FAQ Campaign setup

Can I test pages that need to be behind login?

Yes, you can test pages that need to be behind login (identified to a database or behind an .htaccess file) by loading the source code of the page concerned into the editor. The formatted page will then be displayed in the editor so you can modify it as you would for other tests. A Chrome extension simplifies the process.

Can I delete a variation of a test that is running?

The general principle governing AB Tasty’s operation requires that a visitor assigned to a variation must see the variation during the entire test. If you tried to delete a variation during a test, we couldn’t arbitrarily assign the visitor to another variation without distorting the test data. This is why it is not possible to delete a variation once a test has been started. If you try to delete a variation, you won’t get an error message, and the test will still be present in the editor interface and the report. If you want to delete a variation, you must first stop the test, duplicate it, delete the variation in a new test, and start the new test.

Can I test several successive pages in the customer path or conversion funnel?

Yes, you can carry out this type of test, as long as you have a Premium account. This is particularly useful if you want to modify an item on the first page of the funnel and that modification needs to appear on subsequent pages. The user who is directed to version 2 of the page will be automatically directed to version 2 of the second page, and so on.

Can I test similar pages even if there is no homogeneity in their URLs (i.e. no common pattern)?

Yes, of course. Targeting by URL is the default option because in 80% of cases it is possible to target the required pages. Otherwise, you can use advanced targeting methods that trigger the test depending on the presence of a cookie, a variable JavaScript, or an element in the page’s source code.

Can I assign an exact percentage of my traffic to each version?

Yes, you can set the % of your traffic that will be allocated to each version. By default, the traffic is equally shared between the original and the variations (50/50 if there’s just one variation plus the original, 33/33/33 if there are 2 variations, etc.), but that is fully configurable in the test options.