When you are spending over a million dollars in paid social alone, and comparable amounts across programmatic and search, the budget is never perfectly distributed. It cannot be. Campaign plans are built weeks or months in advance based on forecasts, historical performance, title release schedules, and market priorities. But the moment those campaigns go live, reality starts diverging from the plan. Some titles catch fire — a new Marvel series drops and the audience response is immediate, engagement spikes, cost per click drops, and the campaign is clearly capable of absorbing more spend profitably. Meanwhile, other titles plateau. The initial launch momentum fades, the audience saturates, and every additional dollar produces diminishing returns.

This is not a failure of planning. It is the nature of performance marketing at scale. The plan gets you to launch. What happens after launch is a function of real-time audience behaviour, competitive pressure, creative fatigue, and a dozen other variables that no forecast can fully predict. The question is not whether budget misallocation will happen — it will, every time. The question is how quickly you spot it and how decisively you act.

In the Disney+ operation, my official role was to execute. Set up campaigns, manage budgets to plan, deliver reports, maintain zero-error standards. But when you are inside the data every day, monitoring performance across platforms and markets, patterns start to emerge. And those patterns were telling a story that nobody was explicitly talking about: some of our money was in the wrong place.

The signals were not dramatic. They were the kind of thing you only notice if you are paying close attention across multiple dimensions simultaneously. On social, certain ad sets tied to specific Disney+ titles were showing consistent improvement in CTR over successive days — not just stable performance, but an upward trajectory that suggested the content was resonating and the audience was not yet saturated. At the same time, their CPA was trending downward, which meant each additional conversion was getting cheaper. These are the classic indicators of an underfunded campaign: strong performance signals with room to scale.

On the other side, I was watching ad sets that had been running at full budget for weeks start to show the opposite pattern. CTRs were flattening or declining, CPAs were creeping upward, and frequency numbers were rising — meaning we were showing the same ads to the same people more often, with less impact each time. These campaigns were not failing in any dramatic sense. They were still technically delivering. But they were past their peak efficiency, and every dollar spent there was worth less than a dollar spent somewhere else.

The gap between these two situations — the underfunded performers and the overfunded plateau — was where the opportunity sat. It was not a crisis. It was not a fire. It was the kind of slow, structural inefficiency that burns money quietly if nobody is looking for it.

I could have simply flagged these observations in a report and moved on. That was technically what my job required. But I wanted to do something more useful than surface a problem — I wanted to present a solution.

I started pulling data across platforms and markets, building a view of where budget was producing the highest return and where it was producing the lowest. This was not a simple spreadsheet exercise. The metrics are different across platforms — Meta reports in one way, TikTok in another, DV360 uses entirely different attribution models. I had to normalise the data enough to make meaningful cross-platform comparisons while respecting the fact that a dollar on Meta and a dollar on DV360 are doing fundamentally different things in the funnel.

From this analysis, I built specific reallocation recommendations. Not vague suggestions like “we should move some budget from Campaign A to Campaign B.” These were structured proposals: move this specific amount from this ad set to that ad set, shift this market's programmatic allocation toward social for the next two weeks, reduce spend on this title's always-on campaign and redirect it toward the new title launch that is showing stronger early signals. Each recommendation came with a rationale grounded in the data — the trend lines, the efficiency metrics, the audience saturation indicators.

In a global agency operation, you cannot just move money around because you think it is a good idea. Every budget change on an account like Disney+ goes through a chain of approvals. The offshore strategy team needs to review and agree. The client needs to be informed and aligned. The execution needs to be coordinated across platforms and markets so that nothing breaks in the process.

I presented my recommendations to the offshore team through a combination of structured reports and direct conversations. The key was framing the proposals not as criticism of the existing plan but as responsive optimisation — the plan was good, the data has since told us where to adjust, and here is a specific way to capture more value from the same total budget. This framing mattered. Nobody wants to hear that their plan was wrong. Everyone wants to hear that there is an opportunity to make it work harder.

The recommendations were reviewed, discussed, and — in several instances — approved and implemented. The budgets moved. The campaigns that received additional funding continued to perform. The campaigns that were scaled back did not suffer — they simply stopped burning money in a space where the returns had already been captured.

The reallocation led to measurable improvements in both CTR and CPA across the affected campaigns. The same total budget, distributed more intelligently, produced better results. The campaigns that were underfunded responded to the additional spend with continued strong performance. The campaigns that were overfunded did not lose meaningful volume when scaled back, confirming that the excess spend was indeed past the point of diminishing returns.

Budget reallocation sounds technical and unglamorous. It does not have the narrative appeal of a creative overhaul or the novelty of building an AI agent. But I would argue that this is the most fundamental skill in performance marketing: the ability to look at a large, complex, multi-platform media operation and identify where the money could be working harder.

It requires three things. First, you need to be inside the data deeply enough to spot patterns that are not obvious from a summary dashboard. Second, you need the analytical rigour to build a case that is specific, defensible, and actionable — not just a gut feeling. And third, you need the stakeholder management skills to navigate the approval process, frame the recommendation constructively, and get it implemented.

This was not a one-time exercise. It became a recurring part of how I approached the work. Every week, I was looking at the data with a dual lens: is everything executing correctly (operations), and is everything allocated correctly (strategy)? That dual perspective — execution plus optimisation — is what I believe separates a good analyst from someone who is just running campaigns.