How to interpret referring domains broken down under a tracking ID?

[cunnyong_goh]

For your information, the tracking ID is tagged with an url posted on facebook and twitter. Using that report, I understand that there are about 26 thousands unique visitors who have clicked that URL that was tagged with that tracking ID. By breaking down for referring domains, I see that about 14.9k of unique visitors came from facebook while about 9k of unique visitors came by typing/bookmarking the url of my website. There are also many other random referring domains associated with this tracking ID.
How could that happen if I put that URL with a tracking ID only on facebook and twitter? How to explain that?
 

Does referring domain mean an immediate site that refer visitors to my website? Or it means any site prior to landing on my website?
 

Is it correct that by breaking down a tracking ID by referring domain, what I see on the report is what website led visitors to my website via the URL that is tagged by a tracking ID.
By that, I mean, if I see website A as a referring domain, I can say that someone clicked my URL that tagged with a TRACKING ID on website A and landed on my site.
 

Another question is, since the URL tagged with that tracking ID was posted on facebook and twitter only. How to interpret the referring domains (Type/Bookmark)?

a) Does that mean they came to my website by typing/bookmarking and later on, they went to facebook/twitter and happened to click that link (with tracking ID)
b) They came to our site by opening the URL tagged with that tracking ID by right clicking and opening the URL in a new tab? (I tried that, and that action didn't pass any value to referring domain variable.) 
c) The URL was shared to a visitor by their friend through another medium such as facebook messenger or other messenger. Then, the visitor open the link by copying and pasting on a browser OR clicking the link straightaway.(I tried that, and that action didn't pass any value to referring domain variable.) 
d) None of the above

4 questions above. Thank you for reading this lengthy thread. Please the find the attachment to understand better.Thanks!

I did read all the  following definitions from Analytics Help Home but don't quite know how to interpret the report I attach with this thread. 

Referrers

Displays the domain or URL where your visitors came from before they arrived at your site, the methods visitors use to find your website, and the number of visits to your site that came from these referring locations.

Referring Domains

Shows the domains that referred the customers that most impacted your site’s success metrics. Referrers fall into two main categories: Domains and URLs. Domains refer to the domain name, and appear as the base domain without the query string or subdirectories attached. URLs include the base domain name, as well as any query strings or subdirectories.

Referrer report: Displays all previous URLs linked to your site that do not match internal URL filters.

Referring Domain report: Similar to the Referrer report, it combines all referring URLs to just their root domains.

http://microsite.omniture.com/t2/help/en_US/reference/index.html#kb-analytics-comparing-referrer-ref...

[AkhilSharma]

Hi, 

It was very well written and detailed question. Thanks for that as it really helps us to understand the problem. Please find my reply below:

Ques 1:How could that happen if I put that URL with a tracking ID only on facebook and twitter? How to explain that?

Answer:Other websites might be figuring as to someone might have pasted the link on that website itself.In referrer domains report, in order to qualify as a referrer the webpage needs to have a link to your site and the link must be clicked. Since most of the websites are for best promotions and offers, so must be that they had pasted your link in their website.

Ques2:Does referring domain mean an immediate site that refer visitors to my website? Or it means any site prior to landing on my website?

Answer: Referring Domain Report:A referrer is the Web page location your visitors were viewing before coming to your site. In order to qualify as a
referrer, that web page also needs to have a link to your site, and the link must be clicked. Use of the Back or
Forward buttons does not create a "referral", nor does the use of the History or Favorites (Bookmarks). The Referring
Domains Report shows you the domains your visitors came from before they were on your site. The Referrers report
is similar but shows you the complete URLs instead of only the domains.
The Referring Domains Report provides an overview, such as how many referrers are coming from "abcsite.com"
versus "xyzsite.com." The Referrers report provides more detailed information about the unique referring URLs for
each of these sites.

Ques 3: Another question is, since the URL tagged with that tracking ID was posted on facebook and twitter only. How to interpret the referring domains (Type/Bookmark)?

Answer:“Typed/Bookmarked line items occur in reporting where a referrer for an image request is not present.” So in other words, if SiteCatalyst does not see a referrer value, then it simply can not tell you where that visitor came from, so they get dropped into the Typed/Bookmarked bucket.

It might be  two case scenarios:
1) People have returned to your website directly by bookmark or typing it after going from the facebook/twitter link.
Check for  the total unique visitor across unique visitor coloumn (30841) versus the total which is de-duplicated on the top (26,635).

2)people have copy pasted the link from the facebook/twitter directly onto the browser without clicking it which resulted in the typed/bookmarked section.

Ques 4: Is it correct that by breaking down a tracking ID by referring domain, what I see on the report is what website led visitors to my website via the URL that is tagged by a tracking ID.
By that, I mean, if I see website A as a referring domain, I can say that someone clicked my URL that tagged with a TRACKING ID on website A and landed on my site.

Answer: Yes 

Suggestion: Also do check the orginal referrer report which will complete the picture as to who was the original referring domain for the visit.

Let me know if this helps.

[cunnyong_goh]

First of all, thanks for replying. That is very helpful but I need a little more clarification. 

Ques 4: Is it correct that by breaking down a tracking ID by referring domain, what I see on the report is what website led visitors to my website via the URL that is tagged by a tracking ID.
By that, I mean, if I see website A as a referring domain, I can say that someone clicked my URL that tagged with a TRACKING ID on website A and landed on my site.

Answer: Yes 

Suggestion: Also do check the orginal referrer report which will complete the picture as to who was the original referring domain for the visit.

Regarding this the above suggestion, I am worry that the report would show the very first referring domain, prior to clicking the URL tagging with a TRACKING ID, that led visitors to my website. How to filter that out? 
Situation 1: A visitor came to my website from abc.com but on the same day, he went to facebook and happened to click the link that has a tracking ID and landed on my website again. Would A's original referring domain be abc.com while referring domain be facebook(by break down tracking ID report)?
Situation 2: A visitor came to my website from abc.com on Day 1. On day 2, he went to facebook and happened to click the link that has a tracking ID and landed on my website again. Would A's original referring domain be abc.com while referring domain be facebook(if I break down tracking ID report and set the time period to day 2)? what about if the time period is day 1 and 2 while other variables stay the same?

What I am trying to do here is to figure out how many percent of the visitors, who click the URL with a tracking ID, came from twitter, facebook or other sources. Thus, should I use original referring domain / referring domain? 

For your information, using google chrome, I tried to check if a value is to passed referrer variable on twitter and facebook and found out the following:
On twitter, if I opened the shortened twitter link in a new tab, the referrer variable would receive a value. However, by selecting go-to the shortened twitter link (from the right click drop down menu), the referrer variable would not receive any value. Besides that, by copying and pasting a shortened twitter link on a browser URL bar would not pass any value to referrer variable as well.
On facebook, if I did the above steps, none of them will pass a value to the referrer variable. 
I suppose, all the above will be categorized as Type/Bookmark then.

[AkhilSharma]

In that case you can try build the landing page as a segment and add the tracking code in the segment as a rule and check for the reports within the strict dates when the page became live.
the segmentation takes care of the other pages as it will filter out only the visits coming from the tracking id and landing pages directly.

So build rule for visit when landing page was your landing page and tracking id was <your tracking id>.

Or else there is no way to differentiate between people who had opened the url in a new tab either from facebook or twitter or just done the copy paste and reached the marketing page.

[cunnyong_goh]

Alright, I did that segment by setting entry page as my landing page and tracking ID as my tracking ID and went to original referring domains report and referring domains report. Comparing data from both reports for the same time frame, I expected to see same number. However, that did not happen. 
So, that led me to think that even by setting the same time frame, the original referring domain will still look for the very first referrer that brought the visitors, those are in my segment, to my website. Is that right?
What I meant was, if a visitor came to my site through ABC.com on Sept 1st. On Sept 3rd, he came back to my site by clicking a link that I shared on facebook. 
By segmenting using rule(entry page as the landing page and tracking ID as my tracking ID) and day as Sept 3rd only, I would see that visitor's referring domain as facebook while his/her original referring domain as ABC.com?
[First Question above]

That is quite similar to what i asked in my last reply. Hopefully, you could answer this time. Thanks. 

On the other hand, why cant I do it by breaking down a tracking report by referring domain/original referring domain? Wouldn't that limit the visit to that who has clicked that URL that was tagged with a tracking ID?
[Second Question above]

Thanks! 

[AkhilSharma · Accepted solution]

What I meant was, if a visitor came to my site through ABC.com on Sept 1st. On Sept 3rd, he came back to my site by clicking a link that I shared on facebook. 
By segmenting using rule(entry page as the landing page and tracking ID as my tracking ID) and day as Sept 3rd only, I would see that visitor's referring domain as facebook while his/her original referring domain as ABC.com?
[First Question above]
Answer: Yes your are right.Even if your page did went live on 3rd september, it will show the original referring domain as ABC.com and referring domain as facebook.
For new  visitors coming for the first time from facebook both referring domain and original referring domain will be  same.

On the other hand, why cant I do it by breaking down a tracking report by referring domain/original referring domain? Wouldn't that limit the visit to that who has clicked that URL that was tagged with a tracking ID?
[Second Question above]
Answer: Nopes cross pollination of your visit (as shown in  example above) will create difference in your report. So even if your filter it by your tracking id, original referring domain will show results for two types of visits:
1) first time visit by facebook, twitter or other referral medium
2) repeat visitors reaching earlier through lets say ABC.com and then by Facebook, twitter or other referring domains.

However, the purpose of the exercise will again be defeated as to identify strictly referral from facebook, twitter or other website.
For the possible workaround from an analyst point of view we have to assume certain assumptions and then did our analysis this we can support by facebook analytics and link shortener analytics as well.

Also you can try the following steps:
In the original referring domain report(already filtered by the required tracking id, same segmentation) break down by traffic sources>referring domains.This will help you to link the original domain and referrer domain report together.
Also this will help you understand two section  of your visitor( you can try this additional rule in your segmentation):
1) new visitor medium breakup to reach your landing page.(visit number equals 1)
2) repeat visitor medium breakup to reach your landing page.(visit number greater than 1)

Let me know if that helps.