Audience Targeting
Not every experiment should run for every visitor. Audience targeting lets you focus your tests on the segment where the change actually matters.
How to set it up
When creating an experiment, expand the Audience section in Step 1. By default, experiments run for all visitors. Add conditions to narrow it down.
Available targeting options
By device
Run an experiment only for mobile visitors, only for desktop, or both.
Use case: Your mobile checkout has a different UX problem than desktop. Target mobile-only so desktop traffic doesn't dilute the signal.
New vs returning visitors
Test different experiences for first-time visitors vs people who've been to your site before.
Use case: New visitors don't know your brand. Show them social proof and a softer intro offer. Returning visitors already trust you — show them a direct CTA.
Traffic source (UTM)
Run experiments only for visitors arriving from a specific campaign, source, or medium.
Use case: You're spending money on Google Ads sending visitors to a landing page. Run your landing page experiments only for that paid traffic — organic visitors have different intent.
Examples:
- UTM Source =
google— only Google traffic - UTM Medium =
cpc— only paid click traffic - UTM Campaign =
spring-sale— only a specific campaign
Referrer
Target visitors coming from a specific website.
Use case: Visitors from Product Hunt vs visitors from your newsletter behave very differently. Test messaging targeted at each source separately.
Quick presets
| Preset | What it targets |
|---|---|
| All visitors | No restriction (default) |
| New visitors only | First-time visitors to your site |
| Mobile only | Phones and tablets |
| Paid traffic | UTM medium = cpc / paid |
| From Google | UTM source = google |
Effect on sample size
Targeting a smaller audience means the experiment takes longer. If only 30% of your traffic is mobile, a mobile-only experiment needs ~3× as long as an all-visitors experiment.
Datsnip's sample size estimator automatically accounts for this when you enter your baseline and traffic estimate.
Ideas by audience
New visitors — Test your welcome offer. "Get 10% off your first order" vs "Free shipping on your first order." New visitors are the ones who see this — no need to distract returning customers.
Returning visitors — Test a "Welcome back" loyalty offer, or a more direct pitch that skips the intro. They know you.
Paid traffic (CPC) — These visitors arrived with intent. Test bolder CTAs and fewer distractions on landing pages for paid campaigns.
Mobile visitors — Test sticky add-to-cart bars, simplified forms with fewer fields, and tap-friendly button sizes specifically for mobile.