google-logoIf you had a chance to interview a Googler, what would you ask?  Do you love your work?  What would you do differently if you were in charge?  Well, there’s one question that was asked by some of the folks at Search Engine Land, and that was what sort of algorithms do the Googlers themselves love?  Gary Illyes of Google was asked this in part two of their interview with the search quality analyst.

To date, Gary said that his favorite algorithms are Panda, Penguin and the localization project he worked on previously.

He loved Panda because “people found very interesting ways to try to game our ranking algorithms and Panda figured out how to deal with those weird things.”

He also enjoyed the localization project because he speaks “several languages,” and especially with Japanese, Chinese or Korean, because the language and queires don’t have any spaces, and provides some interesting challenges.

He said:

I speak several languages working with that, with the language team for example was really really really cool. There are tons of languages where for example where you don’t have space and that the concept of space in the language, like Japanese or Chinese or Korean.

And when people type these like sentences or these super long phrases in the search box and then you have to figure out for example how to segment those because our indexing algorithm is working based on tokens not like cool phrases. That that was really really fascinating. I thought super interesting.

But right now, Penguin is his favorite algorithm because of how it devalues over demotes.

Here is the audio snippet:

 

Here is the transcript:

Danny Sullivan: Gary, what’s your favorite signal?

Gary Illyes: Favorite signal?

Danny Sullivan: Yeah. No, your favorite signal for ranking?

Gary Illyes: Oh, for ranking.

Danny Sullivan: Oh, you got 200. Just give us one. Give us one unusual, weird signal that nobody knows about. You can totally tell us; nobody would care. Not going to ruin anything. Just one good signal.

Gary Illyes: I worked on local or localization. Basically ranking different in different countries and based on and based on the language. And that was really fascinating.

I speak several languages working with that, with the language team for example was really really really cool. There are tons of languages where for example where you don’t have space and that the concept of space in the language, like Japanese or Chinese or Korean.

And when people type these like sentences or these super long phrases in the search box and then you have to figure out for example how to segment those because our indexing algorithm is working based on tokens not like cool phrases. That that was really really fascinating. I thought super interesting.

Danny Sullivan: That’s cool.

Gary Illyes: But it wasn’t the secret. I think we already said that we didn’t say we use language Panda. But I will tell you why.

I really like the Panda algorithm or Panda signal because people found very interesting ways to try to game our ranking algorithms and Panda figured out how to deal with those weird things.

So that’s also great.

And now Penguin, Penguin is probably now the most favorite.

Source