Google Takes a While to Evaluate Quality Changes

Webmaster 2 6:16
So I have a two-part question about Google evaluating website quality. So in the past office hours, you’ve talked about if the overall quality of a website is an issue, specifically, if it has 10s of 1000s of pages or more, that aggressively deindexing low-quality pages is a good idea, which makes sense. So removing bad pages from the index that haven’t had a chance to meet a certain quality threshold just yet, with good user-generated content.

And I guess all of this in an effort to make Google reconsider the quality of the website as a whole. So we have a website and that camp, and in the last couple of months, we’ve made an effort to do just that. And in past office hours, you’ve, you’ve said, it can take months, or even up to half a year for Google to recalculate the quality and treat a website differently in search. So question one would be do you sort of stand by that time frame? Can anything affect the timeframe?

And is it sort of correlated to the size of a website? And then a quick follow-up for question two is, you know, if we’re a small team, and that’s with SEO being our primary acquisition channel, how would you suggest measuring that that was indeed the problem that we tackled? So for instance, like, would you look for gradual improvement? Or would it sort of be like flat for many months, and then you sort of would see, you know, things shoot up if we did, indeed do the right thing? Because I guess it’s a scary proposition to sort of remove then pages from your index, and then wait six months? And then see if you’ve done enough, you know, does that all make sense?

John 7:58
Now, I think the timeline is probably about right. I’m sure there are situations where you can go a little bit faster. But on a whole, that’s, that’s probably something which, which takes that, I don’t know, half-year or longer almost, or things to kind of settle down. And it’s, I think, partially due to us needing to reindex everything on the site, understanding the site again, partially. Also, just because of the quality signals that we collect, they just take a long time to kind of be built up.

And so that’s something where a couple of months, half a year or maybe even a bit longer, I think that would be kind of the norm, I’m sure there are situations where it can be faster. But for the most part, it’s probably in that range. With regards to recognizing when you’re kind of on the right track, I think that’s super hard. Because, on the one hand, you have the situation where you’re removing pages, so you might see a drop in traffic from those pages.

I think, overall, if you look at the metrics for your site, the pages that you remove are probably pages that don’t get a ton of pageviews. So maybe that’s something that wouldn’t pull down your overall traffic from search, or not, not that much. But it takes a long time for things to kind of start seeing more and more traffic and kind of being picked up as being higher quality. So one thing I will try to do is to find—try to find some other proxy metrics that you can use for recognizing the quality of your site.

And that could be something like, like looking into analytics and the looking at things like I think, like, they call it the engagement rate now or other metrics where you have kind of time on site where you have something where when you look at the metrics now or when you compare to the metrics that you have, maybe a month ago, before you started working on this, you can tell that actually those were pretty bad. And with these improvements, we see kind of this shift in user behavior.

And it’s not so much that we will use that user behavior directly in search, but it—but it’s, it’s something that’s more like a leading indicator for you to let you know that you’re on the right track. So that’s kind of the direction I would look at it there. On the one hand, kind of, you shouldn’t see a big drop from the pages that you remove, you should see kind of like this slight subtle shift in user behavior through the other metrics that you track. And then over the course of, I don’t know, a couple of months, you probably see like a gradual increase. And then at some point, you see maybe a bigger jump happening.

Webmaster 2 10:50
Okay. And thanks, that really helps. So would one strategy that you recommend be: could you test a small subset of pages, take some of the better-quality ones, and maybe do like a campaign around them to sort of get them shared around a little bit if you’d believe and that they’re good quality? Would that expedite the reconsideration of just those pages by any chance? Or is it kind of more like, no, if you have a website quality issue as a whole, those might see a suppression, too, no matter what?

John 11:23
I think for tracking the metrics yourself, if you’re looking at analytics, that’s definitely a good way to look at it. Where if you can drive traffic there and say, like, instead of, I don’t know, 50 people a day coming to these pages, you have 1000 people coming to those pages, you have a little bit clearer metrics. From an SEO point of view, from a kind of a Google search quality point of view, I don’t think taking a small subset of a page of a site and improving the quality of that would be enough to make us say, “Oh, this is a different part of a site.”

And this is high quality. And this is kind of medium quality, we’ll treat it differently. I think, for the most part, we would look at it or try to look at it as an overhaul thing. And it might be that the smallest set that you improve, kind of move the overall average up a little bit higher. But that’s not something I would kind of wait for. Because if you’re talking about a half a year, and you improve 10 percent of your pages, and after half a year, you realize, oh, I need to improve even more. Like you, you’ve kind of wasted a lot of time.