Integrating Twitter Into Web Analytics [Inside Omniture SiteCatalyst]

At last week’s annual Omni­ture Sum­mit, there was a lot of buzz about social media and Twit­ter specif­i­cally.  In my Site­Cat­a­lyst Power User ses­sion, one of the things I cov­ered was an idea about how you can lever­age Omni­ture Site­Cat­a­lyst to mon­i­tor your company’s brand rep­u­ta­tion in tools like Twit­ter.  This con­cept seemed to really res­onate with the Sum­mit audi­ence, so much so, that I was given an oppor­tu­nity to share it at the clos­ing ses­sion.  The fol­low­ing will describe the con­cept in greater detail for those who could not attend my Sum­mit ses­sion.  DISCLAIMER: What I pre­sented is a “proof of con­cept” and is no way a for­mal intro­duc­tion of a new Omni­ture product.

Busi­ness Sce­nario
So this idea started with me think­ing about how cool Twit­ter is and how it could be used for mar­ket­ing pur­poses.  As an Omni­ture Twit­ter ambas­sador (Omni_Man), I see peo­ple talk­ing about Omni­ture all of the time on Twit­ter.  Some­times this chat­ter is pos­i­tive, some­times it is neg­a­tive.  I usu­ally try and send my col­leagues at Omni­ture “tweets” that I think might be rel­e­vant to them, but this can be very time con­sum­ing.  So I said to myself, “Site­Cat­a­lyst has a Data Inser­tion API that is used to inject non-website data into Site­Cat­a­lyst and Twit­ter has an API asso­ci­ated with its search​.twit​ter​.com web­site, so if you put the two together, why couldn’t you pass Twit­ter infor­ma­tion into Site­Cat­a­lyst?”  Doing this would allow you to do many cool things which I will describe below.  I enlisted the help of one of our Omni­ture Con­sult­ing geniuses and within 24 hours, we had a work­ing pro­to­type.  The fol­low­ing exam­ples of this func­tion­al­ity will use Com­cast (who I co-presented with) as an exam­ple, but these rep­re­sent test data and is not meant to imply that Com­cast is using this func­tion­al­ity.  In my ses­sion, I posed the fol­low­ing hypo­thet­i­cal busi­ness sce­nario:  You are the web ana­lyt­ics man­ager at Com­cast and your CMO returns from an exec­u­tive retreat where he/she has learned all about Twit­ter and believes that Com­cast needs to do every­thing it can to mon­i­tor what its cus­tomers are say­ing on social media sites.  The CMO calls an emer­gency “all hands” Marketing/PR meet­ing and demands to know the following:

1. How often Com­cast is men­tioned on tools like Twitter
2. If there is ever a spike (pos­i­tive or neg­a­tive) in brand-related terms (in a week, day or even hourly)
3. Who are the peo­ple most often men­tion­ing Com­cast on social media tools and who are they com­mu­ni­cat­ing with the most?
4. When are peo­ple on social media tools men­tion­ing key Com­cast product/service fea­tures that Prod­uct Man­agers should know about

So at this point, the CMO turns to you and asks if there is any­thing you could do to help… What would you say?  Not sure?  Let’s tackle them one at a time…

Mon­i­tor­ing Brand Com­ments
As stated ear­lier, the key to this solu­tion is lever­ag­ing the Site­Cat­a­lyst Data Inser­tion API and the Twit­ter API.  For those of you not famil­iar with the Site­Cat­a­lyst Data Inser­tion API, it is used to send data to Site­Cat­a­lyst when a JavaScript tag is not an option.  By com­bin­ing this with the search​.twit​ter​.com API, we can set a Site­Cat­a­lyst Suc­cess Event for every “Brand Tweet” by pump­ing the results of a “Com­cast” Twit­ter search into Site­Cat­a­lyst.  This allows us to see a met­ric chart of “Brand Twit­ter Com­ments” so we can track it by month, week, day or hour.  How­ever, why stop there?  Site­Cat­a­lyst has a built-in Alert fea­ture that allows you to be noti­fied via e-mail or mobile device when a Suc­cess Event met­ric hits a thresh­old or changes more than a spec­i­fied per­cent­age.  Why not take advan­tage of this fea­ture and send your­self (or oth­ers) an Alert when your brand is men­tioned 25% more this hour than last hour or decreases sig­nif­i­cantly day to day?  This would allow you to stay on top of what is going on in Twit­ter with­out hav­ing to con­stantly mon­i­tor Twit­ter every day/hour!  Below is a screen shot where you can see the “Brand Twit­ter Com­ments” Suc­cess Event and an Alert related to it being set:

Who’s Tweet­ing?
The next thing your CMO wanted to know who are the most active Twit­ter users that are tweet­ing about your brand.  Are there some really good brand advo­cates out there?  Are there peo­ple who are repeat­edly bash­ing your brand?  Are any of your employ­ees “going rogue” and con­fus­ing the mar­ket­place with mixed mes­sages?  Again using the API, it is pos­si­ble to extract the Twit­ter user name asso­ci­ated with every tweet.  In our proof of con­cept we did our best to extract the author and the recip­i­ent, with the lat­ter being more dif­fi­cult since there are times when there is no recip­i­ent or mul­ti­ple recip­i­ents (we are still work­ing on this).  How­ever, by plac­ing both in sep­a­rate Con­ver­sion Vari­ables (eVars), we could break­down the “Brand Tweet Com­ments” Suc­cess Event met­ric by author to see who is twit­ter­ing about your brand the most.  We decided to take this one step fur­ther by cre­at­ing a Con­ver­sion Sub­re­la­tion between the author and the recip­i­ent so you could break one down by the other (note that if there is no recip­i­ent we used “[No Recip­i­ent]”).  This allowed us to see who was tweet­ing with each other the most often.  I imag­ine that this could be use­ful to see what types of peo­ple have formed vir­tual com­mu­ni­ties and some com­pa­nies might con­sider con­tact­ing the key mem­bers of this vir­tual com­mu­nity to gather prod­uct feedback/suggestions or to lever­age them for brand pro­mo­tion.  You could also use SAINT Clas­si­fi­ca­tions to group Authors into mean­ing­ful buck­ets once you knew who they were (i.e. Cus­tomers, Ven­dors, etc…).  The fol­low­ing is a screen shot of the sub­re­la­tion report we created:

Min­ing Impor­tant Prod­uct Key­words
The final thing the CMO tasked us with was dis­cov­er­ing when, in addi­tion to our brand name, social media users were men­tion­ing spe­cific key­words that are prod­uct or ser­vice related.  For exam­ple, if some­one “tweets” about Com­cast and in the same tweet men­tions “speed” it is likely that this tweet is related to high-speed inter­net access and could be inter­est­ing to the Inter­net prod­uct man­ager.  Alter­na­tively, if a “tweet” men­tions Com­cast and also men­tions “Tivo” or “DVR” it is likely they are express­ing an opin­ion in the dig­i­tal TV record­ing arena that would inter­est the asso­ci­ated prod­uct man­ager.  So you have mil­lions of oppor­tu­ni­ties to read what your cus­tomers are say­ing, but who wants to scan through all of those “tweets” to find the rel­e­vant ones, espe­cially if this has to be done manually?

This got me think­ing about SiteCatalyst’s search func­tion­al­ity.  If we had all of the “tweets” in Site­Cat­a­lyst, you could per­form a key­word search and let Site­Cat­a­lyst find all of the com­ments that men­tion a spe­cific key­word.  For exam­ple, let’s imag­ine we use the Data Inser­tion API to pass all Com­cast “tweets” to a Con­ver­sion Vari­able (eVar) and then con­duct a search for the phrase “Tivo.”  Site­Cat­a­lyst would iso­late those “tweets” and then you can book­mark that report and sched­ule it to be e-mailed to the appro­pri­ate prod­uct man­agers at what­ever time inter­val you desire (hourly, daily, weekly, monthly, etc…).  This way, no one at your orga­ni­za­tion would ever have to use or look at Twit­ter, but instead, the infor­ma­tion they need to see would be pushed to them auto­mat­i­cally.  Best of all, there is no limit on how many searches and book­marked reports you can cre­ate so you can cre­ate hun­dreds of dif­fer­ent key­word searches and send them to dif­fer­ent groups of peo­ple using Pub­lish­ing Lists.  The fol­low­ing screen shot pro­vides an exam­ple that shows “tweets” hav­ing been sent into Site­Cat­a­lyst and a user enter­ing the phrase “Tivo” in the search box and a high­light­ing of one “tweet” that would be found:

As you can see, if you know a lot about Site­Cat­a­lyst includ­ing API’s, Con­ver­sion Vari­ables, Sub­re­la­tions, Search­ing, Book­marked Reports, etc… you would be able to amaze your CMO by answer­ing all of his/her ques­tions and be a rock star!

Track Mul­ti­ple Brands
Another con­cept related to this that we have explored is the idea of track­ing mul­ti­ple brands.  There is no rea­son why Com­cast, in this exam­ple, could not also cap­ture “tweets” about its com­peti­tors or sub­sidiary brands to see them side by side.  This would require the use of an addi­tional eVar or poten­tially some addi­tional Suc­cess Events, but we got this work­ing in our prototype.

Next Steps
As men­tioned pre­vi­ously, all of this was done as a proof of con­cept, but as you can see, the con­cept has great poten­tial.  We at Omni­ture are going to explore this topic more and hope that you do the same.  We hope to add more infor­ma­tion about this to the Devel­oper Con­nec­tion.  We will also con­tinue to explore new ideas related to this, but I encour­age you to leave com­ments here with your ideas on how this con­cept can be extended.

If you want to learn the imple­men­ta­tion details of this solu­tion, please refer to the fol­low­ing tech­ni­cal blog post: http://​blogs​.omni​ture​.com/​2​0​0​9​/​0​2​/​2​4​/​i​m​p​l​e​m​e​n​t​i​n​g​-​t​w​i​t​t​e​r​-​d​a​t​a​-​t​r​a​c​k​i​n​g​-​i​n​-​o​m​n​i​t​u​r​e​-​s​i​t​e​c​a​t​a​l​y​st/

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