How should I split my marketing budget across channels?
Split your budget by what each channel adds, not by the credit it gets. Keep most spend where a test backs it, give a small slice to new tests, and move money in steps. Without tests, last-click shares are a first guess, not an answer.
By Joris van Huët, Founder & CEOUpdated 6 min read
Run the numbers for your store: the free multi-channel budget calculator.
Usually, split it by what each channel adds, not by what it gets credit for. Keep most of the budget where a test backs the spend, give a small slice to new tests, and move money in steps. If you have no tests yet, last-click shares are a starting guess, not an answer.
The usual answers are a fixed percentage rule or a ranking by platform ROAS. Both fit neatly in a spreadsheet. Neither asks what a channel would sell if you switched it off, and that is the number a split depends on.
What one store's data shows
One store's anonymised GA4 export, 1 January 2024 to 21 August 2026. It holds shares of revenue only: no ad spend, no order counts.
| What the export shows | Share of revenue | Source cell |
|---|---|---|
| Direct, in last click, first click and touched views | 57.7% | Channels sheet, Direct row |
| Paid Social, in all three views | 0.0% | Channels sheet, Paid Social row |
| Journeys with 1 touch (0.5 days to buy) | 79.5% | Journeys sheet, 1 touch row |
| Journeys with two or more touches, added up | 20.6% | Journeys sheet, two to three, four to nine and ten or more touches rows |
Split this store's budget by credit and the biggest slice goes to Direct, at 57.7% of revenue in the export under every rule. You can't buy Direct. It doesn't take bookings.
Paid Social shows 0.0% in all three views on the Channels sheet, so a split by credit gives it nothing. That is right only if it caused nothing. The export can't say whether paid social ran untagged, ran and did nothing, or never ran.
In the Journeys sheet, journeys with 1 touch carry 79.5% of revenue, at 0.5 days to buy. Journeys with two or more touches hold 20.6% of revenue in the export, about a fifth. Yet they make up 3,656 of the 3,670 distinct path sequences in the export. Most of the variety sits in a fifth of the money, so a dashboard full of long paths is mostly showing you the small end.
The Break-even sheet adds one line of arithmetic: at a 40% margin, break-even ROAS is 2.5x, because 1 divided by 0.40 is 2.5. It is a floor for judging any channel, not a result.
What the export can't show is spend. Without spend there is no return per channel, so the export can rank credit but not value.
Why does the usual answer mislead?
A fixed percentage rule ignores your margin, your buying cycle and what each channel adds. It is a diet plan written before anyone stepped on the scales. It also treats every channel as equally easy to measure. Email and brand search often sit at the end of journeys and collect credit, while video and social can start journeys that end somewhere else.
Ranking by platform ROAS has two leaks. Each platform counts the sales it touched, so the same sale can sit in several reports. And a channel's average return says little about its next euro. Google researchers note that advertising usually has lag effects and diminishing returns. The first thousand euros in a channel can work far harder than the tenth thousand.
Even careful models wobble. The same paper found that the optimal media mix its model produced had a large variance, because its parameter estimates were uncertain.
Platform planners stay inside their own walls. Google's Performance Planner works by shifting budgets between campaigns in your Google Ads account. It won't tell you that a euro does more on Meta.
Brand search shows the trap best. Field experiments at eBay found returns from paid search were a fraction of non-experimental estimates, and brand-keyword ads showed no measurable short-term benefit. Yet a brand click sits right before the sale, so last click hands it full credit.
What can a credit-based split never tell you?
It can't tell you what happens if you cut a channel, because nobody switched it off. It can't price the next euro, because averages hide the curve. And it can't see sales that never touch a tracked click, like word of mouth that turns up as Direct.
That leaves a simple shape. Anchor most of the budget where a test backs the spend. Give a slice to the channel you are testing now. Keep a little for something new. For illustration, a €20,000 month might put €15,000 on tested channels, €3,000 on the one under test and €2,000 on a new bet.
Google's own lift guidance says to test before you make major budget decisions. Move money in steps, then let the next test check the move.
Then hold each slice to your floor. A channel whose tested return sits below break-even shrinks, however pretty its dashboard. A channel that clears it with room to spare gets the next step up.
What to do this week
- See which channels bring new buyers. In Shopify admin, open the Growth page and select View channel report. Add the New customers and Returning customers columns. Pass: you can name the channels that win first orders. Fail: every channel shows mostly returning customers, so your acquisition budget is buying repeat orders.
- Ask Google where its next euro flattens. In Google Ads, go to Performance Planner within the Tools menu. Select the plus icon and plan next month. Pass: the forecast shows extra spend buying fewer extra conversions, so you know where to stop. Fail: your campaigns are unforecastable, so judge Google spend on a test instead.
- Book one test before the next budget meeting. In Google Ads, open Lift measurement in the Goals menu. Start a study for your biggest channel. Pass: you have test dates and regions on paper. Fail: Conversion Lift isn't offered, so plan a regional go-dark test by hand.
Check the homework. Your GA4 Attribution paths export already holds the evidence. Causality Engine reads that one file and shows what each channel caused next to what last-click gave it, in 1 to 2 minutes, for €99 once (excluding VAT), refundable within 30 days. Check the homework
Sources, 1 October 2026: Bayesian Methods for Media Mix Modeling with Carryover and Shape Effects (Google Research). Create and edit a plan with Performance Planner (Google Ads Help). Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment (NBER). Set up Conversion Lift based on users (Google Ads Help). Measuring marketing performance (Shopify Help Center). Set up Conversion Lift based on geography (Google Ads Help).
Related answers
Frequently asked questions
Should I use a fixed percentage rule to split my budget?
As a habit it beats guessing: most money where it works, some on promising bets, a little on new ideas. As a rule, no. It ignores your margin, your buying cycle and what each channel adds, so test the big slices before you trust them.Should I give more budget to the channel with the highest ROAS?
Not on platform ROAS alone. Each platform counts the sales it touched, so one sale can sit in several reports. Check that the channel clears your break-even ROAS, then test whether its sales are extra before you add money.How often should I change my channel split?
At budget cycles and after each test, in small steps. Google's lift guidance says to test before major budget decisions. Small moves keep cause and effect readable, while one big swing makes it hard to tell what changed sales.
Go deeper: Causal attribution, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Keep reading
Terms in this article
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- ExperimentsExperiments are scientific procedures that test hypotheses or demonstrate facts. In marketing, experiments like A/B tests determine the causal effect of campaign changes, enabling data-driven decisions.
- Lift MeasurementLift Measurement: A method to determine the incremental impact of a marketing campaign by comparing exposed and control groups.
- Marketing MixThe marketing mix is the set of actions a company uses to promote its brand or product. It traditionally includes product, price, place, and promotion.
- Marketing Mix ModelingMarketing Mix Modeling (MMM) is a statistical analysis that estimates the impact of marketing and advertising campaigns on sales. It quantifies each channel's contribution to sales.
- Media Mix ModelingMedia Mix Modeling is a statistical technique that measures the collective impact of marketing and advertising on sales. It uses historical data to inform budget allocation.
- MER vs ROASMER (marketing efficiency ratio) is total revenue divided by total ad spend; ROAS is revenue credited to a specific channel divided by that channel's spend. MER cannot be gamed by attribution but is blind to channels; ROAS is channel-level but only as honest as the attribution behind it.
- Word of MouthWord of Mouth is the passing of information from person to person through oral communication. It is one of the most trusted forms of marketing.