Tracking and Logging Your A/B Tests

Highlighted

Tracking and Logging Your A/B Tests

Hi All,

Quick Question - what is the best tool to use to track the results of your email A/B tests?

If there is no efficient tech tool and you track it manually, what is the best way to go about it? Are there available templates?

Thanks!

Chris

Elizabeth Robichaud

4 REPLIES 4
Highlighted
Marketo Employee

Re: Tracking Your A/B Tests

Hi Chris Kowalczyk ,

If you are referring to Email AB test in Marketo, then the email program "dashboard" is where you see the test results. It would also be useful to go through the below URL's

Use the Email Program Dashboard - A/B Test View - Marketo Docs - Product Documentation

Creating an A/B Test program in Marketo

Hope that helps

Best Always,

Karan Hari

Highlighted

Re: Tracking Your A/B Tests

Hi Karan Hari​, Bram Valdez​, and Veronica Holmes​,

I guess I didn't word it appropriately, but I meant more in terms of logging the results. For example, throwing it into an Excel doc and including the test variables, hypothesis, audience, test type, test length, item tested, etc. So going back to my question, if you're logging your results, do you use any type of technology to help with this process or if you do it manually, like mentioned above, do you have anything to add to it? Best practice ideas? Templates?

Thanks again,

Chris

Highlighted
Level 2

Re: Tracking Your A/B Tests

Hi Chris Kowalczyk​,

The A/B Test program is a bit rigid for my taste. If you want to run a straight-up A/B test without having to declare a winner as part of the flow, you can manually create a program using a 50/50 random sample in your own flow.

You can also test your results here if you are uncertain if your sample set is too large or too small to declare your own "winner": https://measuringu.com/ab-cal/

Highlighted
Level 7 - Champion Alumni

Re: Tracking Your A/B Tests

Or just use the A/B test but choose to test on 100% of the audience, that way you still get to use the email program dashboard which random sample won't give you and you don't need to send a "winner".