Normal 0 false false false MicrosoftInternetExplorer4
Hi Folks,
We would love to conduct an A/B Test on an offer with multiple segments in Tessitura. Here's the breakdown:
- Two different emails in Wordfly with two different offers and codes
- In Tessitura, we have multiple segments in the extraction, but we need to promote only 2 segments. We can’t extract using a list because that would only generate one source code for us.
Question: How can conduct an A/B Test in Wordfly when there are multiple segments in an Extraction and we need to promote only two source codes? Is this even possible?
Thanks!
Bradley SanchezRoundabout Theatre Company
I'm not a Wordfly user, but as far as limiting source codes, if I am understanding the question, it sounds like you need to do an extraction that only has two segments. Do you need to combine some segments and save those to a list, combine other segments and save those to a list, and then pull a new extraction that has one segment as one list and one segment as the other list? This would give you only two source codes. A weird kind of workaround, but it is something I have done in the past to separate things out that way. And then I only promote one of the extractions. But, again, I don't use Wordfly. Not knowing more details about your extraction, that's what I could come up with. :o)
Couldn't you save the two segments as two separate lists and promote them separately?
There are multiple different segments in the extraction, but I only need to promote two of them. On top of that, they need to be random (which can occure in Wordfly) because of the A/B Testing.
If you are only promoting two segments, what are you doing with the people in the other segments? If you are mailing them something then ideally they should get a source code promoted. If you are not mailing to them, make them suppression segments and they won’t get included in the output.
As for the random, there are 2 random features in extractions. Random Select is applied when generating counts and is used to randomly distribute constituents between segments, which is generally what you want to do for A/B testing. So each segment would have the same criteria, and then you would use Random Select to randomly distribute people between the two segments and the two mailings. Random Extract is applied upon extraction and randomly chooses which constituents in the segment are output. That’s generally more for capping the number of constituents in a segment though than A/B testing (like when there are 5000 assigned to a segment but you can only mail to 2000 and you want to pick which 2000 randomly).
Kevin Sheehan
Senior Documentation & Learning Resources Specialist
Tessitura Network
+1 888 643 5778 x 329
ksheehan@tessituranetwork.com
Thank you all for your help! Here's what I did:
- I saved the extraction as a list
- Created a new extraction and used the list in the two segments
- Random selected the first segment with 50%
- Promoted the lists, creating two source codes to use for the two different offers
My reasoning for multiple segments but only two promotions is that I'm trying to filter and target different constituencies as precisely as possible. For this particular offer, I broke down multiple segments based on past purchase behavior, but wanted to create only two sources reflecting the two offers. Now that the constituents have been randomly selected in Tessitura, Wordfly doesn't have to randomly select for me; I can just place each segment in its respective email offer.
Bradley – I’m only just getting going in WordFly, but my understanding is that their A/B testing would work independently of anything else you’ve set up in Tessitura. It takes your list (however loaded in) and splits it into two groups randomly. Using the built in testing won’t acknowledge anything you’ve preset in Tessitura.
However, you can, of course, just run two campaigns and set each of them up with the right list and the right insides. Within Extractions, you can use the Random Select/Extract as Kevin explained.
If you want to talk through this more, you can give me a call.
I would also suggest emailing the team at WordFly. They’ve done a nice upgrade from TMS to WordFly, but it’s feedback like this that will help move it to even better integration with Tessitura (and its bells and whistles for targeting and randomizing).
Jamie
From: Tessitura Marketing Forum [mailto:forums-marketing@tessituranetwork.com] On Behalf Of Bradley SanchezSent: Tuesday, August 16, 2011 12:17 PMTo: Jamie O'BrienSubject: [Tessitura Marketing Forum] Mulitple Sources in A/B Testing
This message was sent automatically to you by www.tessituranetwork.com because you subscribed to the Tessitura Marketing Forum. You may reply to this message to post to the Marketing forum or visit the site to search, read and post to the forums. In the interest of keeping the forum posts from becoming cluttered, we encourage you to delete previous message text from your reply before sending. Thank you!