Run a Google Ads Search experiment

Search supports custom experiments, which split one campaign's traffic between a control arm that bids the way your account normally does and a treatment arm that bids on Scowtt Custom Goal. It's the cleanest way to prove the result, because both arms compete in the same auctions over the same period.

How it works: You build a custom experiment on your existing Search campaign and point the treatment arm at Scowtt Custom Goal. Scowtt reviews the setup and reads the result with you against your real funnel outcomes.

Before you start

Scowtt Custom Goal created in your account. See Set up your Google Ads test campaign.
A base campaign that's active, on its own daily budget (not a shared budget), on a standard bid strategy (not a portfolio strategy), and already using Smart Bidding.
Enough conversion volume. Google's guidance for value-based bidding tests is at least 50 conversions in the last 30 days, and experiments need more volume than that for reliable results. Low-volume accounts won't produce a readable Search experiment.

Steps

You: Create a custom experiment

Go to Campaigns > Experiments, click the plus button above the All experiments table and select Custom. Choose the experiment and the variable being tested, click Next, then Select campaigns and confirm the base campaign.
You: Set up the treatment arm

Set the conversion goal to Scowtt Custom Goal and the bid strategy to Maximize conversion value. Add a target ROAS only if your account already has one you can stand behind. Leave the control arm untouched.
You: Check the experiment settings

Under Advanced options, confirm Experiment sync is on. It's on by default and can't be changed after the trial campaign is created. Set the traffic split to 50/50 (cookie-based where offered), set Success metrics to conversion value, and set a duration of at least 44 days.
You: Leave auto-apply off

Setup offers to push the winning arm into the original campaign automatically. Leave this off: the decision to scale should be made together with your Scowtt team, not by the platform.
Scowtt: Review the setup and read the results with you

Your Scowtt team checks the experiment before launch, then reads results with you using the ground truth funnel-stage columns.

Ending the experiment

To keep the result, either apply the experiment to the original campaign or convert it into a new campaign. Performance data is kept either way. To end it without applying, use Edit > End now, or bring the end date forward.

Also available: Campaign Mix experiments (early access)

Google is introducing Campaign Mix experiments, which test up to five groups of campaigns against each other in one experiment, including across campaign types such as Search, Performance Max and Demand Gen. Like a custom experiment, traffic is split between the groups, 50/50 by default.

They're currently in early access, so they aren't available in every account yet. Scowtt's recommended approach is still the setup on this page, while Campaign Mix is proven at scale. If it's available in your account and you'd like to use it, talk to your Scowtt team first so the test still compares Scowtt's signal fairly. Google's guide: About Campaign Mix experiments.

Good to know

With sync on, check after any later edit that the treatment arm is still on Scowtt Custom Goal.

Who does what

Step	Who	What changes hands
Create a custom experiment	You	—
Set up the treatment arm	You	—
Check the experiment settings	You	—
Leave auto-apply off	You	—
Review the setup and read the results with you	Scowtt	Setup review, then results