Very Small Mailing: Should I Even Split Test?

14 replies
Hi all;

I'm doing a targeted mailing of about 30 letters to a group of my ideal prospects. I'd like to test a couple of ideas for the campaign, but the mailing is very small and I can not justify a bigger mailing at this time (I'm going after clients for a new venture and can't handle more than 1-2 to start. This is a multi-step campaign so I will be contacting them repeatedly).

Is there any reason to test more than one mail campaign on a list of 30?

Thanks
#mailing #small #split
  • Profile picture of the author wilder1047
    I personally can't see this giving any sort of insights into how well the piece converts.

    I'd test one piece with atleast 100 mailings alone.
    {{ DiscussionBoard.errors[7078366].message }}
  • Profile picture of the author iAmNameLess
    No, that is way too small of a sample size. Even 100 each is too small. In my opinion, 1,000 each is too small but I guess that would give you an idea.
    {{ DiscussionBoard.errors[7078540].message }}
    • Profile picture of the author Just Curious
      I'm thinking about doing a B2B postcard direct mailing. What would a good number of pieces to test my initial mailing?
      {{ DiscussionBoard.errors[7084849].message }}
      • Profile picture of the author iAmNameLess
        Originally Posted by Just Curious View Post

        I'm thinking about doing a B2B postcard direct mailing. What would a good number of pieces to test my initial mailing?
        1,000 minimum in my opinion. Anything less is really too small of a sample size.
        {{ DiscussionBoard.errors[7085645].message }}
        • Profile picture of the author Irish Intuition
          Well, I don't know what is up for sale but not too long ago
          I would not let my client mail more than 40 letters.

          Part of the reason I limited the run is because I wanted them
          to implement a follow up (too many letters equaled too much
          follow up).

          We mailed the first 40, turned that into $32,000 in sales. Mailed
          another 40 and ended up at $98,000 in sales (this number is
          padded a bit being that one $15,000 sale was a referral from one
          of the recipients.

          This was for a big ticket service, but it goes to show you don't
          need to mail 1000's to find out. I find that ridicules.

          If you're sending postcards, different story.

          I once created a postcard and my client got crazy and sent out
          100,000 of them.... I was furious. Luckily it did well.
          Signature




          {{ DiscussionBoard.errors[7088061].message }}
          • Profile picture of the author ewenmack
            Originally Posted by Irish Intuition View Post

            Part of the reason I limited the run is because I wanted them
            to implement a follow up (too many letters equaled too much
            follow up).
            Was the follow up by phone or multi mailings to same 40?

            Thanks.
            Ewen
            {{ DiscussionBoard.errors[7088561].message }}
            • Profile picture of the author Irish Intuition
              Originally Posted by ewenmack View Post

              Was the follow up by phone or multi mailings to same 40?

              Thanks.
              Ewen
              The follow up for this project was by phone.
              Signature




              {{ DiscussionBoard.errors[7088634].message }}
  • Profile picture of the author iamrichard
    i agree both above posts. 30 is too small to have a valid conclusion. I would be using split test with mails of at least in range of 300 -500.
    hope it helps
    {{ DiscussionBoard.errors[7082478].message }}
  • Profile picture of the author sundaymorning
    I heard for split test you are suppose to at least do 500 to see results on it. Minimum.
    {{ DiscussionBoard.errors[7082709].message }}
  • Profile picture of the author MyNiches
    It's too small to really draw any conclusions, but it's not too small to *try*. Why not send 15 to one group, 15 to another group, do your split testing and see? If one group gets 5 responses, and the other gets none, you'll at least have a better idea what to send out next time than if you had sent all 30 to one group and gotten 1 or 2 responses.
    {{ DiscussionBoard.errors[7084863].message }}
  • Profile picture of the author SJJPFTW
    Yeah don't worry about split testing. The sample size is waaay too small to give a result that is meaningful, unless you get a MASSIVE swing in results i.e no sales from one sample all from another.

    Chances are the difference would not be that big I wouldn't bother with the extra effort.
    Signature

    If you want to learn more about making money with small business owners for FREE click here to be notified when review copies of my new offline course are available.

    Please remember if something I do helps you, the thanks button helps me :-) ------------------>

    {{ DiscussionBoard.errors[7088137].message }}
  • Profile picture of the author Ouroboros
    I was a Manufacturing Engineer (i.e. Industrial Engineer) for 10 years and often had to make decisions based on a very small universe of samples. If I were only contacting 30 prospects, and had two possible emails, I would split the group. The difference I guess is your control and analysis.

    Split-Testing is a science in itself, but I'd rather send 2 messages to 30 people and get a hit on one of them, than just send the same message to 30 and not hear anything.
    Signature

    Need a Simple Product/Service to Market to Offline Clients? Sell Them DFY Custom Videos. https://www.fiverr.com/users/gigsiteguy

    {{ DiscussionBoard.errors[7088398].message }}
  • Profile picture of the author Jool
    maybe 100 emails so you can do split testing. what software or application you are using to send emails?
    Signature

    We Focus on Your Growth, You Focus on Finding Clients While Label SEO Services

    {{ DiscussionBoard.errors[7088488].message }}

Trending Topics