- Othman Laraki of Twitter Search provided some additional insight into the company's latest offering. He said that ultimately Twitter addresses a hyper-local question on a massive scale. True, true. He also stated that results in real time are less about order and more about context, and that -- overall -- we need to think about SEO differently.
- Meredith Ringel-Morris from Microsoft provided the results of an interesting study comparing search and social. She stated that the rise of social represents the rise of more an "asking" than a seeking behavior. The study compared the results from asking and finding answers on social networks, versus seeking (coincidentally, social won the test as I searched for Ringel-Morris's name for this article, as it wasn't listed in the agenda - thanks @GilR and @I80equipment09). One interesting study finding (which seems to support common sense), was that users liked getting reinforcement in both search and social. In other words, they liked getting strong responses from their networks, and also seeing similar recommendations appearing in the search engines.
- Ringel-Morris also painted a picture of little bot friends attaching to our networks (like Bing) that are able to answer questions based on conversation. As it begins to manifest, I'm sure this friend will probably be very helpful at times, and highly annoying at other times.
- Brian Theodore gave a rundown on Yahoo's real-time product, and observed that Twitter language is much like a command line. In his words, "MS-DOS lives on." It is a bit puzzling to me that Yahoo does not account for recency factors in its own engine - this is, after all, real-time search. Theodore also stated that the real challenge for the growth of real-time search is changing consumer behavior. In my own ponderings on this topic, it is clear that general audiences, much less the search crowd, are not using real time search to any great degree. The value to the user is there in a big way, but they simply do not know how to use it. Perhaps it will be the job of the engines to push the results enough to where consumers can't live without it (QDF , or query deserves freshness).
- Jeremy Hylton at Google also gave a decent rundown on QDF in real0time search), the factor that pushes a real time search into mainstream Web search. He said there are about 10 signals, particularly when "authoritative people" are talking. Social relevancy is alive and kicking.
- My colleague, Rachel Pasqua, director of mobile strategy for iCrossing, presented a very interesting case study on optimizing iPhone Apps for visibility. Downloads beget more downloads, and a key takeaway was to push as many downloads as possible, using a variety of methods like Google Mobile Adwords, standard SEO, and optimization of the app store profile, for which she provided a detailed rundown.
Google May Be Smart, But SEMs Aren't Stupid Either
Perhaps the most controversial and enlightening panel of the whole show was a session titled "Inner View: Google's Keyword Research Tools." Danny Sullivan hosted Google's Baris Gultekin, Group Product Manager, AdWords, and about 200 people were in the room.
I'm certain that Gultekin stayed very close to the party line in discussing the keyword tool, but it ultimately proved how out of step Google is with its core advertisers -- the same audience that funds its primary revenue stream. Danny Sullivan demonstrated how the tool often skews toward upper-funnel head terms and phrases (around the test phrase of "Britney Spears"), and showed that so much raw data is weeded out, it creates problems for marketers on many levels.
Gultekin responded that this type of high-level skewing and "don't worry, we'll find your terms for you" approach is just what Adwords advertisers want. Then Sullivan asked for a show of hands from the Adwords advertisers on who liked the tool in its current state, and none raised their hands. He then asked how many wanted raw data, and just about everybody's hand shot up. Gultekin then stated it was something they were working on.
It was painful on so many levels. We went through this whole dance during the session, and it was finally announced by Gultekin, after the worst of it, that the tool was being changed to incorporate more raw data. Another audience member pointed out that this creates a skewed bidding market for upper funnel terms, in a way that could artificially drive up bids. He was right.
Overall, we learned that the value and perception of the Google Keyword Tool has taken a nosedive, even for those using it directionally, and that was a hell of a payoff for a single session.
Next day I spoke with in-house search marketer (who asked not to be named) who said aid that when he came to a conference, he expected to be able to go to the Google booth and get answers. Whatever his questions, he wasn't satisfied. On top of this, he mentioned his great disappointment at Google's Adword keyword tool, especially in light of the day's previous panel.
A lot of these sentiments may not be new, but clearly the big engine needs to do more to address them.
At the end of the day, it is this core group that provides Google's core revenue stream that it allows it to fund neato things like the self-driving car.