14 {{ upvoteCount | shortNum }}
14 {{ upvoteCount | shortNum }}

How do YOU determine quality of traffic when split testing

by Uncle Dimitry
12 replies
This problem finally got me when I got 8% conversion on free squeeze page after I've Spent A Week Writing Headlines And Text (Crap! I Still Start Every Word With A Capital Letter).

... that page was brilliant, but it got bad result because of low quality traffic. Sometimes affiliates send really good visitors and junk pages convert at 40-60%

So the question is: how to avoid mistakes during split testing and correctly measure traffic quality?
#determine #quality #split #traffic
  • Profile picture of the author Tsnyder
    Wouldn't the only measure of quality be the number of visitors
    who performed the act you desire? (opt-in, purchase, etc)

    Tsnyder
    Signature
    If you knew what I know you'd be doing what I do...
    {{ DiscussionBoard.errors[842208].message }}
    • Profile picture of the author Uncle Dimitry
      Originally Posted by Tsnyder View Post

      Wouldn't the only measure of quality be the number of visitors
      who performed the act you desire? (opt-in, purchase, etc)

      Tsnyder
      Yes, but think about this - we get two variables here. Here they go:
      • traffic quality
      • offer quality

      If I make stupid change to the offer which is supposed to kill conversion rate
      ...and at the same time affiliate will post real good review on their blog, it could end up by rate increase
      Signature
      Udimi.com - buy solos
      {{ DiscussionBoard.errors[842233].message }}
  • Profile picture of the author CashTactics
    The only way to really determine how qualified your traffic was would be to look at your conversion rate.

    You can always pre-qualify traffic in your ad text by stating what is required.

    Example:
    Want free diapers for life
    Fill out this quick survey
    Females 18 and Older

    That tells several things. 1. The person can get free diapers for life. 2. It takes a quick survey to enter... If they don't want to fill out a survey they won't click. 3. Only females 18+ qualify. This will keep out unqualifying traffic.

    Your clicks will be less but your conversion rate will increase. Try it out. Might be worth a shot.

    Kris
    Signature
    Cashtactics.net - Strategic Marketing at Warp Speeds!
    {{ DiscussionBoard.errors[842217].message }}
    • Profile picture of the author Uncle Dimitry
      Originally Posted by CashTactics View Post

      The only way to really determine how qualified your traffic was would be to look at your conversion rate.

      You can always pre-qualify traffic in your ad text by stating what is required.

      Example:
      Want free diapers for life
      Fill out this quick survey
      Females 18 and Older

      That tells several things. 1. The person can get free diapers for life. 2. It takes a quick survey to enter... If they don't want to fill out a survey they won't click. 3. Only females 18+ qualify. This will keep out unqualifying traffic.

      Your clicks will be less but your conversion rate will increase. Try it out. Might be worth a shot.

      Kris
      Kris, that looks like filtering low quality traffic at a stage of ads, before they even hit the site. My problem comes when they already got to my site (aaaaargh )
      Signature
      Udimi.com - buy solos
      {{ DiscussionBoard.errors[842385].message }}
  • Profile picture of the author esh
    how did you know that was a bad traffic?

    EDIT :

    There are umpteen number of reasons for a conversion to happen and not happen.If my particular segment of traffic that i take and see, its small in number, but the conversion is 100% for that segment. So, if you start sending any kind of traffic considering that they are all looking for a solution that your product claims to achieve, you get between 1-10% conversion rates. If i say something has a bad traffic, that means you are getting totally untargeted herds of people that have nothing to do with your product, and they bounce off immediately after coming to your site. So when i see an increase in my bounce rate i look for the kind of that untargeted traffic or bad traffic. If your channel of marketing is search engine, or article or payperclick you must have a fairly decent conversion rate, and you dont have to worry about bad traffic! So i just wanted to know how did you determine what you got is a bad traffic?
    {{ DiscussionBoard.errors[842258].message }}
  • Profile picture of the author JustinBrooke
    Usually the first sign of low quality traffic is a high bounce
    rate in your Google Analytics reports or a low avg time on
    site.

    You should also be able to see where that traffic is coming
    from in your referrer stats. If you see things like traffic exchanges,
    or anything else suspicious you can go into your cpanel and
    block that traffic from hitting your site.

    Also make sure your affiliates know upfront what types of traffic
    you consider wasteful and why and that you'll block it if you see
    it coming in.
    Signature

    You ROCK!,
    Justin Brooke

    FREE: spreadsheet of 182 traffic sources (no opt-in)

    {{ DiscussionBoard.errors[842362].message }}
    • Profile picture of the author Uncle Dimitry
      Originally Posted by JustinBrooke View Post

      Usually the first sign of low quality traffic is a high bounce
      rate in your Google Analytics reports or a low avg time on
      site.
      this might not help since bounce rate depends on the offer itself.

      Originally Posted by JustinBrooke View Post

      Also make sure your affiliates know upfront what types of traffic
      you consider wasteful and why and that you'll block it if you see
      it coming in.
      Great idea. I didn't think about this. I'll even add some keyword ideas for affiliates
      Signature
      Udimi.com - buy solos
      {{ DiscussionBoard.errors[842404].message }}
  • Profile picture of the author ragnartm
    same page, design and copy completely identical, different sources of traffic. Track sales or sign-ups or even clicks. You can do this with something like protrackerplus.com free up to 3 campaigns, or hitsconnect.com paid 20 bucks a month for up to 100 bucks.
    Signature

    Ragnar.

    Quality over quantity. Hire me to write highly shareable, user focused blog posts or articles.

    {{ DiscussionBoard.errors[842381].message }}
  • Profile picture of the author anthony2313
    Banned
    [DELETED]
    {{ DiscussionBoard.errors[842528].message }}
    • Profile picture of the author Uncle Dimitry
      Originally Posted by anthony2313 View Post

      you keep mentioning the traffic but your copy or offer

      could be bad.
      I talk in general, not mentioning concrete ad or copy, just a way to get more accurate split test results
      Signature
      Udimi.com - buy solos
      {{ DiscussionBoard.errors[842659].message }}
  • Profile picture of the author kaisen
    Doesn't google offer a way of split testing different parts of your site for free?
    {{ DiscussionBoard.errors[842687].message }}
  • Profile picture of the author esh
    if people scroll down to read the offer, i dont think your bounce rate will be so high in the range of say 80%

    i recommend you to try crazyegg service and see how the visitor interacts with your website!
    {{ DiscussionBoard.errors[843001].message }}
    • Profile picture of the author MunkMurray
      Hi there,
      I had the same problem lately when testing a squeeze page.
      What i did is copy the page to a different location and use only PPC as traffic to the new location.
      This way I know the traffic quality will stay approximately the same and it doesn't interfere with all my other traffic sources

      hope it helps,
      MM
      Signature
      The Gameon Project - Find your new favorites...
      {{ DiscussionBoard.errors[843095].message }}

Trending Topics