Agentic AI Campaign Beats Traditional DSP: PubMatic's Geloso Beverage Group Case Study
Agentic AI autonomously planned and executed a beverage campaign that outperformed traditional DSP methods across all metrics
Agentic AI Campaign Beats Traditional DSP: PubMatic's Geloso Beverage Group Case Study
PubMatic’s AgenticOS platform planned and executed an entire advertising campaign for Geloso Beverage Group without using a traditional demand-side platform (DSP) or manual optimization desk, outperforming the original plan across every major metric.
Why This Matters Today
Marketing leaders face increasing pressure to demonstrate ROI while managing complex, multi-channel campaigns. This case proves agentic AI can deliver better results with less human intervention, directly impacting marketing efficiency and effectiveness.
The Agentic AI Approach
Using PubMatic’s AgenticOS platform launched in January 2026, the campaign was fully autonomous:
- Planning: AI agents interpreted Geloso’s objectives, guardrails, brand-safety requirements, and creative parameters from natural language instructions
- Execution: Agents handled media buying, placement, and real-time optimization without human oversight
- Optimization: Continuous self-improvement based on performance data, eliminating the need for manual optimization desks
Traditional DSP Limitations
Conventional campaign management requires:
- Human traders to interpret briefs and set up campaigns in DSP interfaces
- Manual optimization desks making periodic adjustments based on delayed reporting
- Multiple handoffs between planning, execution, and analysis teams
- Latency between performance insights and optimization actions
Results Comparison
| Metric | Agentic AI Result | Traditional DSP Baseline | Improvement |
|---|---|---|---|
| Campaign ROI | 1.8x plan | 1.0x plan | +80% |
| Cost per Acquisition | 35% below target | 15% above target | 50% better |
| Brand Lift | 22% increase | 12% increase | 83% better |
| Optimization Speed | Real-time | Daily/weekly | 24x faster |
| Human Intervention | 0 hours/week | 15 hours/week | 100% reduction |
Key Enablers
AgenticOS succeeds where simpler automation fails because it:
- Understands intent: Uses LLMs to interpret complex marketing briefs accurately
- Operates autonomously: Makes decisions within defined guardrails without constant approval
- Learns continuously: Improves performance through reinforcement learning from real-time results
- Integrates natively: Works directly with ad exchanges and publisher APIs
Risks and Considerations
While results are impressive, marketing leaders should note:
- Brand safety: Requires robust constraint definitions to prevent off-brand placements
- Transparency: AI decision-making can be opaque; audit trails are essential
- Skill shift: Teams need to transition from media buying to AI supervision and strategy
- Vendor dependence: Performance is tied to the specific agentic platform capabilities
The Bottom Line
For CEOs and CMOs questioning whether agentic AI delivers tangible marketing advantages, this case provides clear evidence: autonomous agents can outperform traditional methods while reducing operational overhead. The Geloso Beverage Group campaign suggests agentic AI isn’t just experimental—it’s production-ready and delivering superior ROI today.
For guidance on implementing agentic AI in your marketing operations or assessing vendor capabilities, contact admin@infomly.com
Stay ahead of the AI shift
Daily enterprise AI intelligence — the decisions, risks, and opportunities that matter. Delivered free to your inbox.