Marko TomicFlagship 04 · Paid growth operating model
Performance & Growth
Rebuilding paid growth around the right specialist partner, trustworthy measurement and deliberate architecture - then augmenting analysis with autonomous, read-only AI.
Measured
Year-over-year improvement in return on ad spend - not a 371% ROAS.
Cost per click reduced year over year.
German-market Performance Max campaign covering bestseller categories, measured across a full year.
Context
Spend without an operating model
Paid acquisition was running, but the structure around it was weak: measurement that couldn't be fully trusted, no deliberate campaign architecture, no internal capability to direct the work commercially. The problem was not ad spend - it was how that spend was planned, measured and corrected.
Decision
Buy the specialist, build the system
The honest assessment: deep, daily Google Ads execution is a specialization, and building it in-house would have been slow and expensive. So the decision was to source and select the right external partner - and to own everything around them internally: strategy, tracking architecture, measurement, prioritization and commercial oversight.
That division of labor is the point. The specialist brings execution and optimization expertise; the internal role makes that expertise effective and accountable. Good leadership, not reduced ownership.
System
The evolution
- Agency / specialist selection me
- Tracking rebuild me
- Campaign / measurement architecture me + specialist
- Performance improvement specialist executes
- AI performance intelligence read-only, autonomous
- Human specialist execution external partner
The final two steps are the important ones: analysis was later augmented by the autonomous Google Performance Intelligence system - while execution authority deliberately stayed with the human specialist. AI reads; humans decide.
Ownership
Who owns what
I owned
- Identifying need for specialist capability
- Sourcing & selecting the partner
- Onboarding, reporting and decision structure
- Strategy & prioritization
- Tracking architecture
- Measurement environment
- Commercial oversight
External specialist
- Majority of direct campaign execution
- Optimization expertise
AI / system
- Autonomous cross-platform analysis
- Summaries & change highlights
Boundary
- AI never alters ad accounts - read-only by design
Outcome
Sequence mattered
Partner, then tracking, then architecture, then scale. The results - +371% YoY ROAS improvement, −29% CPC, a 1,127% annual ROAS on the German PMax bestseller campaign - came from that sequence, not from any single move. The later AI layer made weekly analysis nearly free.
Reflection
What the obvious solution gets wrong
The obvious move is to hire "a performance marketing person" and hand them the keys. What that misses is the measurement foundation: without rebuilt tracking, even excellent execution is steered by distorted instruments. Fix the instruments first - then scale the spend.
Related