|
|
http://www.theideais.com/
——————————————————————————–
There are many ways to think about SEM and many realms one can focus on. recently I had some thoughts about the algorithm that decides which ads are shown. As I understand it (and correct me if I’m wrong) ads are shown based on the user’s search phrase while calculating several factors such ad the bid the advertiser put on the word, it’s CTR and the destination landing page Quality Score. When I think about it, It looks to me that somehow the algorithm boverlooks the user. If you ask me, matching the ad only to last search phrase the user enteres is not enough. I mean think about it, after all, over 400 million searches are performed every day, not all of them are relevant for ads (probably most of them are not, interesting piece of information. Does anybody have a clue of how many searches are related to ads?). But one thing is certain, all of the searches are performed by potential buyers. What happens is that there are tons of searches that do no show any relevant ads.
So here’s my SEM challenge, and here comes the solution.
The idea is, the searche engines algorithm should incorporate also the users search history in addition to LSA and LSI and some trial and error factors that can determine the type of user and decide which ads to show. I know it sound complicated but bear with me for a while. Think about the search history search engines hold for the user. They actually gather all the user thoughts and interests. If the search engine uses Latent Semantic Analysis on users’ search history It can determine the user’s areas of interest. Now as I said, for searches that have relevant ads, matching an ad to the last search phrase is fine. But many searches are not related to any ads and therefore are waisted advertising space, why not match the ads to some other stuff a user is interested in and not just her last thought. Take for example, a typical user, she often looks for brand names like Fendi, Prada and Gucci (tons of ads), but she’s also a spiritual girl and is interested in Krishna, Rama and other oriental spiritual stuff (no relevant ads). The first group of words is highly relevant but the latter is not and frequently shows no ads. Pop up an ad about Discounted Fendi hand bags and you have a winner here.
I’m not sure incorporating LSA on users’ search history is The Future of Search Marketing but I’m quite confident that it can monetize a hole lot of unused ad space that is currently just waiting to be used. After all we’re all buyers and we all love to get good deals even when we’re just thinking about something else.
