Data Segmentation

Discussion in 'Noob Central' started by bumpies, May 22, 2011.

  1. bumpies

    bumpies VIP

    Joined:
    Apr 18, 2011
    Messages:
    122
    Likes Received:
    1
    Trophy Points:
    18
    Hey Guys,

    I'd love to be able to pick your brains on formulating a data segmentation and sending strategy.

    I have a couple of servers up and running so want start getting down and dirty optimizing my data.

    I have some questions to start off with:

    1. Do you segment openers and clickers or just clickers?

    2. Do you remove these from the list or just add them to a new list.

    3. How deep do you segment your clickers and openers?
    - a single clickers and openers file?
    - high level files (bizopp, weight loss, shopping etc..)
    - low level files (eg individual campaigns)

    4. If you are doing low level files, I'm assuming you keep your clickers in the big data set files and mail them multiple times from both the clickers list and the big list?

    5. If you have replicated data across 3 list say original data, openers and clickers how to you treat them when mailing different offers so you ensure you are not hitting the same valuable clicker emails many times over?

    6. At what point do you dump a list (if ever) after you have removed all the clickers and openers?

    I'm sure ill have more questions but any help on these would be most appreciated :D

    Thanks guys!
     
  2. bumpies

    bumpies VIP

    Joined:
    Apr 18, 2011
    Messages:
    122
    Likes Received:
    1
    Trophy Points:
    18
    Can anyone help me out with some insights to how they do this? :)
     
  3. gspot

    gspot VIP

    Joined:
    Apr 8, 2011
    Messages:
    616
    Likes Received:
    14
    Trophy Points:
    18
    Location:
    I live here
    Don't go too crazy with segmentation or you will limit your results, so opener & clickers can go together. However, if you only want to segment your clickers make sure you query the positive clicks not negative (advertiser unsub)

    I always keep this file separate or you'll double your work load later

    Again, don't over-think the segmentation especially if this is GI. Segmention comes down to what you are looking for as your final result. Therefore, if you want to target woman for a new shopping add, then yes pull the o/c from a similar offer. If not, then pull from a date range, normally your want the freshest data so no more than 30day

    Yes, thats one way to go.However be careful who you are sending this to, if its a WListed domain then this would be a terrible strategy as you will piss off loads of your recipients, however if its GI go for it!

    Cut your O/C list so it distributes across all your drops evenly.

    I clean-up every 30days, if a recipients hasn't o/c in over 90 I dump them FAST!!
     
  4. bumpies

    bumpies VIP

    Joined:
    Apr 18, 2011
    Messages:
    122
    Likes Received:
    1
    Trophy Points:
    18
    Thanks gspot!

    For number 2 on separating the file, do you remove your openers and clickers from the master file? so you can slowly cut out the openers and clickers from the master file? Or do you just hit the master file for 30 days then extract the openers and clickers?
     
  5. gspot

    gspot VIP

    Joined:
    Apr 8, 2011
    Messages:
    616
    Likes Received:
    14
    Trophy Points:
    18
    Location:
    I live here
    it all depends if you've mailed the list already, if its NOT raw data then YES- you should pull from the master file. However, if its raw data that's never been mailed, then mail the data for 30days and then separate the o/c.
     

Share This Page