agency case studies describe work done for a client you can't verify, with numbers you can't check. This one is different, CDL Schools USA is our own venture, running in production right now on the same AI operations engine we build for clients. You can visit the site, inspect the content, and watch the systems work. This is what "we run what we sell" looks like.


The business

CDL Schools USA is a national platform for the commercial driving industry: a directory of CDL schools across the United States, DOT/FMCSA-aligned online training, and white-label training for trucking fleets. Three revenue lines, directory, training, B2B white-label, one operations team of effectively zero full-time staff.

The problem we built against

The CDL training market is large, fragmented, and badly served online: thousands of schools with terrible web presence, prospective drivers searching state-by-state with no authoritative resource, and fleets needing compliant training without building it themselves. Winning required three things that are normally staff-intensive:

  1. A directory with genuinely useful pages at national scale, thousands of school and location pages, each needing unique, accurate, maintained content.
  2. A content operation that could compete for search traffic across directory, training, and regulatory topics, dozens of interlinked articles, guides, and state resources, kept current as regulations change.
  3. Operations that didn't scale headcount with traffic, lead routing, school onboarding, student support, and fleet inquiries handled systematically.

A traditional build would have meant a content team, an ops team, and a support desk. We built systems instead.

What we built

The programmatic directory engine. Structured school data flows through templated page generation with unique-content rules, every school and state page carries genuinely differentiated content (local requirements, pricing context, program details), not find-and-replace text. Data updates propagate automatically; stale pages get flagged for refresh.

The AI content engine. Our content pipeline handles research briefs, drafting, internal-link mapping, schema markup, and publication prep, with human editorial review gates before anything ships. This is the system behind 87 identified content gaps turning into a structured injection roadmap and produced articles across the directory, training, and white-label business lines.

The measurement layer. GA4/GSC instrumentation with content-pillar tracking, so every page's performance feeds back into what the content engine builds next. The system reads its own results.

Lead routing & follow-up. School listing inquiries, student questions, and fleet white-label leads route automatically by type, with follow-up sequences for anything not immediately closed.

How it connects to what we sell

Every component above is a system we build for clients:

CDL Schools USA systemClient equivalent
Programmatic directory engineMulti-location / multi-listing page systems
AI content engine with editorial gatesContent operations for SEO-driven businesses
Measurement + feedback loopReporting dashboards and analytics instrumentation
Lead routing & follow-upThe same follow-up automation we install for contractors

The stack runs on infrastructure we own, built with the same Build & Transfer discipline we bring to client engagements, documented, monitored, and owned outright.

Honest lessons (what we'd tell a client)

  • Editorial gates are non-negotiable. AI-drafted content without human review eventually publishes something wrong. Every pipeline we build, ours or yours, keeps a human approval step where accuracy matters.
  • Programmatic pages live or die on unique value. The directory works because each page answers a real local question. Thin templated pages would have been penalized or ignored; the engineering effort went into differentiation, not volume.
  • The measurement loop is the product. The content engine improves because performance data feeds the next brief. Content operations without instrumentation are just publishing.

The proof is public

Visit cdlschoolsusa.com, the directory, the state guides, the training platform, all of it runs on this stack. If you want the same architecture applied to your business, directory, content operation, lead handling, or all three, book a fit call and we'll walk you through exactly how it maps.

Want the same stack for your business?

Thirty minutes, no deck. If a cheaper tool fits, we will say so on the call.

Related