Marketing Budget Allocation: A Practical Split for 2026

Start with the number of customers you need, multiply by your customer acquisition cost (CAC), add fixed costs, then split the result 70-20-10 across proven, emerging and experimental channels. Most budgets fail earlier than that: they begin with a percentage of revenue borrowed from a company ten times your size.

This guide shows the maths, the split and the tests that tell you when to move money. It is written for Australian e-commerce and DTC teams, where one wrong channel bet is a real hit to cash.

Key Takeaways

• Set the total from goals and CAC, not from a revenue percentage. Australian government guidance gives 3% to 5% of turnover as a rule of thumb, which is a sanity check, not a plan.
• Split spend 70% proven, 20% emerging, 10% experimental, and make each channel earn promotion with six months of data.
• Brand effects are the larger part of long-run return. Kantar’s models put them at 60% of total marketing ROI, so a budget that only funds the bottom of the funnel is under-spending.
• Judge channels on marginal return and incrementality, not platform-reported ROAS. Reallocate when a test says the next dollar is worth less elsewhere.
• Re-cut the split quarterly. Never re-cut it weekly.

Table of Contents

• Set the total first: three ways, ranked
• What goes into the budget
• Split it across channels: the 70-20-10 rule
• Allocate across the customer journey, not just the platform
• Use a media mix model when the spend justifies it
• Know when to move money: incrementality and marginal return
• A quarterly rhythm that stops panic reallocation
• Perspective
• Work with a team that plans the whole budget
• FAQ
• Sources

Set the total first: three ways, ranked

There are three common ways to pick a total. Only one is worth trusting alone.

1. Percentage of revenue. The Small Business Development Corporation in Western Australia gives a general rule of thumb of 3% to 5% of annual turnover, and says new businesses may need more at the start to build a profile. Use it as a floor and ceiling check. Large-company surveys run higher: Gartner’s 2025 CMO survey found budgets flat at 7.7% of company revenue, but its respondents were mostly firms with revenue above US$1 billion. Your DTC brand is not that firm. Copying their number is copying their cost base.

2. Last year plus or minus. Easy, and it bakes in last year’s mistakes. If you wasted 30% then, you will waste 30% again, with inflation.

3. Objective and task. This is the one to use. Semrush’s worked example multiplies a target customer count by CAC, then adds fixed marketing costs: ($250 x 100 customers) + $13,000 = $38,000. The mechanism is simple. The budget now answers a business question (what does it cost to buy the growth we want?) instead of a guess.

Two cautions on the maths. First, CAC is not a constant. It rises as you push spend into colder audiences, so a plan built on today’s blended CAC overstates what extra budget will buy. Second, business.gov.au advises tying the budget to your marketing goals and then evaluating whether activities worked. If you cannot say what “worked” means before the money goes out, you are not budgeting. You are donating.

A worked example

Say you sell skincare online, with a 60% gross margin and an average order of $80. You want 2,000 new customers next quarter. Your blended CAC is $40, so acquisition costs $80,000. Add $15,000 for creative production, tools and email/SMS software. The total is $95,000.

Now check it against revenue. If quarterly revenue is $1.2 million, you are spending 7.9%. That is above the 3% to 5% rule of thumb, which is fine for a brand in growth mode, but it tells you to confirm the $40 CAC is real and not a platform-reported number. (Rule of thumb ≠ target.)

What goes into the budget

A budget with one line called “ads” hides where the money leaks. Aprimo’s guide lists five components: ad spend, outsourcing (agencies and freelancers), research, technology and events. For an e-commerce brand, we would reshape them like this:

Creative and conversion deserve their own lines because they multiply the media line. Meta treats near-identical ads as one ad, so fifteen thumbnails on one script get retrieved like a single ad. You paid for variety and got one creative. Budget for genuinely different concepts, not different crops.

And a note on the retention line: Gartner’s survey found paid media took 30.6% of marketing budgets (2025, large firms). That is the share a typical big marketer puts on media. It says nothing about what is right for you.

Split it across channels: the 70-20-10 rule

The simplest working split is 70-20-10. Growth Method defines it (updated June 2026) as:

• 70% NOW: channels with six or more months of positive performance data.
• 20% NEXT: channels with early positive signals, not yet proven at scale.
• 10% NEW: untested ideas.

Ideas graduate from NEW to NEXT to NOW. That graduation path is the useful part. It stops two common habits: starving experiments so nothing new ever works, and funding every shiny platform at 25% because someone on a podcast said so.

Why it works: it matches how channels behave. A proven channel has a known CAC curve, so you can scale it with predictable returns until it saturates. An unproven one has a wide range of outcomes, so you cap the downside at 10% and learn cheaply.

Caveat: 70-20-10 is a heuristic from a content site, not a law of nature. If your whole business is one channel at 95%, the rule tells you to diversify. If you launched six weeks ago, you have no proven channels and the 70% bucket is empty. Early on, run 50% on the best single bet and treat the rest as tests.

What this looks like for a DTC brand

Here is an illustrative split for a brand with two proven paid channels. These are our planning starting points, not benchmarks, and your data should overwrite them.

If you want depth on the paid-social side, see our Meta and TikTok ads work and how we treat Google Ads for ecommerce.

Allocate across the customer journey, not just the platform

Platforms are how you buy. The journey is why you buy. Split by both.

Top of funnel (reach and brand). This is where most Australian DTC budgets are thin. Kantar’s February 2024 playbook reports that, on average across its models, the long-term effect from brand equity is larger than the short-term effect and represents 60% of total marketing ROI. It adds that this varies by category and market. The same playbook warns that when brand effects are not modelled, you underestimate where diminishing returns begin and, on average, under-invest in media by 50%. Sit with that. Half.

Middle (consideration). Creator content, YouTube, product education. This is the layer that makes retargeting pools bigger. Retargeting only harvests people someone else already persuaded.

Bottom (conversion). Search, retargeting, email, SMS and the landing page. Cheap per conversion, small in total volume. Pouring more money here does not create demand. It just gets you a bigger share of the demand you already made.

After the sale (retention). Repeat purchase is the cheapest revenue you will buy. Give email and SMS a real line, not scraps.

A rough starting range for a growth-stage DTC brand, which we would test against your own numbers rather than treat as a norm: roughly half on prospecting and creative, a quarter on conversion-stage media, and the remainder on retention and measurement. If your funnel is mostly bottom-of-funnel today, move 5 to 10 points up the funnel for one quarter and read the result, rather than flipping the mix overnight.

Longer buying cycles change the timing. If your product takes weeks to decide on, payback slips and so does the point where you judge a channel. Learn the stages and metrics in sales cycle management, then set a review window that fits your cycle, not the platform’s default 7-day attribution.

Use a media mix model when the spend justifies it

A media mix model (MMM) answers the question your platform dashboards cannot: across all channels together, what is each one actually contributing? Marketing Science in Python describes MMM as using aggregated time-series data, typically weekly sales and media spend by channel, so it works without user-level tracking. It can measure, simulate and optimise, including allocating a fixed budget across channels to maximise an expected outcome.

That is attractive in a world of consent limits and signal loss. But MMM needs history, spend variation and enough channels to separate. A brand spending $20,000 a month on two channels does not have a modelling problem. It has a testing problem. We wrote up when media mix modelling pays off and when it doesn’t if you want the thresholds.

The short rule: below a few channels and a year or so of varied weekly data, run lift tests. Above it, add MMM and use tests to calibrate it.

Know when to move money: incrementality and marginal return

Platform ROAS counts conversions the platform touched. Some of those people would have bought anyway. Reallocating on that number is how brands cut the channel that creates demand and feed the one that claims credit for it.

Incrementality fixes the question. Skai defines it (January 2025) as the additional impact beyond what would have happened without the activity, and it can be used to calculate incremental ROAS (iROAS) and incremental cost per acquisition (iCPA). Note that Skai sells measurement software, so treat its claim that these metrics beat attribution as a vendor’s view. The definitions are the useful part.

A named example: FabFitFun used test-and-control incrementality testing and cut its retargeting budget by 81% and customer acquisition cost by 16%, then moved the money into channels beyond digital, including TV and direct mail. The figures come from the measurement vendor that ran the programme, and the page gives no timeframe, so read them as directional, not as a forecast for your brand. The mechanism is what transfers: retargeting was claiming sales that were happening anyway, so the marginal dollar there was worth little.

How to run it without a data science team:

1. Pick one channel. Retargeting or branded search is the usual suspect.
2. Hold out a group. Use the platform’s lift tool, or switch a channel off in matched regions.
3. Run it long enough. Cover at least one full purchase cycle. Two weeks is rarely enough for considered purchases.
4. Compare iCPA to blended CAC. If the incremental cost is far above what the dashboard said, the channel is over-funded.
5. Move a slice, not the lot. Shift 10% to 20% of that channel’s budget and re-read.

We covered the setup in incrementality testing for Meta ads. If your team is also bringing AI tooling into reporting, an AI readiness assessment shows what staff already use before you wire it into measurement.

Two blended metrics that keep you honest

Platform ROAS is a per-channel view. Marketing efficiency ratio (total revenue divided by total marketing spend) is the company view. When ROAS rises and MER falls, the channels are trading credit, not creating sales. Our MER vs ROAS workflow sets budgets with both and protects prospecting from being cut for looking expensive. For Google specifically, check the attribution columns before changing bids.

A quarterly rhythm that stops panic reallocation

Moving budget every week is its own tax. Algorithms need a learning period, and you never see a clean read. Use this cadence instead:

• Weekly: watch spend pacing and creative fatigue. Do not move budget between buckets.
• Monthly: compare channel CAC and MER to plan. Flag, but do not act on, single-month swings.
• Quarterly: re-cut the 70-20-10 split. Promote a NEXT channel if it has held its CAC target for the period. Retire a NOW channel only if a lift test backs it.
• Annually: rebuild the total from goals and CAC, then run the revenue-percentage sanity check.

(Fast review ≠ good review.)

Perspective

Most budget advice is a pie chart with no mechanism. Seventy here, twenty there, and nobody says what the money does once it lands. We think that is the real problem, and it comes from how agencies are built.

When creative sits in one team, media in another and conversion in a third, each team optimises its own number. Media buys cheap clicks. Creative wins awards. The landing page gets fixed next quarter. The budget is allocated by channel because the org chart is organised by channel.

We run strategy, creative production, paid media and conversion as one senior team. The strategist who reads the numbers also briefs the creators and signs off the edit. So when the data says a channel is saturated, the fix is not only a lower bid. It might be a new concept, a different offer or a faster page. Allocation becomes a decision about the whole funnel, not a tug of war between line items.

That is also why we aim to deliver at least 3 times your investment in us, and why we judge the work on the bottom line, not just revenue. A budget that looks tidy but ignores margin is a nicer-looking mistake.

Harry Sekhon

Work with a team that plans the whole budget

If you want the split built, tested and run by one team, that is what we do. We handle brand and creative strategy, ad creative production, paid media across Meta, TikTok, Google and YouTube, conversion optimisation and email/SMS, with no hand-offs between them.

See the full range on our services page. We don’t publish prices, because the right scope depends on your numbers. Happy to talk through your CAC and your current split, no pitch required.

FAQ

What is a typical marketing budget?

For small Australian businesses, the Western Australian small business agency gives a 3% to 5% of annual turnover rule of thumb, with more for new businesses. Large firms surveyed by Gartner reported 7.7% of revenue in 2025. Treat both as a check on a bottom-up number built from CAC and customer targets.

What is the 70-20-10 rule in marketing?

It puts 70% of budget on channels with six or more months of positive data, 20% on channels showing early signs, and 10% on new ideas, as Growth Method describes it. Ideas move up a tier as they prove themselves. It is a planning heuristic, so adjust it if you have no proven channel yet.

What are the five components of a marketing budget?

Aprimo lists ad spend, outsourcing, research, technology and events. E-commerce brands usually also split out creative, conversion work and retention so each has its own line and its own target.

How do I know when to move budget between channels?

Move it when an incrementality test, or a sustained gap between a channel’s CAC and your target, says the next dollar earns less than it would elsewhere. Shift 10% to 20% of that channel’s budget, then re-read after one purchase cycle. Do not act on a single week of platform ROAS.

Sources

• 8 steps to marketing your business, SBDC Western Australia
• Develop your marketing plan, business.gov.au
• Gartner CMO Spend Survey coverage, Demand Gen Report
• How to Create a Marketing Budget in 9 Steps, Semrush
• Marketing Budget Approval, Aprimo
• The 70-20-10 Rule, Growth Method
• Marketing Effectiveness through the wide lens, Kantar
• MMM fundamentals, Marketing Science in Python
• Incrementality Measurement, Skai
• Why FabFitFun Uses Incrementality, Measured
• Sales cycle management, Foyai
• AI readiness assessment for marketing agencies, Calibre Forum

Recommended

• PMax vs Search: How to Split Your Google Ads Budget
• Facebook Ads Scaling: How to Add Spend Without Losing Profit
• MER vs ROAS: Six Step Workflow to Set Budgets and Protect Prospecting
• Paid Media Strategy: How to Build One That Pays Back