Most media budget targets are set the same way every year: take last year's numbers, adjust for growth ambitions, allocate more to what the platforms say worked, and defend the result in a spreadsheet. It is a process that feels rigorous because it is structured. But the inputs are incomplete, and the targets reflect that gap.

The problem is not effort. It is the data those targets are built on.

Why last-click data produces the wrong budget

Last-click measurement, and the platform-reported ROAS that most brands rely on for planning, credits the final interaction before a purchase. That bias has a predictable effect: paid search and retargeting look like the strongest performers, because they sit at the point of demand capture. Paid social, video, and upper-funnel channels look weak, because they generate demand that converts elsewhere.

When a brand builds its 2026 budget on that picture, it doubles down on the bottom of the funnel and starves the channels doing the upstream work. Short-term ROAS improves, because the budget is now concentrated on people who were already going to buy. Over time, the pipeline of new customers narrows.

Les Binet and Peter Field's analysis of the IPA Databank, first published as The Long and the Short of It in 2013 and updated in Effectiveness in Context in 2018, consistently shows that brands allocating too little to demand generation underperform on long-term revenue growth, even when short-term conversion metrics look healthy. Across nearly 1,000 effectiveness case studies, the optimum budget ratio held at roughly 60% toward brand building and demand generation and 40% toward demand capture. Binet and Field note that the optimal ratio varies by category - online and direct-response markets tend to skew closer to 50:50 - but the underlying principle holds: most retail eCommerce brands are significantly over-indexed toward demand capture relative to what the evidence supports.

The difficulty is that most measurement tools make it nearly impossible to see the demand generation half of that equation clearly.

What does good budget-setting require?

Three inputs are needed, and most brands are missing at least two of them.

Causal measurement of incremental contribution. Incremental revenue is the revenue a channel caused, not the revenue it was present for. A channel that appears in 40% of conversion paths may be contributing far less in true incremental purchases - or far more, if it is generating demand that converts on a different channel entirely. Last-click measurement cannot reliably distinguish between the two, because it only credits the final interaction before purchase. Impression-led measurement, grounded in a Daily MMM that accounts for the full channel path, gives each channel the credit it is earning. Unlike platform-reported figures, incremental contribution reflects what each channel is genuinely causing - not just where conversions happened to be recorded.

Channel benchmarks. Internal data tells you how a channel is performing relative to your own history. It cannot tell you whether that performance is strong or weak relative to your category. A 3x ROAS on paid social may be above benchmark for a fashion brand and below benchmark for a beauty brand. Without competitive context, teams optimize against the wrong ceiling and miss channels their best-performing competitors are quietly scaling. Independent analysis from Varos, comparing ROAS across 7,000 non-Fospha brands with a combined annual media spend of $7 billion, found that Fospha clients outperform that market benchmark by 30%. That gap is consistent with brands allocating against incremental evidence rather than platform-reported correlation, though the comparison reflects overall client performance rather than a controlled test of allocation method alone.

Forward-looking spend forecasts. Budget targets need to hold through a year of changing conditions. The only way to pressure-test an allocation before committing to it is to understand the incremental return of spend at different levels per channel - seeing where headroom remains, and where you are already approaching diminishing returns. That replaces gut-feel planning with evidence of what each dollar is likely to return.

How do you translate this into a practical planning process?

The planning process that produces reliable 2026 targets has four stages.

Start with incremental baselines, not platform numbers. Pull channel-level incremental revenue contribution from your measurement model, not platform-reported ROAS. Unlike platform-reported figures, incremental contribution reflects what each channel is genuinely causing, not just where conversions happened to be recorded.

Apply category benchmarks to identify the gaps. Compare your channel ROAS and CAC against cohorted benchmarks for your category and revenue tier. The channels where you are significantly below benchmark are candidates for increased investment. The channels where you are above benchmark may be approaching saturation.

Use incremental forecasting to prioritize before finalizing allocations. Per-channel saturation curves show where spend has headroom to grow efficiently and where you are past the point of maximum return. Focus budget increases on channels with demonstrated headroom, not on spreading increases evenly.

Align marketing and finance on shared KPIs before the budget is locked. That alignment gap is one of the most common reasons budget plans stall between planning and execution. Finance needs to trust the numbers behind the allocation. That requires measurement that is transparent, validated, and explainable without a data science degree.

Where incremental forecasting changes the planning process

Beam, Fospha's incremental forecasting tool, is built for exactly the planning problem described above. Built on Fospha's always-on Daily MMM (Core), Beam uses Bayesian modeling and causal inference to forecast the incremental return of every dollar at different spend levels. It works by showing you, for each channel, where headroom remains and where spend is approaching diminishing returns - updated daily so the picture you plan from reflects this week's performance, not last quarter's.

Beam's Spend Strategist surfaces concrete budget guidance for each channel: which are positioned to scale efficiently, which should be approached cautiously, and which are already past their point of maximum return. Brands like Underoutfit used Beam's saturation curves to identify headroom in YouTube and TikTok, scaling YouTube spend 315% and TikTok spend 93% month-over-month while cutting blended CAC by 15% and generating $3.3 million in incremental revenue in a single month. Footasylum used Beam to identify that TikTok could support 2x daily spend while remaining above ROAS targets, and to give finance the causal evidence needed to release that budget.

Beam's outputs are grounded in the same validated measurement layer that finance can interrogate - closing the gap between what marketing believes and what the CFO will approve.

Common questions

Q: How do I know if I am over-investing in demand capture channels?

If your paid search and retargeting ROAS looks strong while your overall new customer acquisition rate is declining, that is a reliable signal of demand capture over-investment. Click-based measurement tends to over-credit bottom-funnel channels because it assigns all value to the final interaction - a bias that is well-documented across categories, though the degree varies by channel mix and purchase cycle. An impression-led Daily MMM will show the incremental contribution of each channel independently of where the conversion was recorded, and typically reveals that upper-funnel channels are driving more incremental revenue than click data suggests.

Q: What is the right brand-to-activation budget split for a retail eCommerce brand?

Les Binet and Peter Field's IPA research points to roughly 60% brand building and 40% activation as the optimum ratio for sustained long-term growth across most categories, though they note that online and direct-response markets tend to skew closer to 50:50. In practice, most retail eCommerce brands run considerably more heavily toward activation, because that is what click-based measurement makes look effective. Shifting toward a healthier balance requires measurement that can quantify the incremental contribution of demand generation channels, not just demand capture.

Q: Can I use Fospha's benchmarks before committing to the platform?

Fospha publishes aggregated benchmark data in its annual State of Retail Commerce report, which is available publicly. This covers channel ROAS, CAC trends, and spend allocation patterns across retail eCommerce categories. That data gives a useful starting point for assessing where your current allocation sits relative to high-performing peers, before any platform decision is made.

Q: How far in advance should we be running incremental forecasts for annual budget planning?

Incremental forecasting is most useful when run before budgets are locked, giving enough time to review saturation curves across your channel mix, align marketing and finance on the outputs, and adjust channel commitments with agency or platform partners. Running forecasts only after the budget is approved reduces them to a validation exercise rather than a planning input.

Related reading

‍