Marketing budget allocation is not one question but three, because the right split depends entirely on where you are. A business proving whether anyone wants the product has completely different priorities from one scaling something that already sells.
Stage 1: does anyone want this?
Revenue is small, the offer is unproven, you're still learning who the customer is.
- 70% paid media — the fastest way to find out if the offer works.
- 25% creative production — you need angles to test, and you have no idea yet which will land.
- 5% tools — the minimum that makes results measurable.
- 0% SEO. It won't produce anything inside your runway. Start it later.
Don't split this stage across five channels. One channel, enough budget to learn, a real answer in six weeks.
Stage 2: it works, make it work reliably
You have a repeatable customer and a cost per acquisition you can quote.
- 55% paid media
- 20% creative — now the main lever, since targeting is largely automated
- 15% conversion and automation — landing pages, follow-up, the things that make each click worth more
- 10% SEO — start now so next year is cheaper
This is the stage where fixing the funnel beats raising the budget, and where most businesses do the opposite.
Stage 3: scale
- 50% paid media, across more than one channel
- 20% creative, at genuine volume
- 15% SEO and content, compounding underneath the paid spend
- 10% retention — email, WhatsApp, repeat purchase. Cheapest revenue you'll ever buy
- 5% experiments — a new channel, a new format, deliberately ring-fenced
Creative is not a design cost. At this stage it's the targeting lever, and it should be funded like one.
The mistakes
- Zero creative budget. ₹3 lakh in media and nothing to make new ads with. The account fatigues and nobody can explain why performance decayed.
- Spreading thin. Five channels at a quarter of the budget each means five accounts that never exit learning.
- Funding SEO before there's runway for it. It's the right investment at the wrong time.
- Nothing for tracking. Then arguing about attribution for a year.
- Ignoring retention. Raising repeat purchase lifts your CAC ceiling, which is easier than lowering CAC.
How to sanity-check your marketing budget allocation
Two questions. First: if our best ad stopped working tomorrow, do we have budget to replace it? If not, creative is underfunded. Second: can we say what a customer costs, confidently? If not, tracking is underfunded.
Everything else is a preference. Those two are structural.
Before committing a budget, run it through the PPC forecast calculator and see what it actually buys.
What you can actually afford to spend each month is set by how long acquisition cost takes to come back rather than by a share of revenue. For a funded company the same constraint is sharper — spending a raise without burning runway is mostly this calculation.
Questions people ask
What is a good marketing budget split?
It changes by stage. Proving demand means almost everything in media and creative and nothing in tooling. Scaling means a real share in creative production and measurement, because both become the constraint.
How much should go to SEO versus paid?
Paid first if you need customers this quarter, because SEO does not produce them in that window. SEO earns its share once paid is working and you want the cost per customer to fall over time rather than rise.
Should the agency fee come out of the media budget?
Count it separately, then look at the ratio. Hiding the fee inside the media number makes the account look more efficient than it is and hides the point at which management costs too much relative to spend.
How often should the split change?
Quarterly, or whenever the stage changes. Monthly reallocation is usually a reaction to noise rather than a decision.
Want help setting the split for your stage? Tell us where you are.