cookieMonsterFirst, a lit­tle overview of the pur­pose of cook­ies in terms of ana­lyt­ics (not milk!).

Sim­ply put, in the ana­lyt­ics world, cook­ies are used to track and count unique vis­i­tors.  Cook­ies come in two fla­vors, first party and third party.  First party cook­ies are set by the same domain as the page the user is cur­rently view­ing, while third party cook­ies are set by a domain dif­fer­ent from the page the user is cur­rently view­ing.  As browsers add fea­tures to sup­port greater Inter­net pri­vacy, the accep­tance of third party cook­ies is diminishing. 

When cook­ies are rejected, Adobe uses a fall­back method to deter­mine a unique vis­i­tor. The fall­back method varies depend­ing on the ana­lyt­ics code ver­sion imple­mented on your web site.  Broadly speak­ing, there are three pos­si­ble fall­back meth­ods, the orig­i­nal method using IP address and user agent string, the new fall­back method using the s_fid first party cookie, and finally, the option to go with a true first party cookie.   This blog post out­lines all three meth­ods to help you make the right cookie deci­sion for your ana­lyt­ics needs (with or with­out milk!).

Third Party Cookie – Orig­i­nal Fall­back Method:

All code ver­sions pre­vi­ous to H.25.3 use the orig­i­nal fall­back method to han­dle cookie rejec­tion.   When a third party or first party cookie is rejected by the browser, Adobe uses a com­bi­na­tion of user’s IP address and the user agent string to iden­tify visitors.

Things to keep in mind regard­ing vis­i­tor count:

  • Mul­ti­ple users may be counted as a sin­gle vis­i­tor:  If two users are on the same net­work and use the same browser, which rejects third party cook­ies, then both users will appear to have the same IP address and poten­tially the same user agent string, result­ing in two sep­a­rate users being counted as a sin­gle visitor.
  • A sin­gle vis­i­tor may be counted mul­ti­ple times:  If a user’s IP address changes halfway through the visit or the user agent changes in the future, a new visit and vis­i­tor will be counted for the same user.

Third Party Cookie – New (s_fid) Fall­back Method:

delete_cookies

For ana­lyt­ics code ver­sions from H.25.3 onwards, the code con­tains a new fall­back vis­i­tor iden­ti­fi­ca­tion method.  If the vis­i­tor id cookie is rejected, the code will set a first party cookie using the site’s cur­rent domain.  The new first party cookie, s_fid, is cre­ated and set to a ran­domly gen­er­ated unique ID with an expi­ra­tion of two years.  If an s_fid, first party cookie can­not be set, then the Adobe solu­tion will use the orig­i­nal fall­back method of IP address and user agent combination.

Things to keep in mind regard­ing vis­i­tor count:

  • Vis­i­tor accu­racy matches true first party cook­ies:  This solu­tion sets a first party cookie and its rejec­tion rate is as low as any first party cookie implementation.
  • Vis­i­tor Unique­ness:  The method­ol­ogy used to gen­er­ate the s_fid value does not guar­an­tee a unique value for each vis­i­tor. Fac­tors that impact unique­ness include the vol­ume of s_fid val­ues set across the site and the num­ber of con­cur­rent hits from vis­i­tors who arrive at the same mil­lisec­ond. It’s impor­tant to note that the odds of this are very low, and you shouldn’t notice an even remotely sig­nif­i­cant dif­fer­ence between vis­i­tor counts using this method­ol­ogy and using CNAME first-party cookies.
  • Decrease in vis­i­tors after code update:  If you update your ana­lyt­ics code to H.25.3 or higher, you may see an over­all decrease in your vis­i­tor count because of the decrease in over­all cookie rejection.

First Party Cookie:

For all ver­sions of the ana­lyt­ics code, first party cookie track­ing is the rec­om­mended approach.  First party cookie imple­men­ta­tions pro­vide the most accu­rate count of vis­i­tors because of the very low cookie rejec­tion rate, as well as the guar­an­tee of a unique and per­sis­tent vis­i­tor id value.

Things to keep in mind regard­ing vis­i­tor count:

  • Decrease in vis­i­tors after update from orig­i­nal:  If you choose to tran­si­tion from the orig­i­nal fall­back method to first party cookie, you will see an over­all decrease in unique vis­i­tor count result­ing in a more accu­rate report.
  • Increase in vis­i­tors after update from s_fid:  If you choose to tran­si­tion from the s_fid fall­back method to true first party cookie, dur­ing the tran­si­tion phase you will see an increase in vis­i­tor count because Adobe will dis­re­gard the s_fid value and instead set a new cookie value result­ing in vis­i­tors being recounted.

Again, it is our rec­om­men­da­tion to always use first party cookie track­ing when pos­si­ble.  First party cook­ies will always give you the most accu­rate vis­i­tor count.  How­ever, if you need to use third party cook­ies, the new s_fid fall­back method will pro­vide you with a good degree of accu­racy. Now that you know how each method works, I hope you can fig­ure out which approach best meets your needs!

0 comments