Case studies

Doubled. Tripled. Ceilings that held for years, broken.

Every dossier below is a real result from a real business: doubled, tripled, ceilings broken, engines launched from zero. Client names are held in confidence; the numbers speak plainly enough.

Days of analysis became minutes. Record volume at record-low acquisition cost.

Multi-brand consumer portfolio · $2B+

Every data source unified into one intelligence layer (analytics, ad platforms, attribution, email, app), personal data stripped at the connector by design. AI agents took over the analyst work: budgets reallocated continuously against hard guardrails, ad quality rebuilt and verified across hundreds of ad groups, anomalies flagged to the team before they cost money. Cost per click fell roughly 40% year over year while record order volume arrived at record-low acquisition cost. The spreadsheets were retired.

A growth ceiling that held for years, gone. Record revenue. Record new customers.

Consumer business · $2B+

A mature business of this size isn't supposed to find new gears. New-customer acquisition had plateaued for years, and everyone had a theory about why. I rebuilt the acquisition system instead of debating the theories. The ceiling broke: record revenue and record new-customer volume in the same year. The back half of the first six months added twice the new customers of the front, and it compounded from there.

Chronic laggards became the portfolio's fastest growers. Several doubled. Some tripled.

Multi-brand consumer portfolio

I proved the playbook on the flagship first, then ran it across the sibling brands: same system, tuned per brand. The brands that had trailed the portfolio for years became its fastest growers. That's the test of a real system: it works more than once.

Zero to a working acquisition engine. A new revenue line where none existed.

Enterprise · co-branded financial product launch

No customers, no funnel, no playbook: a brand-new product category for the business. I built the go-to-market from scratch: positioning, channels, acquisition mechanics, the full engine. It launched, it acquired, and it became a revenue line the company didn't have before.

A dozen-plus channels piloted. Failures killed in weeks. Winners scaled into permanent infrastructure.

Multibillion-dollar business

Every channel pilot launched with kill criteria written before a dollar was spent. The losers died fast and cheap: weeks, not quarters. The winners were systematized, scaled, and handed off as permanent acquisition channels, including programs spanning hundreds of ad groups. Discipline on the way in is what makes scale possible on the way out.

The 'failing' channel was the cheapest source of new customers in the building. The data was lying.

Consumer business · $2B+

Click counts tripled overnight with no movement in cost or sales, and the apparent conversion-rate collapse nearly got a top acquisition channel shut down. I traced the phantom traffic to platform prefetching, proved it with zero-second engagement residuals, and verified every real order against matched identities. The channel stayed, at roughly twice the new customers per dollar of anything else. Bad measurement kills more growth than bad ads.

Growth engineering

Your ceiling is next.

Start a Conversation