Last night’s SEMPO/AIMA event includ­ed speak­ers from Google, Bing, and Newell-Rub­ber­maid. All pan­elists came to dis­cuss their efforts in social search, or the incor­po­ra­tion of social net­works into search.


The algo­rithms of the major search engines increas­ing­ly try to bring the user results that are rel­e­vant to them. Social search is one means to this end. If search is able to incor­po­rate the rec­om­men­da­tions of your friends and online con­tacts through tools such as Facebook’s “Like” but­ton and Google’s “ 1” but­ton, then the results are more like­ly to be mean­ing­ful to the user.

The engines have inte­grat­ed social into search with strong visu­al cues, includ­ing images of your friends and bright­ly col­ored icons. It is there­fore more impor­tant than ever to have your con­tent shared via the social graph (A term often used by Face­book, but gen­er­al­ly mean­ing the map of rela­tion­ships online).

Take­aways for brands are to make it easy for users to share your con­tent. Place the “Like” but­ton and the “ 1” but­ton on your arti­cles, videos and pic­tures. Some brands have even begun incen­tiviz­ing their users to share their con­tent, pro­vid­ing dis­counts for “Like” and ” 1″ clicks. A much smarter approach is tak­ing this oppor­tu­ni­ty to improve the qual­i­ty of your web con­tent casi­no jame­shal­li­son so that users will gen­uine­ly want to share your con­tent. If you have been toy­ing around with the idea of build­ing out a blog, guide, wid­get, or oth­er asset that will be use­ful to your user base, it’s time for those ideas to go live.


As for the future of per­son­al­ized search there were three issues touched on last night that weren’t direct­ly addressed by Google or Bing, but were left as food for thought.

  1. Users can “Like” con­tent, but can’t “Dis­like” con­tent. There is no way with­in the cur­rent sys­tem for users to con­tribute their opin­ion when they come across a par­tic­u­lar­ly bad site, ser­vice or prod­uct. Of course the absence of a “Like” will not con­tribute to a link’s edge in the search engine results page as much as hav­ing one would, but are the engines going to penal­ize sites that have not been “liked”? Will sites with­out social shar­ing links be penal­ized?
  2. What does “Like”ing con­tent real­ly mean? Stan­dard star rat­ings for exam­ple allow you to show just how much you like a prod­uct by giv­ing it some­where between 1 and 5 stars. Is a “Like” an unequiv­o­cal endorse­ment of a prod­uct? See­ing that your friends rec­om­mend a prod­uct may help nudge a user in a cer­tain direc­tion, but there are always pros and cons along­side any rec­om­men­da­tion. Fur­ther, we are all guilty of post­ing cer­tain things on our social media pro­files that we may not ful­ly endorse, but post instead to main­tain our online per­sona. For exam­ple, I may lis­ten to 80’s rock 95% per­cent of my day, but post Mozart to my Face­book wall to main­tain my intel­lec­tu­al online per­sona.
  3. As search becomes increas­ing­ly per­son­al­ized, the infor­ma­tion we are exposed to will become more and more tai­lored to our per­son­al inter­ests, and poten­tial­ly impor­tant infor­ma­tion could be edit­ed out. Eli Paris­er, author of “The Fil­ter Bub­ble” says in on the sub­ject that “the search engine algo­rithm sends us what it thinks we want to see, not nec­es­sar­i­ly what we need to see.” As search becomes more per­son­al­ized, we may not be exposed to the impor­tant, uncom­fort­able, and chal­leng­ing infor­ma­tion that we oth­er­wise would have. Poten­tial solu­tions to this prob­lem are giv­ing the user more con­trol over the per­son­al­iza­tion of their search­es and pro­vid­ing more trans­paren­cy into the per­son­al­ized search algo­rithm.

I look for­ward to see­ing how the search engines will address all three of these issues. With the speed that the social graph has been incor­po­rat­ed into search, I wouldn’t be sur­prised if we see answers soon.