Custom conversion tracking on a redirect A/B test

pmeyerdk 07-02-2018

Hi all,

I want to setup an A/B test in Target, which redirects the user to another page (e.g. the variant made outside of Target).

That is quite easy, so no problems there

But now the problem arises:

I want to track goal conversion using Target and not Analytics.

In the control I could possibly add some sort of JavaScript for tracking conversions, since the conversions doesn't have a separate receipt page.

But since the variant is redirecting the user to a different page, it isn't possible to add script code in the UI, where we can do almost everything, when not creating redirects.

I don't have any experience with setting up custom conversion goals, but still guess that would be the way to go, if possible for a redirect variant, as it seems to be in regards to the non-redirected control.

It doesn't seem easily possible to find examples for this use case.

Can anyone help me further with this?

Thanks

Best regards,Peter Meyer

Accepted Solutions (1)

Accepted Solutions (1)

SarahEOwen 10-02-2018

Hi -

Glad you are able to usually use the benefits of A4T   To set-up Target-metrics, you will want to use the Rules in DTM.

Here's an example of setting up an mbox on a page to use as a Target-metric. Steps #1 - 6 will create an mbox on the page you are interested in. Steps #7-13 will use the mbox you just made in steps #1-6 as the Target-metric.

1. In DTM, open a Page Load Rule

2. Give it a name like "Target - Order Confirmation Page"

3. In the Conditions portion of the rule, put in the code that denotes the Order Confirmation page is loaded/is being shown (i.e. URL contains a specific value .... or the dataLayer contains specific values)

4. In the Adobe Target portion of the rule:

     a. In the Create Mbox Around field, enter #orderConfirmPageMboxDiv

     b. In the Mbox Name field, enter orderConfirmPage

     c. In the Mbox Parameters, pass any values from the page to Target (this is optional)

5. In the JavaScript/Third Party Tags portion of the rule:

     a. Go to the Sequential HTML tab and click Add a New Script

     b. Give it a name like Target - Checkout - Order Confirmation div

     c. Put in this code  <div id="orderConfirmPageMboxDiv" style="display:none;visibility:hidden;"></div>
           Note: whatever you used as the value in #4a above needs to be value of the id in the code

6. Leave a comment about what you did in the Add a Note section and hit Save

7. Now go to Target.

8. Get to the Goals & Setting part of your Test.

9. In the Reporting section, click Add a New Metric

10. In the first drop-down, select Conversion (because this example is about getting to the Order Conf. page)

11. In the second drop-down, select Viewed an mbox.

12. In the drop-down that appears, select orderConfirmPage (or whatever you used as the value in #4b above)

13. Finish creating your activity.

Here are some helpful articles from Adobe too:

Mbox rule example for Dynamic Tag Management   (page-load rule)

Mboxes on JavaScript Events  (event based rule)

Mboxes

Thanks -

Sarah

Answers (3)

Answers (3)

pmeyerdk 12-02-2018

Thank you Sarah!

Though we are not using DTM, but Tealium iQ, this is very helpful input.

I understand that I have to create en Mbox for the conversion, and then use that as a conversion point.

Thanks again 🙂

Best regards,

Peter

SarahEOwen 07-02-2018

Hi Peter -

Would you mind if I asked why you don't want to use Analytics as the metrics?

My absolute favorite is to use Analytics as the metrics because then I have every single data point captured in Adobe Analytics available and I don't have to do any additional set-up (i.e. setting up custom conversion goals).

Thanks -

Sarah

pmeyerdk 08-02-2018

Hi Sarah,

Normally, I would also prefer to use Analytics as the metrics.

But from time to time, we cannot always get to the conversion point using Analytics, without setting up new tracking.

Then it would of course make sense to do that, but when looking at release cycles we are placed under, it might be way faster to then just do some of the tests with Target data only.

Best regards,

Peter