Custom Report on Traffic Origin

I just completed a report which might interest many marketers but isn’t available by default in WebTrends. It is a report showing distribution of traffic origin among Organic, Paid Search and Paid Referral (Marketing Campaigns):

trafficorigin

Here is how to create it:

  1. Create a dimension called traffic origin
  2. Base it on Entry request
  3. Click Advanced
  4. Override Default String to “Organic”
  5. Apply regular expression “.*(WT\.mc_id|WT\.srch)=.*”
  6. Translate with a translation file containing
    SOURCE,LABEL
    WT.srch,Paid Search
    WT.mc_id,Paid Referrer
  7. Save
  8. Create a custom report
  9. Use the dimension you defined beforehand
  10. Add any visit based measure (Visits, Pages/Visit, Orders, Product Views, etc.)
  11. That’s it :)

Remarks:

  1. The regular expression extracts either “WT.mc_id” (if present) or “WT.srch” (if present) or <empty string> (which then is set to “Organic” by the overridden Default String.
  2. The Lookup Table then translates “WT.srch” to “Paid Search” and “WT.mc_id” to “Paid Referral”.
  3. As the custom dimension is itself based on a standard visit based dimension, all visit based measures (Visit Duration, Visit Page Views, etc.) are available in the Measures tab.
  4. This assumes that incoming links (first hit in a visit) never contain WT.srch and WT.mc_id paremeters at the same time.
  5. You shouldn’t, unless you know what you are doing, use visitor based measures with this report.

Any remarks or questions are welcome! :)

VN:F [1.3.2_665]
Rating: 5.0/5 (5 votes cast)

2 Responses to “Custom Report on Traffic Origin”

  1. admin says:

    Hi

    The report will still work, but only one of both parameters will be considered.

    I tested your example, and the result was WT.mc_id.

    I tried to play with parameter order in the URL and the regexp, as well as removing leading and trailing “.*” and the results seem quite random.

    Looking a bit closer, I suspect that the choice of the extracted portion of the URL depends on the regexp’s greedyness, meaning that the “biggest match” (even if not all of it is finally extracted) will determine the result.

    Hope this helps!

    VN:F [1.3.2_665]
    Rating: 0.0/5 (0 votes cast)
  2. OG says:

    Hi,

    what if we have a SEM campaign using both and still we wanted to use this report?

    assume that we have ?WT.srch=1&WT.mc_id=aaa

    Any thoughts?

    thx
    OG

    VA:F [1.3.2_665]
    Rating: 0.0/5 (0 votes cast)

Leave a Reply


SiteSensor Browser Pedia Collector