Can I A/B test campaigns in Atomato?

No. Atomato has no built-in A/B testing, so it can't split shoppers between two versions of a message. You can still compare two versions by running them one after the other and reading each campaign's results.

Why two live campaigns are not a test

Atomato shows every active campaign to every shopper who matches its targeting. If A and B are live at once, the same shoppers see both, so neither result stands on its own. On Free, only one NotiFeed and one NotiPop campaign can be active at a time anyway.

Make version B

1. Click Campaigns in the Atomato sidebar, find version A, point to Edit in the Settings column and choose Duplicate as draft. The copy opens in the editor with the same messages and targeting.
2. Rename it under Campaign name, for example by adding "B".
3. On the Setup tab, change one thing, such as the title, the image or the button text.

Leave the Design button alone. It styles the whole feed, or every NotiPop, so changing it changes A too.

Run them one after the other

On Scale and Enterprise+, schedule both versions:

1. In version A, open the Schedule tab.
2. Choose Run the campaign according to the schedule with a defined start and end date, tick Set end date and set when A ends. If A is live, leave its start as it is.
3. In version B, set the start to A's end, and an end that gives B the same length of time.
4. Click Save campaign on each.

Set B's dates before you save it: if A had a schedule, saving the copy unchanged puts it live at once. New start times must be later than now. See how to schedule a campaign to start and end.

On Free, the Schedule tab is locked, and saving B while A is active shows Starter plan limit reached (Starter is the app's name for Free). Switch by hand: when A's period is over, click its name in the campaign list, click Pause campaign and confirm. Then make your change to B, save it, and click Start Campaign on its summary page. Pause B after the same length of time.

Compare the results

Click each campaign's name in the campaign list to open its Campaign summary. Set the date range to the days that version ran, then compare Impressions, Clicks, Click rate (CTR) and Confirmed revenue. Analytics update every 6 hours (UTC), so wait until B's last day has been counted. See how Atomato attributes revenue for how orders are credited.

Read the results fairly

• Run each version for the same number of whole weeks, so both cover the same weekdays.
• Change only one thing, so you know what made the difference.
• Treat a gap between the two as a reason to test again, not as proof.

For ideas on what to change first, see how to improve a campaign that is not getting clicks.