Loading ...
Sorry, an error occurred while loading the content.

Google analytics campaing tracking accuracy

Expand Messages
  • davidehmer
    Just wondering if anyone can explain the results Google Analytics provides for campaign tracking. I have just tested redirects (see below for example) to
    Message 1 of 4 , Jul 2, 2011
    • 0 Attachment
      Just wondering if anyone can explain the results Google Analytics provides for campaign tracking.

      I have just tested redirects (see below for example) to record several parameters from users who will key in several vanity urls (offline promotion)

      REDIRECTS (6 in all):
      redirect 301 /workshops/adelaide-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop

      redirect 301 /workshops/melbourne-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop

      redirect 301 /workshops/sydney-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop

      redirect 301 /workshops/adelaide http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop

      redirect 301 /workshops/melbourne http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop

      redirect 301 /workshops/sydney http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop

      The redirects work as expected. In testing I have used 2 computers and 3 different browsers. Cookies and Javascript are on in all browsers. I visited each or the 6 urls at 3 separate times of day.

      When checking results in GA's Traffic sources > Campaigns, I would have expected to see 3 visits for each source (18 in total) with 1 page per visit.

      However, in Campaign overview the following results display:
      Campaign Source Visits Pages per visit
      Adelaide+Workshop long-letter-adelaide 1 4
      Melbourne+Workshop long-letter-melbourne 1 6
      Sydney+Workshop long-letter-sydney 1 6

      When also displaying 'source', the following displays:
      Campaign Source Visits Pages per visit
      Adelaide+Workshop long-letter-adelaide 1 1
      Melbourne+Workshop long-letter-melbourne 1 3
      Sydney+Workshop long-letter-sydney 1 3
      Adelaide+Workshop long-letter-adelaide 0 0
      Melbourne+Workshop long-letter-melbourne 0 0
      Sydney+Workshop long-letter-sydney 0 0

      Could it be that all activity for each time of time of day was treated as one session and only recorded it as one visit. (hence 3 visits rather than 18) And that it recorded the other activity as subsequent page accesses on those visits.

      If so this makes my testing misleading but in real world practice it should work fine?

      Any thoughts about what is happening here would be most welcome.

      Thanks
      David
    • Yuhui
      Hi David, You mentioned that you tested using 3 different browsers at 6 times of the day. Could you provide more information about: - Did you test the URLs
      Message 2 of 4 , Jul 3, 2011
      • 0 Attachment
        Hi David,

        You mentioned that you tested using 3 different browsers at 6 times of the day. Could you provide more information about:

        - Did you test the URLs consecutively at each time of the day? E.g. Machine 1 Browser A: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
        Machine 1 Browser B: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
        etc.
        - Which URLs did you test with each browser?
        - When were the 3 times of the day? I assume that it's morning/afternoon/night, i.e. a break of more than 30 minutes between each time.
        - You didn't clear any of your browsers' cookies before/after each test, right?

        Also:
        - Can you re-check the reports that you shared with us? In your message, you listed 6 "long-letter" workshops for the 3 cities, but you only have 3 "long-letter" and 3 "short-letter" workshops.
        - Which reports are you looking at? It's strange that Adelaide, Sydney and Melbourne long-letter workshops give 2 different Pages per Visit numbers.
        - How soon after your tests did you check the reports?

        Having this information will help us clarify our thinking and better help you.

        Regards,
        Yu Hui
        Twitter: @yuhui

        --- In webanalytics@yahoogroups.com, "davidehmer" <davidehmer@...> wrote:
        >
        > Just wondering if anyone can explain the results Google Analytics provides for campaign tracking.
        >
        > I have just tested redirects (see below for example) to record several parameters from users who will key in several vanity urls (offline promotion)
        >
        > REDIRECTS (6 in all):
        > redirect 301 /workshops/adelaide-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop
        >
        > redirect 301 /workshops/melbourne-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop
        >
        > redirect 301 /workshops/sydney-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop
        >
        > redirect 301 /workshops/adelaide http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop
        >
        > redirect 301 /workshops/melbourne http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop
        >
        > redirect 301 /workshops/sydney http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop
        >
        > The redirects work as expected. In testing I have used 2 computers and 3 different browsers. Cookies and Javascript are on in all browsers. I visited each or the 6 urls at 3 separate times of day.
        >
        > When checking results in GA's Traffic sources > Campaigns, I would have expected to see 3 visits for each source (18 in total) with 1 page per visit.
        >
        > However, in Campaign overview the following results display:
        > Campaign Source Visits Pages per visit
        > Adelaide+Workshop long-letter-adelaide 1 4
        > Melbourne+Workshop long-letter-melbourne 1 6
        > Sydney+Workshop long-letter-sydney 1 6
        >
        > When also displaying 'source', the following displays:
        > Campaign Source Visits Pages per visit
        > Adelaide+Workshop long-letter-adelaide 1 1
        > Melbourne+Workshop long-letter-melbourne 1 3
        > Sydney+Workshop long-letter-sydney 1 3
        > Adelaide+Workshop long-letter-adelaide 0 0
        > Melbourne+Workshop long-letter-melbourne 0 0
        > Sydney+Workshop long-letter-sydney 0 0
        >
        > Could it be that all activity for each time of time of day was treated as one session and only recorded it as one visit. (hence 3 visits rather than 18) And that it recorded the other activity as subsequent page accesses on those visits.
        >
        > If so this makes my testing misleading but in real world practice it should work fine?
        >
        > Any thoughts about what is happening here would be most welcome.
        >
        > Thanks
        > David
        >
      • davidehmer
        Hi Yu Thanks for your reply. Some clarifying comments to your question below. ... You mentioned that you tested using 3 different browsers at 6 times of the
        Message 3 of 4 , Jul 3, 2011
        • 0 Attachment
          Hi Yu

          Thanks for your reply. Some clarifying comments to your question below.

          -------------------------------------------------------------

          You mentioned that you tested using 3 different browsers at 6 times of the day.
          Could you provide more information about:

          - Did you test the URLs consecutively at each time of the day? E.g. Machine 1 Browser A: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
          Machine 1 Browser B: 12:01pm Adelaide short-letter, 12:03pm Melbourne
          short-letter, 12:05pm Sydney short-letter, etc.
          etc.

          Test 1 was at 10am on machine 1. Only one browser (FF3.6) visited all six urls. (each redirects as noted in previous post)
          Test 2 was at 4pm on machine 2. Only one browser (FF3.6) visited all six urls. (each redirects as noted in previous post)
          Test 3 was at 8pm on machine 2. Only one browser (IE8) visited all six urls. (each redirects as noted in previous post)

          - Which URLs did you test with each browser?

          As above

          - When were the 3 times of the day? I assume that it's morning/afternoon/night,
          i.e. a break of more than 30 minutes between each time.

          As above

          - You didn't clear any of your browsers' cookies before/after each test, right?

          That's right. Did not clear cookies.

          Also:
          - Can you re-check the reports that you shared with us? In your message, you listed 6 "long-letter" workshops for the 3 cities, but you only have 3 "long-letter" and 3 "short-letter" workshops.

          Just noticed that the formatting in my email got munged when posted. Here's that data again in a format that should be better for email

          In Campaign overview the following results display:
          Campaign: Adelaide+Workshop
          Visits: 3
          Pages per visit: 2.33

          Campaign: Melbourne+Workshop
          Visits: 1
          Pages per visit: 7

          Campaign: Sydney+Workshop
          Visits: 1
          Pages per visit: 6

          When also displaying 'source', the following displays:
          Campaign: Adelaide+Workshop
          Source: short-letter-adelaide
          Visits: 2
          Pages per visit: 3

          Campaign: Adelaide+Workshop
          Source: long-letter-adelaide
          Visits: 1
          Pages per visit: 1

          Campaign: Melbourne+Workshop
          Source: short-letter-melbourne
          Visits: 0
          Pages per visit: 0

          Campaign: Melbourne+Workshop
          Source: long-letter-melbourne
          Visits: 1
          Pages per visit: 3

          Campaign: Sydney+Workshop
          Source: short-letter-sydney
          Visits: 0
          Pages per visit: 0

          Campaign: Sydney+Workshop
          Source: long-letter-sydney
          Visits: 1
          Pages per visit: 3

          - Which reports are you looking at? It's strange that Adelaide, Sydney and Melbourne long-letter workshops give 2 different Pages per Visit numbers.
          - How soon after your tests did you check the reports?
          The stats above are about 48 hours old now.

          Regards
          David


          --- In webanalytics@yahoogroups.com, "Yuhui" <yuhuibc@...> wrote:
          >
          > Hi David,
          >
          > You mentioned that you tested using 3 different browsers at 6 times of the day. Could you provide more information about:
          >
          > - Did you test the URLs consecutively at each time of the day? E.g. Machine 1 Browser A: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
          > Machine 1 Browser B: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
          > etc.
          > - Which URLs did you test with each browser?
          > - When were the 3 times of the day? I assume that it's morning/afternoon/night, i.e. a break of more than 30 minutes between each time.
          > - You didn't clear any of your browsers' cookies before/after each test, right?
          >
          > Also:
          > - Can you re-check the reports that you shared with us? In your message, you listed 6 "long-letter" workshops for the 3 cities, but you only have 3 "long-letter" and 3 "short-letter" workshops.
          > - Which reports are you looking at? It's strange that Adelaide, Sydney and Melbourne long-letter workshops give 2 different Pages per Visit numbers.
          > - How soon after your tests did you check the reports?
          >
          > Having this information will help us clarify our thinking and better help you.
          >
          > Regards,
          > Yu Hui
          > Twitter: @yuhui
          >
          > --- In webanalytics@yahoogroups.com, "davidehmer" <davidehmer@> wrote:
          > >
          > > Just wondering if anyone can explain the results Google Analytics provides for campaign tracking.
          > >
          > > I have just tested redirects (see below for example) to record several parameters from users who will key in several vanity urls (offline promotion)
          > >
          > > REDIRECTS (6 in all):
          > > redirect 301 /workshops/adelaide-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop
          > >
          > > redirect 301 /workshops/melbourne-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop
          > >
          > > redirect 301 /workshops/sydney-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop
          > >
          > > redirect 301 /workshops/adelaide http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop
          > >
          > > redirect 301 /workshops/melbourne http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop
          > >
          > > redirect 301 /workshops/sydney http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop
          > >
          > > The redirects work as expected. In testing I have used 2 computers and 3 different browsers. Cookies and Javascript are on in all browsers. I visited each or the 6 urls at 3 separate times of day.
          > >
          > > When checking results in GA's Traffic sources > Campaigns, I would have expected to see 3 visits for each source (18 in total) with 1 page per visit.
          > >
          > > However, in Campaign overview the following results display:
          > > Campaign Source Visits Pages per visit
          > > Adelaide+Workshop long-letter-adelaide 1 4
          > > Melbourne+Workshop long-letter-melbourne 1 6
          > > Sydney+Workshop long-letter-sydney 1 6
          > >
          > > When also displaying 'source', the following displays:
          > > Campaign Source Visits Pages per visit
          > > Adelaide+Workshop long-letter-adelaide 1 1
          > > Melbourne+Workshop long-letter-melbourne 1 3
          > > Sydney+Workshop long-letter-sydney 1 3
          > > Adelaide+Workshop long-letter-adelaide 0 0
          > > Melbourne+Workshop long-letter-melbourne 0 0
          > > Sydney+Workshop long-letter-sydney 0 0
          > >
          > > Could it be that all activity for each time of time of day was treated as one session and only recorded it as one visit. (hence 3 visits rather than 18) And that it recorded the other activity as subsequent page accesses on those visits.
          > >
          > > If so this makes my testing misleading but in real world practice it should work fine?
          > >
          > > Any thoughts about what is happening here would be most welcome.
          > >
          > > Thanks
          > > David
          > >
          >
        • Yuhui
          Hi David, Apologies for the late reply. I haven t really encountered this nor gave it any though before, so what I m saying next is my best guess. Your results
          Message 4 of 4 , Jul 8, 2011
          • 0 Attachment
            Hi David,

            Apologies for the late reply.

            I haven't really encountered this nor gave it any though before, so what I'm saying next is my best guess.

            Your results are due to attribution. The part that clued me in was the 0 visits for some campaign sources.

            That sounds suspiciously like what is a known "phenomenon" with the Keywords report. Some keywords report 0 visits as well. In a Google Analytics video, Avinash explained that such results were because some people came to the site via another source (e.g. referral), then within 30 minutes, came back to the site via an organic search. The first source (i.e. referral) got the credit for the visit, leaving the organic search keyword with 0, so that in total, it is still 1 visit per person.

            Assuming that the above scenario extends to *all* traffic source reports, then that would explain your results. You visited your 6 campaign URLs within the same session, so GA is attributing the results to the various campaigns to ensure that the total number of visits remains the same.

            Your pages/visit results would also reflect this, i.e. I suspect that you were almost always likely to start your test with "Adelaide+Workshop / short-letter-adelaide" and end with "Melbourne+Workshop / short-letter-melbourne" or "Sydney+Workshop / short-letter-sydney".

            Hope I'm providing you with the right information and that it makes sense!

            Regards,
            Yu Hui
            Twitter: @yuhui

            --- In webanalytics@yahoogroups.com, "davidehmer" <davidehmer@...> wrote:
            >
            > Hi Yu
            >
            > Thanks for your reply. Some clarifying comments to your question below.
            >
            > -------------------------------------------------------------
            >
            > You mentioned that you tested using 3 different browsers at 6 times of the day.
            > Could you provide more information about:
            >
            > - Did you test the URLs consecutively at each time of the day? E.g. Machine 1 Browser A: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
            > Machine 1 Browser B: 12:01pm Adelaide short-letter, 12:03pm Melbourne
            > short-letter, 12:05pm Sydney short-letter, etc.
            > etc.
            >
            > Test 1 was at 10am on machine 1. Only one browser (FF3.6) visited all six urls. (each redirects as noted in previous post)
            > Test 2 was at 4pm on machine 2. Only one browser (FF3.6) visited all six urls. (each redirects as noted in previous post)
            > Test 3 was at 8pm on machine 2. Only one browser (IE8) visited all six urls. (each redirects as noted in previous post)
            >
            > - Which URLs did you test with each browser?
            >
            > As above
            >
            > - When were the 3 times of the day? I assume that it's morning/afternoon/night,
            > i.e. a break of more than 30 minutes between each time.
            >
            > As above
            >
            > - You didn't clear any of your browsers' cookies before/after each test, right?
            >
            > That's right. Did not clear cookies.
            >
            > Also:
            > - Can you re-check the reports that you shared with us? In your message, you listed 6 "long-letter" workshops for the 3 cities, but you only have 3 "long-letter" and 3 "short-letter" workshops.
            >
            > Just noticed that the formatting in my email got munged when posted. Here's that data again in a format that should be better for email
            >
            > In Campaign overview the following results display:
            > Campaign: Adelaide+Workshop
            > Visits: 3
            > Pages per visit: 2.33
            >
            > Campaign: Melbourne+Workshop
            > Visits: 1
            > Pages per visit: 7
            >
            > Campaign: Sydney+Workshop
            > Visits: 1
            > Pages per visit: 6
            >
            > When also displaying 'source', the following displays:
            > Campaign: Adelaide+Workshop
            > Source: short-letter-adelaide
            > Visits: 2
            > Pages per visit: 3
            >
            > Campaign: Adelaide+Workshop
            > Source: long-letter-adelaide
            > Visits: 1
            > Pages per visit: 1
            >
            > Campaign: Melbourne+Workshop
            > Source: short-letter-melbourne
            > Visits: 0
            > Pages per visit: 0
            >
            > Campaign: Melbourne+Workshop
            > Source: long-letter-melbourne
            > Visits: 1
            > Pages per visit: 3
            >
            > Campaign: Sydney+Workshop
            > Source: short-letter-sydney
            > Visits: 0
            > Pages per visit: 0
            >
            > Campaign: Sydney+Workshop
            > Source: long-letter-sydney
            > Visits: 1
            > Pages per visit: 3
            >
            > - Which reports are you looking at? It's strange that Adelaide, Sydney and Melbourne long-letter workshops give 2 different Pages per Visit numbers.
            > - How soon after your tests did you check the reports?
            > The stats above are about 48 hours old now.
            >
            > Regards
            > David
            >
            >
            > --- In webanalytics@yahoogroups.com, "Yuhui" <yuhuibc@> wrote:
            > >
            > > Hi David,
            > >
            > > You mentioned that you tested using 3 different browsers at 6 times of the day. Could you provide more information about:
            > >
            > > - Did you test the URLs consecutively at each time of the day? E.g. Machine 1 Browser A: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
            > > Machine 1 Browser B: 12:01pm Adelaide short-letter, 12:03pm Melbourne short-letter, 12:05pm Sydney short-letter, etc.
            > > etc.
            > > - Which URLs did you test with each browser?
            > > - When were the 3 times of the day? I assume that it's morning/afternoon/night, i.e. a break of more than 30 minutes between each time.
            > > - You didn't clear any of your browsers' cookies before/after each test, right?
            > >
            > > Also:
            > > - Can you re-check the reports that you shared with us? In your message, you listed 6 "long-letter" workshops for the 3 cities, but you only have 3 "long-letter" and 3 "short-letter" workshops.
            > > - Which reports are you looking at? It's strange that Adelaide, Sydney and Melbourne long-letter workshops give 2 different Pages per Visit numbers.
            > > - How soon after your tests did you check the reports?
            > >
            > > Having this information will help us clarify our thinking and better help you.
            > >
            > > Regards,
            > > Yu Hui
            > > Twitter: @yuhui
            > >
            > > --- In webanalytics@yahoogroups.com, "davidehmer" <davidehmer@> wrote:
            > > >
            > > > Just wondering if anyone can explain the results Google Analytics provides for campaign tracking.
            > > >
            > > > I have just tested redirects (see below for example) to record several parameters from users who will key in several vanity urls (offline promotion)
            > > >
            > > > REDIRECTS (6 in all):
            > > > redirect 301 /workshops/adelaide-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop
            > > >
            > > > redirect 301 /workshops/melbourne-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop
            > > >
            > > > redirect 301 /workshops/sydney-workshop http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=short-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop
            > > >
            > > > redirect 301 /workshops/adelaide http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-adelaide&utm_medium=post&utm_campaign=Adelaide%2BWorkshop
            > > >
            > > > redirect 301 /workshops/melbourne http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-melbourne&utm_medium=post&utm_campaign=Melbourne%2BWorkshop
            > > >
            > > > redirect 301 /workshops/sydney http://www.literacysolutions.com.au/workshops/workshop-registration.php?utm_source=long-letter-sydney&utm_medium=post&utm_campaign=Sydney%2BWorkshop
            > > >
            > > > The redirects work as expected. In testing I have used 2 computers and 3 different browsers. Cookies and Javascript are on in all browsers. I visited each or the 6 urls at 3 separate times of day.
            > > >
            > > > When checking results in GA's Traffic sources > Campaigns, I would have expected to see 3 visits for each source (18 in total) with 1 page per visit.
            > > >
            > > > However, in Campaign overview the following results display:
            > > > Campaign Source Visits Pages per visit
            > > > Adelaide+Workshop long-letter-adelaide 1 4
            > > > Melbourne+Workshop long-letter-melbourne 1 6
            > > > Sydney+Workshop long-letter-sydney 1 6
            > > >
            > > > When also displaying 'source', the following displays:
            > > > Campaign Source Visits Pages per visit
            > > > Adelaide+Workshop long-letter-adelaide 1 1
            > > > Melbourne+Workshop long-letter-melbourne 1 3
            > > > Sydney+Workshop long-letter-sydney 1 3
            > > > Adelaide+Workshop long-letter-adelaide 0 0
            > > > Melbourne+Workshop long-letter-melbourne 0 0
            > > > Sydney+Workshop long-letter-sydney 0 0
            > > >
            > > > Could it be that all activity for each time of time of day was treated as one session and only recorded it as one visit. (hence 3 visits rather than 18) And that it recorded the other activity as subsequent page accesses on those visits.
            > > >
            > > > If so this makes my testing misleading but in real world practice it should work fine?
            > > >
            > > > Any thoughts about what is happening here would be most welcome.
            > > >
            > > > Thanks
            > > > David
            > > >
            > >
            >
          Your message has been successfully submitted and would be delivered to recipients shortly.