Skip to main content

A/B test goals

Every test has one primary goal. The goal defines what Analytics treats as success for lift and decisions.

Available goals also depend on the test type:

Test typeGoals available
Atomicat page variantAll goals
Split URLAll goals (tracked on each variant’s Atomicat page after redirect)
External redirectTraffic split only (external hosts are not tracked by Atomicat)

Available goals​

GoalWhat it is forExtra field
Traffic split onlyDivide traffic between variants. Atomicat does not treat a conversion goal for winner/lift. Use when you measure sales in Hotmart, Stripe, etc.None
Checkout clickVisitor clicks an Atomicat Checkout button on the page.None
Any button clickVisitor clicks a button, link, or other important clickable on the test page.None
Click selectorVisitor clicks an element matching your CSS selector.CSS selector
URL visitVisitor reaches an Atomicat thank-you/next page, or clicks an outbound link whose URL matches your fragment.Atomicat page or external URL text

Which goal should I pick?​

  • Selling with the Atomicat Checkout component → Checkout click.
  • Optimizing any CTA / link engagement on the page → Any button click.
  • Optimizing one specific button or link → Click selector.
  • Sending visitors to a thank-you or next Atomicat page → URL visit → Atomicat page.
  • Counting clicks out to a payment or external URL → URL visit → External URL.
  • Only splitting traffic and measuring elsewhere → Traffic split only.
Traffic split only

In Analytics you will see visitors, page views, and traffic health (SRM). Conversion rate, lift, and “best variant” are not used for this goal — that avoids confusing numbers from refreshes.

Tips for goals with extra fields​

  • URL visit (Atomicat page): pick a published thank-you or next-step page. After the visitor hits the test, Atomicat keeps context (cookie + same-site links) so a normal click through to thank-you usually attributes.
  • URL visit (External): enter text that appears in the outbound link (for example part of your Hotmart URL). Atomicat counts the click on the test page, not a page view on the external host.
  • Click selector: use a stable selector like #buy-button or .cta-primary. Avoid selectors that change between page edits.

Frequently asked questions​

Can a test have more than one goal?

Each test has one primary goal, which drives the lift and decision calculations. Checkout clicks appear as their own summary card only when the primary goal is Checkout click.

Can I change the goal after the test starts?

Not while it is running. Goal, variants, and traffic split are locked so results stay comparable. Archive the test and create a new one with the new goal.

Why are my conversions zero?

Check that the goal matches something that actually happens on the variant pages: the CSS selector matches an element, the Atomicat goal page is reached with experiment context, or the outbound link contains the external URL text you configured.

Should I use Traffic split only?

Yes when you only need Atomicat to send people to different pages and you judge winners in your payment or CRM tools. If you want Atomicat to compare conversion rates, pick Checkout click, Any button click, Click selector, or URL visit.