Most marketing budgets fall apart in week three. Not because the number was wrong, but because nobody wrote down what the money was supposed to buy. A spreadsheet with "Ads — 40,000" on one line is not a marketing budget. It is a wish.
A real marketing budget is a set of decisions you have already made, written down before the pressure arrives. It tells you what to cut when revenue dips, what to protect when a competitor outspends you, and what to do with a surprise 20% if a good quarter hands you one. This guide walks through building that kind of budget — how much to spend, how to split it, and how to keep it honest once the quarter is underway.
How much should you actually spend on marketing?
The benchmark you will hear most often is roughly 5–15% of revenue, with B2C at the higher end and B2B at the lower. It is a fine starting point and a terrible finishing point, because it says nothing about what stage you are in.
Three situations, three very different answers:
- You have product-market fit and steady revenue. Percentage-of-revenue works. Pick a number in the 7–12% band, hold it for four quarters, and judge it on pipeline rather than feelings.
- You are pre-fit or launching something new. Percentage-of-revenue is circular — you have little revenue, so you spend little, so you get little revenue. Budget from runway instead: what can you spend for six months without threatening payroll?
- You are in a land-grab. A closing window, a competitor stumbling, a channel that is temporarily cheap. Budget from opportunity, accept a worse short-term margin, and set a hard date to reassess.
The mistake is using the wrong model for your stage. A seed-stage startup copying an enterprise 8%-of-revenue rule will starve itself. An established business budgeting from "opportunity" every quarter is just overspending with better vocabulary.
A budget is not a prediction of what you will spend. It is a decision about what you will not.
Work backwards from pipeline, not forwards from last year
Last year's number plus 10% is how budgets drift out of touch with the business. Try the reverse:
1. Start with the revenue target. 2. Divide by average deal size to get deals needed. 3. Divide by close rate to get qualified opportunities needed. 4. Divide by lead-to-opportunity rate to get leads needed. 5. Multiply by your cost per lead.
That last number is your demand-generation budget — the floor, not the whole thing. If the number that comes out is absurd relative to the business, you have learned something more useful than a budget: either the target is fantasy, or your funnel maths needs fixing before more money goes in. Our guide to the marketing funnel covers the conversion rates to sanity-check each step.
The 70/20/10 split, and when to break it
A budget split that has held up well:
- 70% to what works. Channels with known, measured returns. This is the boring majority and it should be.
- 20% to scaling what is promising. Things that showed early signal but have not been pushed hard.
- 10% to genuine experiments. New channels, new formats, things that will probably fail.
The 10% is the part everyone cuts first, and cutting it is how teams end up with one channel and no idea what to do when it gets expensive. Protect it in writing. If a quarter is tight, shrink it to 5% — do not zero it.
Break the split when:
- You are single-channel dependent. If 80% of pipeline comes from one source, your experiment budget needs to be 20–25% until that changes. Concentration is the real risk, not the experiment.
- Something is clearly working and under-funded. If a channel is returning well above blended CAC and you are capping it at 20% out of tidiness, that is bookkeeping beating judgement.
- You have no measurement yet. Before you can say "what works," you need attribution you trust. Spend on the measurement first; otherwise the 70% is guesswork wearing a suit.
Finally know what’s actually working.
Track reach, engagement, and ROI across every channel in one dashboard. 15-day free trial, no card required.
See your analytics →What belongs in a marketing budget
Media spend is the visible part and often less than half the total. A complete budget has five buckets.
1. Media and paid distribution
Ad spend across search, social, retargeting, sponsorships, affiliate payouts. The easiest to track, the easiest to over-weight.
2. Tools and software
Your CRM, email platform, scheduler, analytics, design tools, SEO software. This one creeps. Audit it every six months — most teams are paying for two tools that do the same job and one nobody has logged into since March.
Consolidation is usually the fastest saving available. Moving from six point tools to one platform can cut the line item noticeably and removes the integration tax nobody budgets for: the hours spent moving data between systems.
3. People and contractors
Salaries if marketing carries them, plus freelancers, agencies, editors, designers. If you use contractors, budget by output rather than hours — "four blog posts and eight videos a month" is a number you can hold someone to.
4. Content and creative production
Photography, video, design, tooling for production. Underfunding this is a false economy: creative quality is usually the biggest lever on ad performance, and a cheap ad that does not work costs more than an expensive one that does. Ad creative testing is where that money proves itself.
5. The buffer
5–10% unallocated. Something will come up — a conference worth attending, a partnership that appears in May, a competitor's misstep worth answering. Teams without a buffer either miss those or raid the experiment budget.
Building the thing: a worked example
A B2B SaaS doing ₹4 crore ARR, targeting ₹6 crore, with a 9% budget — ₹54 lakh for the year, ₹4.5 lakh a month.
- Media — 38%, about ₹1.7L a month
- Tools — 12%, about ₹54,000
- Contractors — 26%, about ₹1.17L
- Content production — 16%, about ₹72,000
- Buffer — 8%, about ₹36,000
Now the part most budgets skip — what each bucket is expected to return:
- Media should produce 60 qualified leads a month at ≤ ₹2,800 each.
- Content should produce 12 pieces, with organic traffic up 15% by Q3.
- Tools should shrink as a share of budget as revenue grows, not stay flat.
Write the expectation next to the number. A budget line without an expected outcome cannot be evaluated, only spent.
Seasonality: the reason flat monthly budgets fail
Dividing the annual number by twelve is tidy and usually wrong. Almost every business has months that convert better and months that do not, and spending the same in both is a quiet way to waste a chunk of the year.
Look at last year's revenue by month before you allocate. Most teams find something like this:
- Two or three months carry a disproportionate share of revenue. Retail has the festive run-up, B2B has the weeks either side of budget cycles, education has enrolment windows.
- One or two months are genuinely slow. Spending hard into a slow month buys impressions from people who were never going to buy this quarter.
Weight the media budget towards the months that convert — 1.4x in peak months, 0.6x in slow ones, balancing out across the year. Keep the always-on channels level; it is the campaign spend that should flex.
The one thing not to cut in a slow month is content and SEO. Those compound on a lag, so the work you do in a quiet month is what makes the next peak cheaper. Slow months are the right time to build evergreen content precisely because the team has the capacity.
Plan, approve, publish — together.
A shared calendar, approvals, and roles so your whole team ships on time. 15-day free trial, no card required.
Try Skyfliq free →Where marketing budgets leak
Four patterns turn up again and again in budget reviews.
Zombie subscriptions. Annual plans that renew silently for tools replaced eighteen months ago. Export your card statement, list every recurring charge, and make someone name the owner for each. Anything without an owner gets cancelled.
Agency scope drift. A retainer set for one scope quietly covering three. Re-baseline annually against deliverables, not history.
Always-on campaigns nobody reviews. The retargeting campaign that has been running since 2024 because it was set up well and never questioned. Put a review date on every always-on campaign.
Production spend with no distribution budget. Teams routinely spend ₹80,000 on a video and nothing on getting it seen. A useful rule: for every rupee of production, plan at least a rupee of distribution. A good asset nobody sees is a worse investment than an average asset that reaches people.
Reviewing without re-planning every month
Monthly re-planning is how budgets become a second job. Use a cadence with different questions at each level:
- Monthly — pacing. Are we on track to spend what we said? Over-pacing means the quarter ends dark; under-pacing usually means something is stuck in approvals.
- Quarterly — allocation. Has anything moved between the 70/20/10 tiers? A promising experiment should graduate to the 20%, and something in the 70% should be demoted when returns fade.
- Annually — the model itself. Is the percentage still right? Is the funnel maths still true?
Track spend against outcome in one view rather than two. If your spend lives in a spreadsheet and your results live in analytics, nobody joins them up under deadline. Pulling campaign spend and pipeline into a single dashboard is a small piece of setup that changes how the conversation goes — see our notes on building a social media report for a structure that translates well to full-budget reporting.
Defending the budget
At some point someone asks marketing to justify its number. Three things make that conversation short.
Talk in pipeline, not impressions. "We generated 640 qualified leads at ₹2,600, which became ₹1.9 crore of pipeline" ends the discussion. "We got 2.1 million impressions" invites it.
Show what a cut actually costs. Not "we would be less effective" but "a 20% cut removes roughly 120 leads a quarter, which at our close rate is about ₹35 lakh of pipeline." Make the trade visible.
Bring the experiment results, including the failures. Nothing builds trust in a budget like showing you killed something that was not working before anyone asked. It proves the 10% is a discipline, not a slush fund.
Turn followers into customers.
Capture leads, nurture them by email, and close — all where you already post. Free for 15 days, no card required.
Start free →A simple starting template
If you are building your first proper budget, start here and adjust:
- Set the annual number from revenue target and funnel maths, sanity-checked against a percentage-of-revenue band.
- Split 70/20/10 across proven, promising and experimental.
- Allocate across the five buckets — media, tools, people, production, buffer.
- Write an expected outcome beside every line.
- Set review dates — monthly pacing, quarterly allocation, annual model.
- Name an owner for each line, including the tools.
That is a page, not a spreadsheet epic. The page is the point: a budget nobody can read in five minutes is a budget nobody will follow in month three.
The honest part
The best budget you can write in year one is a decent guess with good instrumentation attached. You will be wrong about at least one channel. The teams that end up with a reliable budget are not the ones who forecast brilliantly — they are the ones who measured honestly, moved money quickly when the data disagreed with the plan, and kept a little aside for the thing they had not thought of yet.
Build for that, and the number takes care of itself.