How to set a marketing budget.

/ 01The problem: why most marketing budgets are wrong

Marketing budget setting is one of the most consequential decisions a business makes, and one of the least rigorous. Most companies set their marketing budgets using methods that bear no relationship to what the business actually needs to achieve. The result is chronic underinvestment dressed up as prudent management, or spending that feels generous but is allocated to the wrong things.

The Gartner 2025 CMO Spend Survey found marketing budgets have flatlined at 7.7% of company revenue for the second consecutive year, down from 9.1% in 2023. Fifty-nine percent of CMOs say they have insufficient budget to execute their strategy.

The IPA’s 2024 research (Ian Whittaker, Liberty Sky Advisors) found that 52% of financial analysts view a marketing spend cut as a positive cost-saving measure, while only 36% recognise it as short-termism with long-term consequences. By contrast, 47% viewed an R&D cut negatively. Marketing is held to a lower standard of protection than R&D, partly due to a lack of understanding of how it works, and partly because marketers haven’t been rigorous enough in how they set and defend their budgets.

/ 02The six common methods

2.1 — Percentage of revenue

The most common method. Take last year’s revenue, apply a percentage (typically 5–10%), and that’s the budget. Industry benchmarks abound: SaaS firms spend 15–25%, consumer goods 10–20%, B2B services 5–10%.

What’s right about it: it’s simple, fast, and gives the CFO a number they recognise. It provides a rough sanity check against industry norms. For SME’s, it is often the most practical method.

What’s wrong with it: it’s backwards. It makes marketing a consequence of revenue, when marketing should be a driver of revenue. It assumes the past is an accurate predictor of what the business needs next and it almost never is. It ignores competitive context entirely: two companies with the same revenue in the same category may need wildly different budgets depending on their relative market share and business objectives.

Verdict: useful as a sanity check. Easy and practical to apply. Uses data available to any business. Dangerous as a primary method. It answers “what can we afford?” not “what do we need?”

2.2 — Competitive parity

Match what competitors spend. The logic is that if the industry has arrived at a certain level of spend, it must reflect some collective wisdom.

What’s right about it: it acknowledges that marketing is a competitive activity, not a standalone line item, and it’s useful for benchmarking.

What’s wrong with it: it assumes your competitors know what they’re doing. They often don’t. Companies of different sizes, strategies and life stages need different investment levels. And it’s inherently defensive: it maintains the status quo rather than enabling growth.

Verdict: a useful input, not a method. Know what competitors spend, but don’t let it set your budget.

2.3 — The “affordable” method

Set the budget based on what’s left after all other costs are covered. Also known as the “what’s in the tin” method.

What’s wrong with it: everything. This treats marketing as a residual, not an investment. It guarantees underinvestment during periods of growth opportunity and over-cutting during downturns which is precisely the opposite of what the evidence says you should do.

Verdict: the worst method available. If this is how your budget is set, the method is your first problem.

2.4 — Zero-based budgeting (ZBB)

Rebuild the budget from zero every cycle. Every programme must earn its funding through a business case. No historical entitlement. No ‘copy and paste’.

What’s wrong with it: Often the issue is the application, moreso than the method. It’s often used as a cost management tool, not a growth tool. In this application, it answers “are we spending efficiently?” but not “are we spending enough?” Applied without a growth framework, it becomes a euphemism for cuts — the Kraft Heinz disaster is the canonical example (see the cautionary tales). Brand building is hard to justify in annual business cases because its effects compound over years, not quarters; ZBB structurally biases against it. And it’s resource-intensive: without access to the right data, it can be hard to do.

What’s right about it: it forces rigour, eliminates legacy waste, and makes marketers justify every line item. Done as a planning discipline aligned with a proper strategy, it works: when Unilever adopted zero-based budgeting in 2016, Ritson publicly defended the move as "a better, more strategic way to plan your marketing" — and Unilever's brands did not suffer Kraft Heinz's fate. Same method, opposite application.

Verdict: excellent as a discipline for allocation and efficiency within an already-determined budget envelope, and when used in a strategic manner. Dangerous when just used as a way to cut costs.

2.5 — Objective and tactics

Define your marketing objectives, identify the tactics required to achieve them, estimate the cost of those tactics, and the sum becomes the budget. The most rational method in textbooks.

What’s right about it: it links spending to outcomes, forces strategic clarity, and makes the budget a function of strategic objectives, not inertia.

What’s wrong with it: it assumes you can work out what it costs to achieve each objective, which requires data most companies don’t have. Without econometric modelling to estimate the cost of objectives, it often defaults to guesswork dressed in strategic language.

Verdict: the right philosophy. But it needs a quantitative backbone (ESOV + econometrics) to become a practical method.

2.6 — Share of voice / excess share of voice (SOV/ESOV)

Binet and Field’s analysis of the IPA Effectiveness Databank established that market share growth is driven by the relationship between share of voice (SOV) and share of market (SOM). When SOV exceeds SOM, brands grow. When SOV falls below SOM, brands decline. The difference, excess share of voice (ESOV), is the single most reliable predictor of market share change in the evidence base. It should be noted that the effectiveness databank is a collection of award submissions, typically larger B2C companies with larger budgets. There are some criticisms and limitations of the dataset, and how much emphasis is placed on the insights derived from it, is a decision for each marketer to make. In a practical sense, as marketers we are used to, and need to be able to work with, imperfect data.

The data: an ESOV of 10 percentage points predicts approximately 0.7% market share growth per year. The relationship holds across categories, geographies, and B2B/B2C contexts (though the multiplier varies).

What’s right about it: it’s grounded in the largest empirical evidence base in marketing effectiveness. It connects budget to competitive context, what you need depends on where you are relative to competitors, not just your own revenue. It provides a clear, defensible logic: “we need X share of voice to grow Y market share, and that costs Z.” And it naturally protects against underinvestment during downturns, when competitors cut, your SOV increases at lower cost.

What’s wrong with it: SOV data is not free or easy to get, it requires category-level advertising expenditure data (from sources like Nielsen, WARC, or Pathmatics). The 0.7% per 10-point ESOV is an average; it varies by category, creative quality, and channel mix. It’s a planning heuristic, not a guarantee. And it tells you how much to spend, but not where to spend it.

Verdict: a strong, evidence-based method available for setting the total budget. It answers the right question: “what do we need to invest to grow at the rate we’ve set?” In reality, not as practical for smaller businesses.

/ 03The recommendation: a composite, evidence-based approach

There is no single method. The optimum approach is a composite that uses the right method at each stage of the budget-setting process. After reviewing the evidence base, the IPA Effectiveness Databank (Binet & Field), WARC’s Anatomy of Effectiveness, Ehrenberg-Bass Institute research, the IPA’s Marketing is an Investment report (Whittaker, 2024), Mutinex/Mi3 MMM research, and Ritson’s Triple-Cooked Budgets framework. The recommendation is a three-stage model.

Stage 1: set the total budget using ESOV

Start with share of voice relative to share of market. This determines the total investment envelope. The logic is simple: if your SOV is below your SOM, you are harvesting, not growing. If you want to grow, your SOV must exceed your SOM by a meaningful margin.

• Determine your current SOM from industry data, internal sales data, or credible estimates.
• Determine category-level advertising expenditure (Nielsen Ad Intel, WARC, Pathmatics, or industry reports).
• Calculate current SOV = your ad spend ÷ total category ad spend.
• Set target ESOV based on growth ambition. Rule of thumb: 10 points of ESOV ≈ 0.7% share growth per year (Binet & Field). Adjust for category dynamics and consider annualisation.
• The required ad spend = target SOV × total category ad spend.
• Cross-reference against % of revenue (5–10% range per Grace Kite / Ritson) as a sanity check, not a ceiling.

Stage 2: allocate between brand and activation using the 60/40 heuristic

Once you have the total budget, split it between long-term brand building and short-term sales activation. Binet and Field’s analysis of the IPA Databank consistently shows the optimal split is approximately 60% brand, 40% activation for the average B2C brand. For B2B, the weighting shifts to approximately 46% brand, 54% activation.

This is a starting point, not a rule. The right ratio depends on brand maturity, category dynamics, and business lifecycle. A startup with zero awareness may need to skew toward activation early and shift toward brand as it scales. A mature brand defending share may need to over-index on brand to maintain mental availability.

Key principle: long-term effects are not the accumulation of short-term effects. Brand builds pricing power and mental availability. Activation harvests existing demand. You need both, and they work through different mechanisms.

Stage 3: optimise channel allocation using econometrics (MMM)

With the total budget set and the brand/activation split determined, the final stage is channel-level allocation. This is where marketing mix modelling (econometrics) earns its keep.

MMM connects marketing investment to revenue and profit outcomes by modelling all paid, owned and earned media, pricing, promotions, distribution, CRM, events, and macro variables. It identifies diminishing returns curves by channel, incremental contribution (what actually caused growth vs what took credit for it), and budget reallocation scenarios.

Why MMM and not last-click attribution? Platforms that sell advertising also grade whether that advertising worked. That’s a fundamental conflict of interest. Last-click attribution ignores anything without a digital footprint and anything that happened before the last click and has systematically biased budgets toward short-term performance channels at the expense of brand building. MMM is the independent, unbiased alternative.

/ 04How to do it: the practical playbook

Theory is easy. Implementation is where most budget-setting frameworks die. Here is a step-by-step playbook that assumes you have access to the resources required for the full composite model.

Step 1 · Build your competitive context picture

• Source category advertising expenditure data (your media agency, Nielsen Ad Intel, WARC Adspend Database, Pathmatics/Sensor Tower, or industry body reports).
• Calculate your current SOV and SOM. If exact data is unavailable, use credible proxies — James Hankins’ Share of Search method uses Google Trends branded search volume as a proxy.
• Map competitor SOV positions. Who is over-investing? Who is harvesting?

Step 2 · Set your growth target and required ESOV

• Agree with the business on a market share growth target.
• Apply the ESOV multiplier (10 points ESOV ≈ 0.7% share growth, adjusting for category). Work backwards to the required SOV.
• Convert required SOV into a monetary figure using category total spend data.

Step 3 · Apply the brand/activation split

• Default: 60/40. Adjust based on brand maturity and competitive situation. Tracksuit Brand Tracking have a free calculator on their website which takes into consideration the variables.
• Measure each piece differently: ROI for activation, brand tracking metrics (mental availability, unprompted awareness, consideration) for brand.

Step 4 · Implement MMM for channel optimisation

• Engage all stakeholders (CEO, CFO, data, sales, agencies) before implementation begins.
• Resolve what outcome you’re modelling (sales vs leads vs revenue) upfront.
• Audit and coordinate internal data sources before onboarding.
• Treat MMM as a business intelligence tool, not a marketing tool — it will surface pricing and product issues, not just channel mix answers.
• Use it to have conversations with the C-suite using hard revenue and profit numbers, not marketing proxies.

Step 5 · Apply ZBB as a downstream efficiency tool

• Within the allocated budget, apply zero-based discipline to every programme and campaign.
• Every line item must earn its place. But the total envelope is set by ESOV, not by ZBB.
• This is where ZBB earns its keep — as a waste elimination tool within a growth-oriented framework.

Step 6 · Defend the budget using investment language

• Frame every conversation as investment, not spend. The IPA research shows this language shift materially affects how boards perceive marketing budgets.
• Build a DCF (discounted cashflow) model for brand investment where possible. The IPA report provides a worked example showing £100/year marketing spend generating a total present value of £4,189 over a sustained period.
• Connect marketing to CFO priorities: increasing cashflow (46% of CFOs’ priority) and expanding markets (32%) — not just “brand awareness.”

/ 05The SME alternative: when you don’t have an econometrician on speed dial

The composite model above requires resources that many small and medium-sized businesses don’t have: MMM platforms, category advertising expenditure data subscriptions, and dedicated analysts. That doesn’t mean SMEs should fall back on gut feel. Here is the best available alternative: Ritson’s Triple-Cooked Budgets method, adapted with Share of Search as a proxy for SOV/SOM.

Step 1 · Set the total budget at 5–10% of revenue

Grace Kite’s analysis (endorsed by Ritson) of three independent data sources shows that 5–10% of turnover on advertising delivers the highest ROI for most businesses. Ritson recommends aiming for 10% for maximum effectiveness and competitive advantage. For an SME in a growth phase, treat the lower bound as a floor, not a target.

Step 2 · Use Share of Search as your ESOV proxy

You can’t afford Nielsen Ad Intel. But you can use Google Trends. James Hankins’ Share of Search method (championed by Les Binet) uses branded search volume as a proxy for market share. Track your share of branded search queries against competitors over time. If your share of search is declining, your brand is weakening, regardless of what your performance dashboards say.

This won’t give you the precision of true SOV/SOM data, but it’s free, it’s directional, and it’s a vastly better signal than “we spent 8% of revenue last year so let’s do 8% again.”

Step 3 · Apply the 60/40 split (adjusted for your stage)

SMEs and startups should start activation-heavy, closer to 20–30% brand, 70–80% activation and shift toward 60/40 as they grow. The key insight: don’t ignore brand entirely. Even 20% of a small budget, consistently invested in emotional, broad-reach creative, compounds over time. You can combine brand and activation in the same campaign: emotional storytelling with a clear call to action.

Step 4 · Invest in creativity over production value

Creativity is a multiplier of effectiveness (WARC, IPA, Binet & Field). For SMEs, this is the single biggest lever. A distinctive, emotionally resonant campaign on a modest budget will outperform a generic, forgettable campaign on a large one. Most advertising is ignored. Being dull is an expensive mistake.

Step 5 · Measure what you can, and do it consistently

• Track share of search monthly (Google Trends is free).
• Track unprompted awareness if you can afford periodic brand tracking.
• Use ROI for activation spend. Don’t try to ROI your brand spend, track it through awareness, consideration, and share of search.

/ 06The cautionary tales

The recession trap

The Ehrenberg-Bass Institute’s recession research provides a crucial insight: recessions change competitor behaviour far more than consumer behaviour. Consumers are habitual. Even in the 1970s recession, arguably the most extreme since WWII, Nielsen records show no contraction in consumer packaged goods purchases: across 50 representative US supermarket categories, sales grew an average of 15% from 1972 to 1979 despite 81% price inflation, with near-identical gains across the same categories in 22 other countries.

The opportunity: when competitors cut spend and media prices drop, brands can substantially increase SOV at lower cost. The trap: cutting marketing in a recession feels prudent but hands market share to competitors who maintain or increase investment. Most importantly, as the IPA data shows, the cost of regaining lost share post-recession exceeds the savings made during it.

/ 07The bottom line

• ESOV (SOV/SOM). Use it for setting the total budget envelope. Don’t use it for channel-level allocation.
• 60/40 split. Use it for brand vs activation allocation. Don’t use it for determining total spend.
• MMM / econometrics. Use it for channel optimisation and proving ROI to the CFO. Don’t use it for setting the total budget, it optimises, it doesn’t set.
• Zero-based budgeting. Use it for eliminating waste within an allocated budget. Don’t use it for determining the total budget.
• % of revenue. Use it for sanity-checking your ESOV-derived budget. Don’t use it for setting the budget itself.
• Competitive parity. Use it for benchmarking context. Don’t use it for budget determination.
• Objective & task. Use it for strategic framing, combined with ESOV data. Don’t use it as a standalone method without a quantitative backbone.

The debate about marketing budget setting has persisted because people keep looking for one method. There isn’t one. The answer is a composite: ESOV sets the envelope, the 60/40 heuristic allocates between brand and activation, MMM optimises channel allocation, and ZBB ensures efficiency within the budget. Each method has a role. None works alone.

For SMEs without econometric resources, the Triple-Cooked method (5–10% of revenue, share of search as an ESOV proxy, 60/40 split adjusted for stage) provides a rigorous, evidence-based alternative that any marketer can implement on Monday morning.

/ 08Sources and further reading

• Binet, L. & Field, P. (2013). The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies. IPA.
• Binet, L. & Field, P. (2017). Media in Focus: Marketing Effectiveness in the Digital Era. IPA.
• Whittaker, I. (2024). Marketing is an Investment. IPA EffWorks / Liberty Sky Advisors.
• WARC (2024). Anatomy of Effectiveness. WARC White Paper.
• Ehrenberg-Bass Institute. A Marketing Guide: What to do in a Recession. University of South Australia.
• Mi3 / Mutinex (2025). Marketers, Money & Measurement: MMM’s Next Wave. Marketers & Money Forum, Sydney.
• Jefferson, M. (2022). “‘Triple cooked chips’: Ritson’s ‘foolproof’ system for marketing budgets.” Marketing Week, 6 October 2022 — reporting Mark Ritson’s Festival of Marketing session.
• Ritson, M. (2020). “It’s time for ‘share of search’ to replace ‘share of voice’.” Marketing Week, 24 September 2020.
• Sutcliffe, C. (2023). “Tesco chief: ‘I don’t see marketing as a cost. I see it as an investment’.” Marketing Week, 4 October 2023.
• Kite, G. / Magic Numbers. Revenue-to-advertising ratio analysis (referenced in Jefferson, 2022).
• Gartner (2025). 2025 CMO Spend Survey.
• Hankins, J. Share of Search methodology.
• Kraft Heinz case study: multiple sources including Branding Strategy Insider, CNBC, Reach3 Insights (2019–2026).