Most marketing organisations cannot forecast future returns with any analytical rigour — they extrapolate from historical averages. That limitation produces a permanent deficit in the board conversation: organisations defending past spend rather than justifying future investment.
The Forecasting Capability That Most Marketing Organisations Lack
The ability to forecast future marketing returns is not a capability that most Australian marketing organisations possess in any rigorous sense. Budgets are typically derived from historical spend levels, adjusted for inflation and strategic priorities, with return expectations based on historical performance averages. This is not forecasting. It is extrapolation — and the difference matters enormously when economic conditions shift, competitive dynamics change, or the organisation is evaluating a material change in budget level or channel mix.
Genuine marketing return forecasting requires three capabilities that are rarely present together. The first is a model of the relationship between marketing spend and business outcomes that captures non-linearity — the diminishing returns at high spend levels and the minimum effective frequency thresholds at low spend levels. The second is the ability to project that model forward using assumptions about the external environment — competitor activity, media cost inflation, category growth trends — that are explicitly stated and regularly updated. The third is the ability to quantify the uncertainty in the forecast and express it in a form that is useful for decision-making rather than paralyzing.
The organisations that lack forecasting capability face a specific and recurring problem: they cannot make credible forward-looking claims about the return on proposed marketing investment. This forces budget conversations into the past — defending what was spent, rather than justifying what should be spent on the basis of expected future return. The CFO’s evaluation of marketing investment is permanently reactive, because marketing cannot provide the prospective analysis that would enable a proactive conversation about capital deployment.
Why Historical Performance Data Is an Insufficient Basis for Forecasting
The instinctive response to a forecasting request is to extrapolate from historical performance data. If a channel delivered a certain return last year, the assumption is that a similar investment this year will deliver a similar return. This assumption is more often wrong than right, for reasons that are structurally predictable rather than random.
The Role of Marketing Mix Modelling in Enabling Forecasting
Marketing mix modelling, when properly constructed and regularly refreshed, provides the foundation for prospective return forecasting in a way that historical performance data alone cannot. The response curves generated by MMM — the mathematical relationship between spend level and output — can be used to simulate the expected return at different investment levels and channel mixes. This simulation capability transforms MMM from a retrospective diagnostic tool into a forward-looking planning instrument.
Organisations that cannot forecast future marketing returns cannot make a prospective capital allocation case. They are permanently defending past expenditure rather than justifying future investment.
The response curves derived from MMM allow planners to answer questions that historical data cannot: what is the expected return if the television budget is increased by 20 per cent and the search budget is reduced by 10 per cent? At what spend level does a given channel reach saturation, and what is the marginal return beyond that threshold? What is the minimum effective budget level for a channel to generate a positive return, given current market conditions? These are the questions that CFOs ask and that most marketing organisations cannot answer with any precision.
The accuracy of these forward projections depends critically on the recency and quality of the MMM. A model built on data that is 18 months old will produce response curves that reflect market conditions, media costs, and competitive dynamics that no longer apply. For forecasting purposes, MMM refreshes at least annually — and ideally biannually in categories with rapidly changing media markets — are necessary to maintain the model’s prospective utility.
Building Scenario-Based Forecasting Into the Budget Process
The most practically useful form of marketing return forecasting for budget purposes is scenario-based: a set of three to five alternative budget scenarios, each with an explicit set of assumptions about media costs, competitive conditions, and channel mix, and each with a projected return range expressed as a confidence interval rather than a point estimate. This format enables genuine budget deliberation rather than the defensive exercise of justifying a predetermined number.
The scenario format is important for a second reason: it makes the assumptions underlying the forecast explicit and therefore debatable. When marketing presents a single projected return figure, the conversation focuses on whether the figure is right or wrong. When marketing presents three scenarios with explicit assumption sets, the conversation focuses on which assumptions the board finds most plausible — a fundamentally different and more productive deliberation. The quality of the budget decision is improved not just by the modelling but by the structure of the conversation the modelling enables.
Forecasting Capability as a Budget Defence Asset
For marketing leaders navigating budget pressure, the ability to produce credible forward-looking return projections is among the most effective defences against arbitrary cuts. An organisation that can demonstrate, using a calibrated MMM-derived model, that reducing the marketing budget by 15 per cent will reduce revenue by 8 per cent over the following 12 months — with a specified confidence interval and explicit assumption set — has a fundamentally different budget conversation than one that can only argue from historical ROAS figures and year-over-year performance trends.
The investment required to build this forecasting capability is not trivial, but it is substantially less than the cost of making a series of poorly informed budget decisions based on inadequate evidence. For boards and CFOs, the practical implication is to ask marketing leadership not just what the current return on investment is, but what the projected return on the proposed investment level will be — and to demand that the answer be supported by a model that makes its assumptions explicit. The organisations that can answer that question well will receive a different quality of board engagement than those that cannot.