Meta recently featured The Media Image on its global advertiser business homepage highlighting the agency’s work in developing a multi-agent “co-pilot” system designed to improve how teams access, interpret and act on performance data.
TMI’s inclusion reflects a broader shift in how performance marketing is evolving. As platforms become more complex and data volumes increase, the ability to translate information into action is becoming a defining capability. We thought you might like a more in-depth look.
From Data Access to Decision-Making
The challenge was not access to data. It was usability.
Performance teams, business intelligence functions and partner managers all rely on Meta’s data but the process of retrieving and interpreting that data is often fragmented. Queries sit across different interfaces, insights require manual assembly and recommendations are not always prioritised in a way that aligns with commercial impact.
Our objective was to simplify this.
The Media Image developed a multi-agent system designed to act as a single conversational interface: users input a query, retrieve relevant data via Meta’s APIs and receive structured, actionable outputs — all within a unified environment.
TMI’s focus was not on automation for its own sake, but on reducing friction between data and decision-making.
How the System Works
At a high level, co-pilot operates as a coordinated set of specialised agents, each responsible for a specific function.
- User Interface: A conversational front-end where users input queries, receive structured responses and provide feedback.
- Orchestration Layer: A central agent that interprets requests and routes them to the appropriate sub-systems, coordinating multi-step responses.
- Performance Insights Engine: Retrieves campaign, ad set and ad-level metrics via Meta’s Insights API, calculating and presenting relevant performance indicators.
- CAPI Troubleshooting Agent: Analyses data quality within Meta’s Conversions API (CAPI — the system used to pass event data from a website to Meta), identifying issues that may affect tracking accuracy and optimisation.
- Recommendation Engine: Surfaces and prioritises Meta-generated recommendations, including Advantage+ features designed to improve performance and efficiency.
Crucially, our system does not operate in isolation; it integrates with existing TMI tools, including our Integration Quality Dashboard, which consolidates event-level data across accounts.
This allows teams to move from identifying issues at a macro level to resolving them at a granular level — quickly and with context.
Driving Efficiency at Scale
The practical impact of the system is best understood through its effect on time and decision-making.
Manual data retrieval and analysis — previously a significant part of an analyst’s workload — is largely automated, allowing teams to spend more time on interpretation, planning and optimization.
The result is a shift in how resources are allocated within TMI.
- Analysts move from data gathering to insight generation
- Response times for technical issues are reduced
- Campaign adjustments can be made more quickly
This aligns with a broader industry trend. As platform-level automation increases, competitive advantage moves upstream — from execution to interpretation.
Measurable Impact
Meta highlighted the results of controlled A/B testing we conducted in December 2025. The co-pilot system delivered:
- 410% increase in landing page views
- 5x reduction in cost per landing page view
- 3.3x faster task completion
- 92% of tasks completed in under one minute
These figures reflect both performance gains and operational improvements.
In practice, this means:
- Campaigns reach more relevant users
- Budgets are used more efficiently
- Teams can act on insights faster
Beyond Efficiency: Expanding Capacity
Operational efficiency is not just about saving time. It changes what is possible. By reducing the time required to manage campaigns and resolve issues, TMI can:
- Support more accounts without increasing headcount
- Manage higher levels of spend
- Invest more time in quality strategic work
This has obvious direct commercial implications. Improved performance drives better results for our clients. At the same time, increased efficiency expands capacity, allowing TMI to scale without proportionally increasing costs.
We have already seen improvements in client experience: faster insights and more responsive optimization have contributed to stronger relationships and higher retention rates.
A System That Continues to Evolve
The system is now being rolled out across The Media Image’s client base, with further development underway. The next phase will focus on expanding the context that the system can interpret.
At present, co-pilot primarily works with structured data — metrics, signals and system outputs. Next, we want to incorporate unstructured inputs, such as creative content, enabling a deeper understanding of exactly what is driving performance.
This reflects a broader shift in marketing technology. As AI systems become more capable, the focus moves from retrieving data to interpreting it in context.
Differentiation Through Capability
As more platform processes become automated, differentiation is increasingly defined by how agencies like TMI use technology rather than simply whether they use it.
Tools like co-pilot enable us to:
- Move faster from insight to action
- Improve consistency across accounts
- Deliver additional value beyond standard platform capabilities
This positions The Media Image not just as an executor of campaigns but as a builder of systems that enhance performance.
A Long-Term Investment
The development of the co-pilot system is part of a broader commitment to innovation.
As adoption of Meta features such as Advantage+ increases, the ability to implement recommendations quickly and effectively becomes more important. Systems that support this process enable teams to focus on higher-value activity, including strategic planning and performance improvement.
Over time, this creates a compounding effect.
Better implementation leads to stronger results. Stronger results lead to increased investment. Increased investment creates more data, which further improves performance.
Perspective from The Media Image
As Tom Hutton, VP of Social at The Media Image, explains:
“We’re incredibly proud of the work our teams have done to design and implement these tools. Automating how we surface KPIs, identify CAPI issues and prioritise optimisation opportunities has fundamentally changed how we operate.
“It’s reduced the time spent on manual processes and allowed us to focus on the areas that drive the most value — strategy, insight and performance improvement.
“The result is faster decision-making, stronger outcomes for clients and a more scalable way of working.”
Linking Technology to Outcomes
The significance of this work is not the technology itself. It is what the technology enables.
By reducing the gap between data and action, co-pilot supports a more integrated, responsive approach to marketing. It aligns with broader themes explored across The Media Image’s work — from integrated channel strategy to data-led performance measurement — where the objective is to build systems that improve over time.
As the industry continues to evolve, the ability to connect data, insight and execution will become increasingly important.
This is one example of how that can be achieved in practice.



