Falling into the bucket of Site­Cat­a­lyst fea­tures that many clients are not famil­iar with, Par­tic­i­pa­tion is a handy fea­ture that can help you deter­mine which of your site con­tent ele­ments are lead­ing to suc­cess and which are not.  In this post, I will explain Par­tic­i­pa­tion and then illus­trate some advanced uses of it that even those famil­iar with it may not have tried.

What is Par­tic­i­pa­tion?
In order to fully under­stand Par­tic­i­pa­tion, I find that it is help­ful to first under­stand Allo­ca­tion (also known as Lin­ear Allo­ca­tion).  If a vis­i­tor to your web­site clicks on ten pages and then com­pletes a Suc­cess Event, Site­Cat­a­lyst “allo­cates” or divides the Suc­cess Event into equal parts and gives each pre­ced­ing page credit.  If the Suc­cess Event is a pur­chase of $100, all ten pages viewed prior to the Suc­cess Event and the page on which the Suc­cess Event takes place would be given credit of $10.  If the Suc­cess Event is a Counter Event, then each page would receive .10 credit.  These val­ues are aggre­gated, rounded and dis­played in the Pages report if you add a Suc­cess Event met­ric as shown here:

Most peo­ple never notice Allo­ca­tion because when they look at the Pages report they are look­ing at Page Views, Vis­its or Unique Vis­i­tors.  You can add mul­ti­ple Suc­cess Events to the Pages report and see the dif­fer­ent page allo­ca­tions for each Suc­cess Event met­ric (note: this func­tion­al­ity is unique to the Pages report).  While Allo­ca­tion met­rics can be use­ful, they have an obvi­ous down­side.  The more pages vis­i­tors view in a visit, the more diluted Allo­ca­tion is for each page.  For exam­ple, if a sec­ond vis­i­tor views 20 pages and then com­pletes the same Suc­cess Event, each page would get only .05 credit.  There­fore, Allo­ca­tion tends to reward vis­its with smaller num­ber of clicks and pun­ish vis­its with many clicks.  Depend­ing upon your busi­ness model and/or site objec­tives, this can be a pos­i­tive or a negative.

Hav­ing cov­ered Allo­ca­tion, let’s now get back on topic and define Par­tic­i­pa­tion.  Par­tic­i­pa­tion is a fea­ture that assigns equal credit to each item that “par­tic­i­pates” in the flow lead­ing to a Suc­cess Event.  So in the exam­ples above, whereas Allo­ca­tion would divide the $100 Suc­cess Event, Par­tic­i­pa­tion would assign full credit of $100 to every page the vis­i­tor viewed up until the point that the Suc­cess Event takes place.  When cus­tomers first hear about this, they are taken aback since the Par­tic­i­pa­tion val­ues are inflated and report totals will not match actual Rev­enue (or equiv­a­lent for a Counter Suc­cess Event).  The key to Par­tic­i­pa­tion is to focus the Par­tic­i­pa­tion val­ues of var­i­ous items in the report to see which pro­duces the most suc­cess.  Since this can be a bit tricky, I will go through another example:

In the above report, we can see a few pages of the Omni­ture Blog, the asso­ci­ated Page Views and Blog Post View Par­tic­i­pa­tion.  In this case, Omni­ture sets a Suc­cess Event each time a vis­i­tor views a Blog Post and through this report you can see how often each page leads to Blog Post Views.  For exam­ple, the “Adam Greco” page had 30 Page Views, but only was viewed in the flow lead­ing to 27 Blog Post Views (my awful head shot must have scared some folks off!).  In other words, the Par­tic­i­pa­tion met­ric is indi­cat­ing that there were 27 Blog Post Views, in the time frame of this report, in which the vis­i­tor viewed Adam Greco’s page prior to open­ing a Blog Post.

Finally, I like to take this a step fur­ther by cre­at­ing a cal­cu­lated met­ric that divides a Par­tic­i­pa­tion met­ric by Page Views.  When doing this, the higher the result­ing num­ber, the more that item (page in this case) gen­er­ates Suc­cess Events for every Page View it gets (effi­ciency).  I use this to do a “hidden-gem” analy­sis where you are look­ing for pages that may not get the most traf­fic, but when they do, con­vert at higher than aver­age rates.  In the exam­ple we have been using, you could call this cal­cu­lated met­ric “Blog Pull-Through” and add it to the pages report as shown here:

Of the pages shown in this report excerpt, the “ASI” post had the most “Blog Pull-Through” in that it was able to get 30 Blog Post Views out of only 8 Page Views or 3.75 Blog Post Views/Page View.

Impor­tant Things To Know About Par­tic­i­pa­tion
The fol­low­ing are some impor­tant things to know about Participation:

  1. Par­tic­i­pa­tion can be turned on for most Site­Cat­a­lyst Suc­cess Events (exclud­ing some forms of Data Sources).  To enable Par­tic­i­pa­tion for a Suc­cess Event con­tact your Account Man­ager or Client­Care.  Once Par­tic­i­pa­tion is turned on for a Suc­cess Event, you add Par­tic­i­pa­tion met­rics to traf­fic reports by using the drop-down box in the “Add Met­rics” dia­log box
  2. Par­tic­i­pa­tion is cal­cu­lated after the Suc­cess Event takes place, but affects pages viewed prior to the Suc­cess Event
  3. Par­tic­i­pa­tion is cal­cu­lated on a per Visit basis
  4. In Site­Cat­a­lyst, Par­tic­i­pa­tion Met­rics are not retroac­tive so they only show data from the point at which they are turned on
  5. Par­tic­i­pa­tion met­rics are avail­able for most Suc­cess Events by default in Omni­ture Discover
  6. Par­tic­i­pa­tion met­rics are pri­mar­ily used with Traf­fic Vari­ables (sProps), not Con­ver­sion Vari­ables (eVars)
  7. By default, Par­tic­i­pa­tion met­rics are tied to the s.pagename (Pages) vari­able, but there are more advanced ways to use Par­tic­i­pa­tion by tying Par­tic­i­pa­tion met­rics to cus­tom sProps (see exam­ple below)

Real-World Exam­ple
In this week’s real-world exam­ple, I will attempt to explain a more advanced ver­sion of Par­tic­i­pa­tion (not for the faint of heart!).  In this exam­ple, a man­u­fac­tur­ing sub­sidiary of Greco Inc. has a web­site con­tain­ing thou­sands of prod­ucts and the Prod­uct team wants to under­stand which prod­ucts site vis­i­tors are look­ing at and, more impor­tantly, which prod­ucts get vis­i­tors to look at the most other prod­ucts.  The first ques­tion is answered pretty eas­ily by set­ting a cus­tom “Prod­uct View” Suc­cess Event and cap­tur­ing the prod­uct ID# in the Prod­ucts Vari­able.  The web ana­lyt­ics team’s first attempt to answer the ques­tion of which prod­ucts get vis­i­tors to look at other prod­ucts was to pass the prod­uct ID# to a cus­tom sProp and enable Pathing.  While this pro­vided some good infor­ma­tion, there were far too many prod­ucts to effec­tively use Pathing analy­sis.  The sec­ond attempt to answer the ques­tion was to enable Par­tic­i­pa­tion on the “Prod­uct View” event as described above, but the issue they ran into was that there were so many pages on the web­site that it was dif­fi­cult to ana­lyze the Par­tic­i­pa­tion met­rics for prod­uct amongst all of the other pages con­tained in the Pages report.

At this point, Greco Inc. turned to Omni­ture Con­sult­ing for guid­ance (shame­less plug!).  Their Omni­ture Con­sul­tant informed Greco Inc. that they could enable Par­tic­i­pa­tion on any sProp, not just s.pagename.  A nice thing about enabling Par­tic­i­pa­tion for the cus­tom sProp (Prod­uct sProp in this case) is that it removes all of the clut­ter sur­round­ing non-product pages which can be really help­ful.  Since they already had prod­uct ID#‘s in a cus­tom sProp, they could tie Par­tic­i­pa­tion to that sProp.  (Con­cen­trate here because this is about to get com­pli­cated!)  By doing this, each time a Prod­uct View took place, the Prod­uct View Suc­cess Event met­ric would be increased and the prod­uct ID# passed to the sProp on the Prod­uct View page would get one Prod­uct View Par­tic­i­pa­tion credit (for itself).  If, within the same visit, vis­i­tors viewed a sec­ond prod­uct, a Prod­uct View Suc­cess Event would take place for the sec­ond prod­uct and the sec­ond prod­uct would get one Prod­uct View Par­tic­i­pa­tion credit (for itself), but more impor­tantly, the first prod­uct they looked at would get a sec­ond Prod­uct View Par­tic­i­pa­tion credit since it was in the flow of another Prod­uct View Suc­cess Event (I told you it was going to be tricky!).  For those keep­ing score at home, here is how the count stands so far:

Just to make sure you get it, let’s go one step fur­ther and assume that the same vis­i­tor looks at one last prod­uct and exits.  At this point, the third prod­uct gets one Prod­uct View Suc­cess Event, one Prod­uct View Par­tic­i­pa­tion (for itself) and the first two prod­ucts get one more Prod­uct View Par­tic­i­pa­tion bring­ing the new totals to:

Once Greco Inc.‘s heads stop spin­ning, they imple­ment this accord­ing to the Omni­ture Consultant’s instruc­tions and by the next day they can see the fol­low­ing report by open­ing the “Prod­uct” sProp report and adding the fol­low­ing metrics:

 

Now Greco Inc. can look at prod­ucts by Page Views, Product View Par­tic­i­pa­tion (how many times each was in the flow of another Prod­uct View) or they can cre­ate a cal­cu­lated met­ric to divide Prod­uct View Par­tic­i­pa­tion by Page Views to see which prod­ucts have the most Prod­uct Views/Page View which can show you which prod­ucts get vis­i­tors to look at other prod­ucts.  Note that Par­tic­i­pa­tion will only count each ele­ment once so, in this case, if a user views the same Prod­uct more than once, it will get a sec­ond Page View, but no more Par­tic­i­pa­tion.  If you antic­i­pate this hap­pen­ing a lot, you can sub­sti­tute Vis­its for Page Views in the above Cal­cu­lated Met­ric.   All of this can be use­ful in answer­ing their ques­tion as to which prod­ucts get vis­i­tors to look at other products.

 

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

 

10 comments
Melvin
Melvin

Excellent post !!! I'm asking to enable this junction on my site from ClientCare. Keep it up !!!

Adam Greco
Adam Greco

James - Great point. It can get confusing between sProps and eVars. I guess the best short answer I can give is that when you want to look at how page-by-page actions impact success, I would use sProps and Participation since they are tied to sProps which collect on every page. When you want to see how a discrete value (i.e. City, Age, Site Tool) impacts success, I would use eVars...

matt
matt

Hey Adam, Excellent post, thanks for spelling this out. Page view participation is *extremely* helpful for us media companies who are always looking to see which pages or chunks of content drive the highest number of page views in aggregate. I was interested in your example above where you talk about taking [PV Participation]/[Page Views] as the "pull-through" metric. This works when you want to make sure that the first page gets credit for multiple product views in a single visit session (your example), but for us it's often the other way around. The business question we want to answer is: on average, how many page views are consumed after a given page is viewed, *regardless* of whether those follow-on page views came from the same page as the one you're measuring? Users are bound to zig-zag, so what we're really looking at is the *first* instance of a given page within the visit. If users click back to the original page, we don't want the pull-through metric to start counting a second set of particpation values toward the same page name. Would the [PV Participation]/[Visits] ratio give us that? Can you see any concerns with using that approach? Thanks again - these posts are a great help. matt

James Dutton
James Dutton

Adam, Useful post as always. One thing I think you might want to touch on, is how participation metrics (which are extremely useful for most of my clients) compare with using evars. With evars offering first touch, last touch and linear attribution and sprops offering allocation and participation; and many Omniture implementations recommending duplicate use of sprops and evars for the same measures there may be some community confusion around best practice for both forms. Keep up the good work, speaking on behalf of the community - we all enjoy your writing. Cheers, J twitter.com/slicecast

Adam Greco
Adam Greco

Rich - Doing examples of Participation for Purchases and Newsletter sign-ups would be too easy! I have been asked by many to provide more advanced things that aren't as obvious which is why I chose the examples I did. Unfortunately, with so many different audiences, you can't please all of the people all of the time! Thanks for the reply... Adam

Rich Page
Rich Page

Hi Adam, Thats a very interesting post on a somewhat confusing, but important topic. How come you didn't chose to do participation examples on something like purchases or newsletter signups? Seems like it would be a better real world example than page and product view participation. Keep up the great posts lately too - the Omniture blog has got much better lately! Rich Page

Adam Greco
Adam Greco

Rich - You are correct. Participation only affects activity after the success event and Participation is only credited for the current visit.

rich chew
rich chew

can you confirm that the participation to a success event is only credited if that page is viewed PRIOR to a success event? in other words, it wouldn't make sense to credit a page for participation of a success event if the page was viewed AFTER the success event, even if it was viewed during the pathing flow. also, it's noteworthy to mention that participation is credited if the pages are viewed during the same visit, correct?

Adam Greco
Adam Greco

Dirk - Yes you can, as long as you have a Success Event for each discrete interaction...

dirk shaw
dirk shaw

Hey Adam, Excellent post and very timely. Is it safe to assume you can tag discrete interactions such as discussion post, a rating or reviews as Participation metrics. The idea is to understand which types of social interactions are driving customers to a Success event. thanks again. dirk