Rudi Shumpert is the mar­ket­ing web devel­oper at Ariba, Inc., the lead­ing provider of spend man­age­ment solu­tions, respon­si­ble for Omni­ture inte­gra­tion. He has 14 years expe­ri­ence in web appli­ca­tion devel­op­ment tech­nolo­gies and has recently been work­ing on improv­ing the effec­tive­ness for Ariba’s web ana­lyt­ics pro­gram. Lately Rudi has been actively inte­grat­ing Ariba’s instance of Omni­ture with a vari­ety of exter­nal data sources, includ­ing Twit­ter and Yahoo pipes. Rudi has also been inno­vat­ing around video track­ing, cre­at­ing new ways to track post-production flash video, Win­dows media, and even YouTube files. We recently chat­ted with Rudi to learn more about his devel­op­ment work with Omniture:

Q: What were some of the dri­vers behind your Twit­ter inte­gra­tion? What data are you pulling and how are you using that to improve Ariba’s mar­ket­ing efforts?

A: For Twit­ter we have a set of search terms includ­ing our brand name that we are mon­i­tor­ing. Cur­rently for those terms, we are cap­tur­ing the user name, full text of the tweet, and the search term. With that we can see by search term the con­ver­sa­tions tak­ing place. This also pro­vides a his­tor­i­cal view into these con­ver­sa­tions as well. Part of our mar­ket­ing effort is to be part of the con­ver­sa­tions about Ariba and the areas we spe­cial­ize in. Being able to have this data and mon­i­tor it within Site­Cat­a­lyst is exciting.

Q: You recently inte­grated Omni­ture with Yahoo Pipes. What did you inte­grate and why?

A: We found a user cre­ated pipe called “Social Media Fire Hose” and it is just that. This par­tic­u­lar pipe will search blogs, news agen­cies, and numer­ous other social media sites on any set of terms you pro­vide. It will also return the data in a JSON for­mat that is easy to con­sume and parse. This data com­bined with the data pulled from Twit­ter, gives us a more com­plete view of the “buzz” on the terms we are search­ing for. And it will pro­vide us with a greater abil­ity to join these ongo­ing conversations.

Q: You recently did some inno­v­a­tive work to improve your abil­ity to track flash video. What prob­lem were you try­ing to solve here and how did you solve it?

A: We have a large col­lec­tion of video files in our online Resource Library for our users. (Webi­nars, demos, etc.) in var­i­ous for­mats (wmv, flv, and even some on YouTube) and from a main­te­nance stand­point I wanted to find one player, one method to be able to track them all, and get the same level of use­able met­rics from each for­mat. The task of con­vert­ing our entire library of wmv files to flv was not a good solu­tion to the prob­lem. A few weeks back I read a blog post about a player from Long­Tail video, called JW player that would play wmv, flv files, & YouTube videos, and also had an open API for being able to work with the play­ers event lis­ten­ers. After a bit of research and test­ing I was able mon­i­tor the events and track all the video events we needed and sub­mit the data into Omni­ture. This method has allowed us to track all the videos in our exist­ing library with­out hav­ing to spend time con­vert­ing videos or embed­ding code into a cus­tom flv player.

Q: What’s next on your inte­gra­tion roadmap with Omniture?

We are in the begin­ning phases of using Gen­e­sis to inte­grate Sales­force & Omni­ture to pro­vide both sales and mar­ket­ing a bet­ter view into the inter­ests of our web vis­i­tors. Aside from that we will con­tinue to enhance the tag­ging of ele­ments on our site to improve the level of detail in the Omni­ture reports.