This past Friday, I asked John Mueller of Google a bit more on if and how Google may use machine learning for adjusting the weights of various ranking signals. The short answer is, Google may or may not do this, depending on the specific ranking signal.
But keep in mind, I narrowed the question specifically to if and how Google may use machine learning for adjusting the weights of individual ranking signals. It may be also the case machine learning is used amongst multiple ranking signals or to maybe even create new ones based on the query. It is hard to know for sure. I'll get into this more later but first let's get into what I asked John.
John said "maybe we use the values directly for machine learning, maybe we adjust them manually, maybe we just start with something manual and then see how it goes."
I asked the question around the 4:11 mark into the video:
Barry Schwartz: So I had a question about Steve the search engine. So in your Google podcast, Search Off the Record, you said you pretty much use that as a way of talking about some ranking signals because you don't like talking about Google ranking signals and one of the things I found was interesting was that you didn't mention anything about machine learning changing the weights of those ranking signals. I found that very interesting because and obviously bing is very into like saying yeah we have lots of ranking signals but we don't know what the weights are at any moment because machine learning and AI changes that on the fly based on tons of factors. Does Google do a lot of that or it depends on the specific ranking factor or signal?
John Mueller: I think it depends is probably the right answer. It's something where for some some signals I know that we do a lot of machine learning to try to figure out how we should integrate them and for others we we don't use that much. And it kind of also depends a little bit on on the specifics of what we're trying to figure out. In the sense of do we have a clear metric that we could kind of base this machine learning system on or are we doing something like I don't know training the machine learning system on clicks and then it just finds the most click-baity titles that we can show and uses those in search. Where you kind of have to watch out for it for things like that. So that's something where for some elements we definitely use machine learning for other elements we don't use it as much.
Barry Schwartz: So could I ask specifically about when is HTTPS it seems like it's [Google is] not, at least Steve, not using it using machine learning to adjust the weights and it seems like the page experience update currently won't launch that way or you don't want to say what Steve or Google…
John Mueller: I don't know. I mean the different weightings of different parts of the algorithms is is something that is pretty tricky to do because you can't just kind of like manually say oh this is weighted ten percent and then suddenly everything else is 10 percent less overall. It's like you kind of need to watch out for the whole system so. I could imagine some parts of that are things that we evaluate with machine learning, maybe we use the values directly for machine learning, maybe we adjust them manually, maybe we just start with something manual and then see how it goes. I honestly don't know.
Here is the video embed:
Now, Google has historically downplayed the significance of machine learning and AI with Google's ranking algorithm. Google even said many algorithms are not suitable for machine learning. It is a topic I am fascinating by because I honestly don't know much about how it works. Whereas Microsoft Bing seems to be all in with machine learning and search.
What do you all think of this?
Forum discussion at YouTube Community.
