The target attendee for PubCon is 30 and under. The first day is just Master's Classes that companies send their employees to. For new people, the sessions are the first time they've had a chance to learn. For the experts and exhibitors, it's a trade show where they network, try to get new business, and see old friends. For the SEO industry, relationships are a traditional part of their customs.
RankBrain - has been around before anyone took notice of it, working in the background. That was the nugget they got from Google. The other was that even Google doesn't truly understand what their "black box" is doing.
Eric starts by asking Gary to explain RankBrain.
Gary: RankBrain is a machine-learning algorithm. It tries to predict what would be a better result based on historical data. It looks at a bunch of signals. It adjusts ranking based on the query. Gary says that it works really well with queries that they’ve never seen before – long-tail queries. They typically use data collected for specific queries to decide how to rank results for a query. But if the query is new, they don’t have that data yet and that’s where RankBrain excels. It makes predictions based on similar queries in the past.
Eric: Is RankBrain doing language analysis?
Gary: You are going too deep here. It’s doing many things.
Eric presses for a list. Gary is dodging. We are six minutes into this talk.
Eric keeps pressing and says that it was originally discussed as being focused on query analysis. It sounds like there may be some other elements?
Gary: There are many elements. It hasn’t taken over the entire algorithm.
Eric: If it’s responsible for 20% of the queries, that’s a lot of queries that RankBrain is potentially acting on.
Gary says yes.
Eric: Why did you end up calling it RankBrain?
Gary has absolutely no idea. Eric saves us all by changing the subject.
Eric: Let’s talk about machine learning more generally. It’s obviously a very big initiative at Google. The new head of search is a big proponent of machine learning. Can you tell us more about what Google is doing with machine learning?
Gary: Machine-learning is a tool. It’s like having a Swiss Army knife. You can use it for many things. It’s not appropriate for everything but you can use it for many cases. You probably don’t want use that Swiss Army knife to punch a hole in concrete because it will not work well. Similarly, RankBrain will not work in every scenario. We are using it where we can. Obviously, we are not trying to deprecate all of our other algorithms. If it’s working well, then we prefer not touching the algorithm. But we are experimenting with what else we can do with ranking.
The emphasis on mobile was a bit of surprise for me and I disagreed with several speakers on what they were saying. (As a side note, when I began to push back on PubCon on my Facebook page, I took heat for it. I removed everything as a result.)
Google reps use PubCon to announce new things and answer questions. SEO's pounce on it.
AMP and SCHEMA were hot because the attendees came to learn about it for their jobs.
People don't come to forums to learn like it was done in the old days. These particular forums lost ground in the past few years and continue to do so, which is a shame because the community here are teachers at the expert level and don't BS around.
Back to mobile...it's Google making the decisions for marketers and website owners. There isn't enough rejection of this and so it continues. I was telling a young person on the bus from the hotel to conference center that I work at a horse stable where the owner and stable help do not have smart phones and what laptops or tablets they have are ancient. They don't have the funds or interest in upgrading. They still need to purchase, learn, search, socialize online, but it's as if they don't exist. The same goes for baby boomers and retirees. Marketers aren't thinking about them and neither are designers. I was definitely on the outside looking in at PubCon.
Edited by cre8pc, 18 October 2016 - 10:49 AM.