Research · September 2, 2026

How Pre-Revenue Companies Build a Marketing Budget

What the published methods actually are, who publishes them, and what the numbers are. Sources and dates on everything.

There is no percentage of revenue, so practitioners swap the denominator. Four are in common use: percent of capital raised, percent of burn, working backwards from a next-round ARR target, and zero-based from the funnel. Only one of the four has a fully published, step-by-step process behind it, and that is Emily Kramer's at MKT1. Everything else is a benchmark range rather than a method.

Read this before the numbers

There is no primary dataset for pre-revenue marketing budgets. Gartner's CMO Spend Survey, The CMO Survey, SaaS Capital and Forrester all measure companies that have revenue, which is exactly the population that does not answer this question. Every number below is either an operator framework, an agency's observed client data, or a synthesis. Treat the ranges as calibration, not as evidence.

1The four denominators

Method	Published figure	Source, date
Percent of capital raised	10–20% of funds raised at seed, producing $50K–$250K/yr	AAJ Consult, How Much Should a Startup Spend on Marketing?, 2026-06-08
	10–25% of raised capital, deployed across the 18–24 months after a Series A	Stackmatix Series A analysis, cited by Zulu Method, 2026-06-30
	10–20% of total seed capital toward go-to-market preparation (brand, site, foundational assets), explicitly not ad spend	C-I Studios, 2026-03-29
Percent of burn	20–30% of monthly burn at pre-seed; 15–30% of quarterly burn pre-revenue generally	Stackmatix, Startup Marketing Spend Benchmarks by Stage, 2026-08-24
	Pre-revenue/seed: sales and marketing combined at 80–120% of burn rate, marketing taking 30–40% of that, so 25–45% of burn	Pipeline Road, 2026-01-29
	30–60% of a small revenue base, framed as percent of burn, producing $0–$100K/yr pre-PMF	AAJ Consult, 2026-06-08
Backwards from a next-round ARR target	Three interchangeable formulas. See section 2.	MKT1 (Emily Kramer), 2024-12-11, updated 2025-06-12
Zero-based from the funnel	New ARR → customers needed → opportunities needed → pipeline needed → channel budget at a cost per opportunity	Pipeline Road, 2026-01-29

2The one real method

MKT1 is the only source with a complete published process rather than a range. Emily Kramer ran marketing at Asana, Carta and Astro, advises 100+ B2B startups, and publishes the template. Her seven steps:

Calculate a ballpark budget using efficiency metric targets
Figure out the hiring plan and marketing headcount budget
Break the remaining budget into categories
Reconcile the target with a bottom-up forecast, and the forecast with the target
Set monthly budget targets
Build the budget and efficiency metrics summary, and check it makes sense
Track through the year with an eye on the efficiency metrics

Step 1 has three interchangeable formulas. Run all three deliberately, because each returns a different number and the spread is the point.

Option A, CAC ratio

marketing budget = (New ARR / CAC ratio) × marketing % of total CAC

Benchmarks: CAC ratio 0.7 to 1.2. Marketing share of total CAC 35 to 45%.

Option B, payback period

marketing budget = (CAC per customer × target # of new customers) × marketing % of total CAC

Benchmarks: payback 12 to 18 months. Target customers = new ARR target / ACV.

Option C, LTV:CAC

marketing budget = (LTV / LTV:CAC ratio) × marketing % of total CAC
where LTV = ACV / annual churn

Benchmark: 3:1, which Kramer notes is the version founders and investors already understand.

The purpose of using these budget calculations is to establish a starting place for your marketing budget. Typically, each method gives you a different budget, which is helpful for setting a target range. Emily Kramer, MKT1, 2024-12-11

Some marketers determine budget based on targeting a percentage of new ARR and/or total ARR. I find the benchmarks for marketing budget as a % of revenue inconsistent compared to the 3 efficiency metrics above. Emily Kramer, MKT1, 2024-12-11

Note what step 2 does: headcount comes out before categories get filled. That ordering is why comparing your number to someone else's is usually meaningless, which is the subject of section 4.

3The published dollar ranges

Stage	Monthly	Annual	Basis given	Source, date
Pre-seed, $0–1M raised	$1,000–5,000	$12K–60K	20–30% of monthly burn	Stackmatix, 2026-08-24
Pre-seed / pre-PMF	—	$0–100K	% of burn	AAJ, 2026-06-08
Pre-revenue B2B tech specifically	$3,000–8,000	$36K–96K	% of monthly runway	Salient PR, 2026-07-16
Pre-revenue to $500K ARR	$5,000–15,000	$60K–180K	Excludes founder time	Pipeline Road, 2026-01-29
Seed, $1–5M raised	$5,000–20,000	$60K–240K	Channel validation	Stackmatix, 2026-08-24
Seed	—	$50K–250K	10–20% of funds raised	AAJ, 2026-06-08
Seed / pre-PMF	$3,000–15,000	$36K–180K	Prove one or two channels	Zulu Method, 2026-06-30
Series A, $5–20M raised	$20,000–80,000	$240K–960K	Scale proven channels	Stackmatix, 2026-08-24
Series A	—	$300K–1.5M	20–30% of ARR	AAJ, 2026-06-08

The ranges cluster tightly: a pre-revenue company that has raised a seed round is described almost everywhere as spending somewhere between $3K and $20K a month, and the gap between the low and high end is mostly whether headcount is counted.

4The two things that make benchmarks useless

Mistake 1, mixing denominators

"The denominator shifts as you scale. Early on you're spending against funds raised or cash burn; later you're spending against revenue or ARR. Mixing those up is the single most common benchmarking error."

AAJ Consult, 2026-06-08

Mistake 2, comparing loaded numbers to media-only numbers

Two companies can both say they spend 10% and be running completely different programs, because one counts fully loaded team cost and the other counts only paid media. A fully loaded number includes salaries and benefits, agencies and contractors, software, paid media, content and creative production, and events.

markcmo.com, 2026-07-10; Artisan Growth Strategies, 2026-07-15

Before quoting any benchmark, pin down two things about it: what the denominator is, and whether headcount is inside the numerator. Most published figures do not say, which is why they disagree so widely.

5What the money goes to at seed

The only source found with a full category split specifically for the pre-revenue to $500K stage:

Category	Share	Monthly	Covers
Outbound tools and prospecting	30%	$1,500–4,500	Apollo, Sales Navigator, sending infrastructure
Content and SEO	25%	$1,250–3,750	Freelance writer, Ahrefs, basic design
Paid ads	20%	$1,000–3,000	Small targeted campaigns for demand capture
Website and landing pages	10%	$500–1,500	Webflow or site improvements
Tools and analytics	10%	$500–1,500	CRM, GA4, email
Events and networking	5%	$250–750	Local events, industry meetups
Total	100%	$5,000–15,000	Excludes founder time

Pipeline Road, SaaS Marketing Budget: Allocation Frameworks, Benchmarks, 2026-01-29. Their stated benchmark inputs: OpenView SaaS Benchmarks 2025, KeyBanc SaaS Survey 2025, Bessemer Cloud Index, Gartner MarTech Survey 2025, plus their own client programs.

6What the budget is for, in the sources' own words

A seed marketing budget is not a branding allowance. It is a proof-of-model budget. The real job of a seed marketing budget is not to maximize visibility. It is to prove a repeatable acquisition model that investors can believe in when the company starts pushing toward Series A. Directive Consulting, The Seed Stage Budget Allocation Guide, 2026-06-09

If you're spending six figures on marketing before product-market fit, you're usually buying noise. This is channel-validation money. Every dollar should be a data point, not a bet. AAJ Consult, 2026-06-08

The 15 to 20% of total funding rule is a starting point, not a strategy. A founder who raised $2M and spent $400K on brand campaigns before proving CAC may be burning runway on the wrong thing. Allocation matters more than total spend. Scalix, Seed-Stage Marketing Budget Guide, 2026-04-09

Pre-revenue budgets should be sized around what experiments need to run to validate channel-market fit. Stackmatix, 2026-08-24

Named mistakes

Spending on visibility before proving demand capture. Founders feel pressure to look bigger right after a raise, which pushes money into channels that create attention but not acquisition clarity. (Directive, 2026-06-09)
Weak measurement. Without attribution and sales feedback, teams optimize toward easier signals: form fills, low-intent leads, traffic growth. Performance looks better than it is. (Directive, 2026-06-09)
Channel sprawl. Named explicitly as a common seed failure. (Directive, 2026-06-09)
Clinging to seed-era budgets at Series A, which starves the engine investors just funded. (AAJ, 2026-06-08)
Broad awareness campaigns and expensive experiments with weak commercial feedback loops. Named as what does not belong in a seed plan yet. (Directive, 2026-06-09)

7What this adds up to

The consensus process, stripped of the agency content wrapped around it:

Pick a denominator that exists. Capital raised or burn, not revenue.
Run MKT1's three formulas against the next-round ARR target to get a range rather than a point.
Take headcount out before allocating anything to categories.
Reconcile the top-down range against a bottom-up funnel forecast. Where they disagree, an assumption is wrong.
Allocate to a small number of channels and treat the spend as a purchase of information about which one works.

The disagreement worth knowing about

Sources split on whether pre-PMF companies should run paid at all. Salient PR says paid acquisition rarely makes sense before product-market fit because there is no validated message to scale. Directive says the opposite, that high-intent demand capture should be the priority precisely because it produces the clearest commercial feedback. Both are arguing about the same thing: whether the goal is to learn what converts or to build a narrative first.