Field Note · Beta Watch · PPC desk
Google's Campaign Mix Experiments. The quiet beta worth a proper look.
Paid search spent the last decade asking the wrong question. Conventional wisdom dictates one campaign wins. The cunning move is testing the whole mix at once, and Google has finally built the tool that lets you do exactly that.
The story starts with a quiet beta. Google Ads has rolled out something rather brilliant, and any brand spending money on paid media should look closely. folkfox sees this one as an antidote to instinct dressed up as insight. Behind the data is a human decision: which mix of campaigns actually earns its place in your strategy? Campaign Mix Experiments, currently in beta, lets you pit different combinations of campaign types and budgets against each other inside one unified experiment, which is precisely the kind of clarity paid search teams have been quietly begging for. Less guessing. More knowing.
Think of it as a fashion shoot with five outfit options, all photographed under the same conditions. You create up to five experiment arms, each one a different mix of campaigns, and Google splits your traffic between them. The results are then normalised, which is the part that matters most for any multi-channel advertiser. You can test across Search, Performance Max, Shopping, Demand Gen, Video, and App campaigns. The only wallflower not invited to the runway is Hotels, which is a small grief, but folkfox can live with it.
Anatomy of an experiment
From one budget to one decision.
This is what a Campaign Mix Experiment looks like when you sketch it out: one budget, five arms, one normalised verdict that your strategy can actually act on.
What can you actually test?
This is where the tool starts earning its place on your paid search shelf. You can interrogate budget allocation across different campaign types, account structure (whether consolidating campaigns or keeping them fragmented performs better), bidding strategies and targeting approaches, and how campaigns interact with each other across channels rather than how one performs in isolation.
That last point is where the cleverness lives. Most testing tools tell you whether one campaign works in a vacuum. This one tells you whether your whole marketing strategy works as a coherent narrative. Different question, far better answer.
Who feels the lift first
The industries that should be running this yesterday.
Most brands stand to benefit, though some verticals feel the impact harder and faster. The industries running multi-channel architectures see the clearest signal, which is why folkfox built playbooks for each.
Compliance-first player acquisition.
Operators juggling Search, PMax, and Shopping in regulated markets get a structured way to prove which mix actually moves first deposits, all while respecting the rules.
Read the playbook →High CAC, high stakes, high scrutiny.
When every saved euro of CAC goes straight to the board deck, evidence-led mix decisions stop being optional and start being the difference between Series B and silence.
Read the playbook →Tour drops, release windows, fandom waves.
Awareness campaigns feeding into App and Search conversion: the full narrative arc of a release cycle has never been clearer.
Read the playbook →Patient acquisition without the cannibalisation.
Lead-led verticals finally get a data-backed way to settle the Search-heavy versus broader-funnel debate, especially in trust-sensitive healthcare communities.
Read the playbook →When attribution is murky, evidence is gold.
App-led brands and SaaS plays can map upper-funnel awareness to lower-funnel conversion inside one experiment, turning "we think it works" into "we have the receipts".
Read the playbook →Boutique brands ready to outfox the algorithm.
If your growth strategy already runs across paid social, paid search, and beyond, this is the conversation worth starting today.
Talk to folkfox →Illustrative · Arm Performance
Where the budget actually moves the needle.
A typical Campaign Mix Experiments readout showing arms compared on ROAS uplift versus the historical baseline. Arm C, the balanced full-narrative mix, outperforms the Search-heavy default by a wide margin.
Illustrative scenario based on a balanced retail mix. Real outcomes vary by industry, audience, and seasonality.
Winning Arm · Composition
The mix inside Arm C.
A balanced full-narrative mix, with no single channel doing the heavy lifting and four campaign types in coordinated rhythm.
Composition shifts by vertical: iGaming arms tend toward PMax, healthcare arms toward Search.
Use it well
Five elegant principles for running a Mix Experiment properly.
This is the folkfox protocol for getting evidence rather than vibes out of every experiment your paid team runs.
-
01
Change one variable at a time.
If you are testing budget allocation, hold every other condition consistent across arms. Changing five things at once is how you end up with interesting data and absolutely no idea what caused it, which is worse than no experiment at all.
-
02
Align total budgets across arms.
Unless budget itself is what you are testing, give each arm the same spend. Comparing a £10k arm to a £40k arm and calling it a fair fight is wishful thinking dressed up as strategy.
-
03
Avoid shared budgets and resist the mid-flight tinkering.
Let the experiment breathe in the nocturnal way good experiments do. Every mid-flight change resets the noise floor and pushes statistical significance further out of reach for your readout deck.
-
04
-
05
Choose your confidence interval and primary metric on day one.
95%, 80%, or 70% are all valid choices, and each implies a different decision risk. Pair it with one primary success metric (ROAS, CPA, conversions, or conversion value) before launch and resist moving the goalposts when results land in your inbox.
The bigger narrative.
What Google is quietly acknowledging is something the cleverest advertisers have known for a while: modern paid performance is no longer about one campaign winning. It is about the right combination working together, and as automation continues blurring the lines between channels, understanding the mix has become the real competitive edge.
Campaign Mix Experiments hands you a cleaner, more honest way to make those decisions, based on what your specific audience in your specific market actually responds to. Less guessing. More knowing. That is a win the folkfox team will happily celebrate.
Want folkfox to run the Mix Experiment for you?
folkfox designs and instruments Campaign Mix Experiments for boutique brands across iGaming, FinTech, healthcare, music, and Web3. You bring the budget. We bring the evidence.
Fresh tracks
Latest from the Google-Ads trail
Read more on this topic
Local Services Ads: the complete 2026 Performance Max guide
Local services ads moved house, not retired, so your Performance Max strategy needs to adapt.
Read the pieceGoogle Ads Performance Planner: 6 Honest Diagnostic Steps
A sudden performance drop usually has a boring diagnosis, and these are the checks that find it.
Read the pieceGoogle Ads Daily Budget Overspend: The Essential 2x Rule
Google's 2x daily budget rule explains most overspend surprises before they become billing disputes.
Read the pieceCampaign Settings Google Ads: 9 Essential Controls
Google Ads offers nine essential controls across account, campaign, and ad group levels.
Read the piece
What will you test in your next mix experiment?
folkfox builds and manages Google Ads experiments that give you clear answers, not guesswork. We can set up your next test and tell you what the data means.