One of the pri­mary rea­sons com­pa­nies uti­lize Omni­ture Site­Cat­a­lyst is to track the effec­tive­ness of online mar­ket­ing cam­paigns (see step three in Avinash’s “Nir­vana” post).  While this is an enor­mous topic, I will use this post to cover the most impor­tant things you need to know to effec­tively track cam­paigns in SiteCatalyst.

What is a Cam­paign?
I usu­ally define a mar­ket­ing cam­paign as any instance where you are delib­er­ately pay­ing money or expend­ing effort to drive traf­fic to your web­site.  I use this def­i­n­i­tion because in order to track the suc­cess of mar­ket­ing cam­paigns you need to have a way to iden­tify the spe­cific traf­fic you have gen­er­ated.  Site­Cat­a­lyst does this through a “Track­ing Code” that is cap­tured when vis­i­tors arrive to your web­site from one of your mar­ket­ing cam­paigns.  While most mar­ket­ing cam­paigns involve costs to gen­er­ate traf­fic, there are some cases, such as inter­nal e-mail lists, that have min­i­mal costs, but are still valu­able cam­paign con­trib­u­tors.  Most clients have a team of peo­ple who are focused on creating/managing mar­ket­ing cam­paigns and deter­min­ing which types (Paid Search, Paid Dis­play, E-mail, etc…) are the best and which cam­paign ele­ments within each cam­paign type pro­duce the best results.  By using Omni­ture Site­Cat­a­lyst to iden­tify the most cost-effective cam­paign ele­ments it is pos­si­ble to squeeze the most value from your lim­ited mar­ket­ing budget.

How Do Cam­paigns Work in Omni­ture Site­Cat­a­lyst?
To use the Cam­paign fea­ture of Site­Cat­a­lyst, the first step is to be sure your Site­Cat­a­lyst imple­men­ta­tion is set-up to cap­ture cam­paign track­ing codes when vis­i­tors arrive to your site.  This is nor­mally done by using a JavaScript plug-in that Omni­ture pro­vides to cap­ture a URL para­me­ter and place it in the cam­paign vari­able (s.campaigns).  For exam­ple, if you buy the key­word “books” on Yahoo, when the user clicks on that Paid Search link, they might come to your site with the URL of “www.mysite.com?id=12345.”  In this case, the code “12345” would be placed in the Site­Cat­a­lyst cam­paigns vari­able for track­ing pur­poses.  In many ways, under­stand­ing Cam­paigns is really a com­pi­la­tion of many of the things we have learned so far in pre­vi­ous blog posts.  The Site­Cat­a­lyst cam­paigns vari­able is sim­ply a Con­ver­sion Vari­able which has full sub­re­la­tions enabled which will attribute any sub­se­quent Suc­cess Events to the cam­paign track­ing code stored in the cur­rent user’s cookie.  As with any other Con­ver­sion Vari­able, you can use the admin con­sole to deter­mine how long you want to keep the track­ing code in the user’s cookie and if you want to give credit to the first cam­paign track­ing code or the last one.

Once your Site­Cat­a­lyst JavaScript file is set-up to cap­ture cam­paign track­ing codes, the next step is to assign track­ing codes to all links that will refer traf­fic to your site.  This is the part of the cam­paign process in which Omni­ture Site­Cat­a­lyst clients make the most mis­takes.  The key to assign­ing track­ing codes to cam­paign ele­ments is that they have to be unique.  As long as the same cam­paign track­ing code is not asso­ci­ated to more than one cam­paign ele­ment you are in good shape.  Using the exam­ple above, the code “12345” is now ded­i­cated to the key­word “books” on Yahoo as a Paid Search ele­ment.  If the com­pany chooses to buy the same key­word on Google, it should use a dif­fer­ent track­ing code so Site­Cat­a­lyst can dif­fer­en­ti­ate between the same key­word on Google and Yahoo.  Unfor­tu­nately, I have seen many clients set the same track­ing code for “books” in mul­ti­ple places and then won­der why it is so dif­fi­cult to show how the key­word per­formed dif­fer­ently on each site.  Along these same lines, you will have many cam­paign types such as Paid Search, Paid Dis­play, E-mail, etc…  While each of these are dif­fer­ent types of cam­paigns, they all need to have track­ing codes and the track­ing codes need to be unique amongst the entire spec­trum.  In this sce­nario, no Paid Dis­play or E-mail link can have the track­ing code “12345” or trou­ble will ensue.

Mak­ing Sense of Cam­paign Track­ing Codes
So now let’s say that you are all set to cap­ture cam­paign codes and have worked with your Paid Search, Paid Dis­play and E-mail ven­dors to add a track­ing code to each des­ti­na­tion link for which you are spend­ing mar­ket­ing dol­lars.  As traf­fic begins to trickle in, you can open up the Track­ing Code report found in the Cam­paigns area and see Click-throughs and also what web­site Suc­cess Events have taken place after vis­i­tors arrived from each cam­paign code:

This sam­ple report shows which cam­paign track­ing codes are get­ting the most clicks and lead­ing to the most Appli­ca­tion Com­ple­tions.  But mak­ing sense of this report is not very easy since you prob­a­bly don’t want to mem­o­rize every cam­paign track­ing code.  If only there were a way in Site­Cat­a­lyst to add meta-data to a Con­ver­sion Vari­able so you could group items in dif­fer­ent ways (hint, hint!).  Those of you who have read the pre­vi­ous posts, of course, know that Clas­si­fi­ca­tions pro­vide this capa­bil­ity so the next log­i­cal step in cam­paign man­age­ment is to iden­tify the ways you want to slice and dice your cam­paign track­ing codes.  For exam­ple, you might want to group cam­paign track­ing codes by:

  1. Cam­paign Name (Spring 2008 Cam­paign, Sum­mer Brand­ing Cam­paign, etc…)
  2. Channel/Type (Paid Search, Paid Dis­play, E-mail, etc…)
  3. Vendor/Website (Google, Yahoo, CNN​.com, etc…)
  4. Link Type (Text, Ani­mated GIF, Flash, etc…)
  5. Busi­ness Unit (Prod­uct Mar­ket­ing, Acqui­si­tion, Cus­tomer Ser­vice, etc…)
  6. Busi­ness Owner (Joe Mur­phy, Sue Smith, etc…)

As we learned in the Clas­si­fi­ca­tions post, you can choose as many attrib­utes as you want and the best part is that Clas­si­fi­ca­tions are retroac­tive, which means you can decide how you want to group cam­paign track­ing codes days, weeks or months after the cam­paign launches and change it as many times as you’d like.  Doing this might pro­duce a report that looks like this:

Impor­tant Things to Know About Cam­paigns
The fol­low­ing are some impor­tant things to know about track­ing campaigns:

  1. The cam­paigns vari­able (track­ing code) comes with full sub­re­la­tions by default so you can break every Con­ver­sion Vari­able down by any track­ing code and any Clas­si­fi­ca­tion of any track­ing code.
  2. If you do a lot of Paid Search, cre­at­ing thou­sands of unique track­ing codes and inputting them into the var­i­ous search engines can be tedious and costly.  Omniture’s Search­Cen­ter prod­uct auto­mates this process and pro­vides many other incred­i­ble ben­e­fits so check it out.
  3. The “Click-throughs” met­ric is only avail­able in Cam­paign reports, though there are more advanced ways to imple­ment this if more flex­i­bil­ity is needed.
  4. You can upload the costs asso­ci­ated with each cam­paign track­ing code to cal­cu­late the Cost per Click or Cost per [Suc­cess Event].  This is a bit more advanced, but if you are inter­ested in this Omni­ture Con­sult­ing would be happy to assist you.

Real-World Exam­ple
Some­one “Twit­ter­ing” with me asked if I could show an exam­ple of how Greco Inc. (our fic­ti­tious com­pany) could mon­e­tize the use social net­work­ing so I am going to do my best in this week’s real-world exam­ple.  In this sce­nario, we’ll imag­ine that the CoolFlow­ers sub­sidiary of Greco Inc. has hired a young “go-getter” out of col­lege who is a whiz at social net­work­ing.  This indi­vid­ual has done research and iden­ti­fied sev­eral blogs on the Inter­net that have active dis­cus­sions about flow­ers and has also iden­ti­fied an inter­nal employee who knows more about flow­ers than any­one on the planet!  Soon, he has his asso­ciate blog­ging on CoolFlow​ers​.com and com­ment­ing on a few key blogs plac­ing links back to her CoolFlow­ers blog.  As described above, a cam­paign track­ing code iden­ti­fier is added to each link going back to the CoolFlow­ers site such as http://www.coolflower.com?id=blogcomment.  Now, as blog read­ers click on the link in her com­ments, they are routed to the CoolFlow­ers web­site where Click-throughs and Pur­chases can be tied to the social net­work­ing ini­tia­tive.  Obvi­ously, this rep­re­sents a rudi­men­tary approach, since it lumps all blog com­ments into one track­ing code, but if it wanted to, Greco Inc. could use “id=blogcomment1,” “id-blogcomment2,” etc… to tie suc­cess to a spe­cific com­ment.  Either way, all of the social net­work­ing track­ing codes can later be grouped together using SAINT Clas­si­fi­ca­tions and com­pared to other mar­ket­ing channels.

This same con­cept can be applied to pro­grams like Twit­ter in which you can post links to your site with a cam­paign iden­ti­fier and then use tools like POPrl to shorten the URL but still record the click as a cam­paign.  To see how this works, feel free to click on the fol­low­ing link http://​poprl​.com/​0cG which will take you to the Omni­ture Con­sult­ing page on omni​ture​.com and indi­cate that you got there because of my blog!

 

Have a ques­tion about any­thing related to Omni­ture Site­Cat­a­lyst?  Is there some­thing on your web­site that you would like to report on, but don’t know how?  Do you have any tips or best prac­tices you want to share?  If so, please leave a com­ment here or send me an e-mail at insidesitecatalyst@​omniture.​com and I will do my best to answer it right here on the blog so every­one can learn! (Don’t worry — I won’t use your name or com­pany name!).  If you are on Twit­ter, you can fol­low me at http://​twit​ter​.com/​O​m​n​i​_​man.

Learn more about Omni­ture Consulting
Learn more about Omni­ture University

 

16 comments
eappleby
eappleby

The images in this post no longer display.  Can you fix that?

Mik
Mik

The drawback of online marketing is anonymity but these web analytic tools is a very useful way to track how effective your online marketing tool is.

kelly P
kelly P

the way our site URLs are set up, we cannot have a? at the end of the URL; we are instead using an ampersand (&id=) for the sitecatalyst tracking code. Will the URLs still track?

Phil
Phil

We have added our external campaign variable 'ecp' to our tweets when and add them in to our bit.ly account prior to posted the links. However there remains a large discrepcancy as to what bit.ly reports and SiteCatalyst reports. Any ideas about this? Does using the getValOnce Plugin on our campaigns create this disparity?

Jim Hazen
Jim Hazen

Ann - do something like this in the s_code.js if your query string param for example is 'cid=' s.campaign=s.getQueryParam('cid'); s.eVar3=s.campaign;

Ann
Ann

We are using Omniture code to track the visits on our pages. i have a campaign code in a url and I need to track that and populate it on one of the eVar variables(eVar3 for example). can you please advice.

Filip
Filip

Hi Adam, very useful post, thank you! I was wondering: you write about SearchCenter, but is there any other way to track search campaigns in detail? Could I, for example, set the tracking code in AdWords to be say ?id=googCPC-853-{keyword} (where keyword is the macro that puts the triggered keyword into the destination URL)? Does that make sense? Thanks, Fillip

Jon
Jon

Great Article Adam. I work for a digital agency, we recently ran a banner campaign on Google via placement/site targeting for a client that uses Omniture SiteCatalyst to determine incoming or referral URLS. Unfortunately we didn't include tracking codes that you suggest to measure success. Can we confirm the incoming URLS that lead users to the clients site via Omniture or will they all appear as if from Google Sites? Would sites that had clicks thrus appear on your logs even if the traffic they drove is small or if the site itself isn't a well known one? Do sites need to be registered somewhere for you to detect them? Sorry for all the questions!! Thanks in advance, ~J

Adam Greco
Adam Greco

Tim - I agree and thanks for pointing that out. I will probably talk about Internal promotions in a future post. Thanks!

Tim Elleston
Tim Elleston

Great post Adam. And you can then take this one step further by using "internal" campaign tracking ID's (i.e. pid=123) to see which internal promotions drive conversions against the original blog campaigns. This way, you'll be able to measure different creative for internal campaigns against external campaigns. I think you need full sub-relations turned on for the internal campaign ID's though. But the correlation report (5x5) created is a real insight into the full user journey and conversion. Tim

Adam Greco
Adam Greco

Beth - Yes. If you do not set up the Paid Search Detection parameter in the Admin Console, then all Search Engine traffic is lumped together into Search Engine and Search Keyword reports. Once you enable Paid Search Detection, data is split into Paid Search reports and Natural Search reports. ClientCare can walk you through setting this up…

Beth
Beth

Hi Adam, Nice post and very well explained. Have one question, if per se campaigns are not set-up in Omniture to track, is campaign visitors bundled with natural search? Thanks, Beth

palani
palani

Hi Adam, Nice Article, Explains very clearly the way things have to be setup. In my past experience, even i have come across scenario's where Unique Campaign Tracking Codes are not generated and that skews the numbers. Thanks

Adam Greco
Adam Greco

Avinash - Thanks! Melissa - I am not aware of any formal benchmarks, but I tend to be concerned if the discrepancy is more that 10%-15%. There are lots of ways to do some auditing using Data Warehouse and "debug" logs if you do suspect an issue.

Melissa
Melissa

Adam - do you have a "normal" discrepancy amount when it comes to comparing vendor click-throughs for emails or banner ads, etc. to what SiteCatalyst registers? This is an issue we come across time and time again when a client runs a banner or sends an email out, and the numbers don't match. I know that the numbers will never match, but what's an acceptable percentage of discrepancy?

Avinash Kaushik
Avinash Kaushik

Adam: You 've made a very complex topic significantly more digestible! And added a touch of social media to boot. : ) Bravo! -Avinash.