Marketing Budget Calculator
A marketing budget range built from published survey data, with every figure sourced — and an honest account of the one thing this tool will not guess.
Your business
$356,948 – $639,650 a year, or 7.1% – 12.8% of revenue. The lower figure is median-anchored and the upper is mean-anchored.[1]
Read the lower end. Only 13% of the survey’s respondents were under $10M in revenue and 32% under $100M, so the small-company bands rest on the thinnest evidence in the dataset — and they read high.[1]
$5,000/mo on SEO: Above the bulk of the market
This sits above the roughly 69% of providers who charge $2,000 or less, in the range where dedicated agency teams operate.[3]
What this calculator will not tell you
It will not split your budget across SEO, PPC, email and social. We went looking for the evidence and it is not there. The only company-level SEO allocation figure published anywhere sits behind a paywall, was reported by a single trade outlet, and was restated there against the wrong denominator — a difference that moves the answer by more than three times. The survey series it comes from also fails an internal arithmetic check: two releases three weeks apart imply more spend on paid online media alone than the same publisher reports for all paid media combined.
Every other channel split in circulation — “SEO should be 25–30% of digital”, “60/40 owned versus paid” — is agency opinion with no survey behind it. A calculator that returns four channel figures is inventing at least three of them. We would rather hand you a narrower tool that is true, and be plain about where the evidence stops.
Why a flat percentage breaks down as you grow
The instinct behind most budget advice is that marketing spend should be a fixed share of revenue. It is a reasonable starting point and a bad finishing point, because marketing returns saturate: each additional dollar buys less than the one before it.
This is not a contrarian opinion. John Little formalised the saturation curve in 1970 in the ADBUDG model[8], and Dominique Hanssens’ review of twenty-five years of research concluded that “the predominant response function is concave” and that advertising is the weakest of the marketing-mix instruments[9]. A meta-analysis of 751 advertising elasticities found a mean of .12 and a median of .05 — implying optimal advertising of 4.6% and 1.9% of sales respectively[7].
It is also settled practice rather than theory. Google’s Meridian[10] and Meta’s Robyn[11], both open source, implement saturation as a required part of the model — the same Hill-function family Little described fifty-six years ago. When the two largest advertising platforms in the world ship diminishing returns as mandatory code, it is not our opinion.
The clearest single demonstration comes from eBay, where large randomised field experiments found that brand-keyword search advertising produced no measurable short-term benefit and non-brand keywords returned negative average returns[12]. That is a paid-search result rather than an SEO one, but it makes the point: a large, well-known brand can spend more on search and earn nothing extra for it.
Where SEO spend actually stops scaling
Here is the part nobody in this industry says out loud: no one has published an SEO-specific response curve. Not academia, not the major vendors. Every “SEO diminishing returns” article is an agency blog post with no data behind it. We looked, and we are not going to invent the curve that does not exist.
What is documented is the market. Four independent surveys of SEO providers spanning 2012 to 2024 agree on the shape: retainers thin out sharply above $10,000 a month and effectively terminate between $25,000 and $50,000. Ahrefs found roughly 2% of providers above $10,000 and 0.3% in the $25,000–$50,000 band[3]. SparkToro found no agency averaging above $50,000[4]. Most tellingly, Moz’s 2012 survey offered buckets all the way to $250,001 a month and the US sample declined to use anything above $25,000[5]. Hand a market an unbounded ceiling and watch it refuse — that is evidence, not an absence of it.
The mechanical reason is that SEO is labour delivered against a site with a finite number of templates, a finite crawl surface and a finite set of addressable queries. At a blended $111 an hour, $50,000 a month buys roughly 450 hours — about three full-time specialists. Past some point you are paying capable people to work on a site that has run out of high-leverage things for them to do, and the sensible move is to bring the work in-house rather than escalate a retainer. That crossover is what the calculator above shows you.
Read the median, not the mean
One last caution about every other budget calculator you will find. US marketing leaders report a mean of 8.96% of revenue — and a median of 5.0%[1]. The distribution is heavily right-skewed by a minority of very heavy spenders. A tool that quotes you “about 9%” is quoting a number that more than half of real companies sit below, which is why the result above is a range with both ends labelled.
For what a specific SEO investment returns rather than what it costs, the SEO ROI calculator models payback against a fixed monthly fee.
Sources
- The CMO Survey, 35th Edition (Duke Fuqua / Deloitte / AMA), February 2026 — Fielded 7–29 Jan 2026. Frame 2,111 US for-profit marketing leaders; 308 responded (14.6% response rate); 97% VP-level or above. Marketing as % of revenue n=154.
- The CMO Survey, 34th Edition, February 2025 — Used alongside the 2026 wave to pool the revenue-band figures, because individual band cells are unstable between waves.
- Ahrefs, “SEO Pricing: How Much Does SEO Cost? 439 People Polled” — N=439 SEO service providers, global, published Dec 2023 / updated Aug 2024. Self-selected sample; recruitment method not disclosed by the publisher.
- SparkToro (Rand Fishkin), “SEO Consultant & Agency Pricing” — US & Canada, fielded late 2015. Agencies n=94, solo consultants n=41. Recruited via the author’s Twitter following; bias disclosed by the author.
- Moz, “SEO Pricing & Costs of Services”, January 2012 — 600+ responses, 490 analysed (US n=287). Notable because the questionnaire offered retainer buckets up to $250,001/month and the US sample is effectively empty above $25,000.
- US Bureau of Labor Statistics, OEWS May 2025, SOC 13-1161 — “Market Research Analysts and Marketing Specialists”: median $78,760, mean $89,490. A true probability sample. Note: SEO is not a BLS occupation, so this is the nearest available proxy, not a measurement of SEO salaries.
- Sethuraman, Tellis & Briesch (2011), Journal of Marketing Research 48(3) — Meta-analysis of 751 short-term advertising elasticities from 56 studies, 1960–2008. Mean elasticity .12, median .05. Table 4 derives optimal advertising at 4.6% of sales on the mean and 1.9% on the median.
- Little (1970), “Models and Managers: The Concept of a Decision Calculus”, Management Science 16(8) — The ADBUDG response model. Establishes the concave/S-shaped saturation curve still used today.
- Allenby & Hanssens (2005), “Advertising Response”, MSI Special Report 05-200 — “The predominant response function is concave… advertising is the weakest of the marketing-mix instruments.” Also documents that smaller competitors carry higher advertising-to-sales ratios than market leaders.
- Google Meridian documentation — media saturation and lagging — Google’s open-source marketing mix model. Implements Hill saturation as a required transform: “as spending on a given media channel… increases, you eventually see diminishing marginal returns.”
- Meta Robyn documentation — features — Meta’s open-source MMM. “The theory of diminishing returns… holds that each additional unit of advertising investment increases the response at a declining rate.” Ships Hill saturation and a budget allocator that solves it.
- Blake, Nosko & Tadelis (2015), Econometrica 83(1) — Large-scale randomised field experiments at eBay. Brand-keyword search ads showed no measurable short-term benefit; average returns on non-brand keywords were negative. Concerns paid search, not SEO.
- US Small Business Administration, Ascent — “Create Your Marketing Plan” — Source of the widely-quoted “7–8% of revenue” rule for businesses under $5M. Included here for context only: the claim carries a footnote marker that appears exactly once in the document, with no corresponding footnote text or citation anywhere in the file.