Gary Illyes (Google): UX Is Ranking Factor On Mobile, Not On Desktop

[aakk9999]

Gary Illyes, webmaster trends analyst at Google tweeted today that User Experience influences Mobile SERPs but currently not Desktop SERPs.

On mobile, sure, since UX is the base of the mobile friendly update. On desktop currently no.
[twitter.com...]

[aakk9999]

I am wondering whether the UX he talks about for mobile is purely based on how the page renders on mobile device or it is actually based on the UX signals from real visitors.

That is, a site not built for a mobile may provide a have a great UX because it is liked by searches who access it a lot on Mobile despite not being built for the mobile. Another site may have a mobile design, but may leave visitors unhapy and they click back to SERPs a lot.

[goodroi]

I don't want to say Gary is wrong because he is much closer to the secret Google formula than I will ever be but my research and experience finds that better UX does help desktop rankings. Maybe Google isn't directly measuring UX but they are measuring something else that is indirectly impacted by UX so improving UX indirectly boosts ranking?

Actually I will say Gary is wrong or at least not being careful with his words. Google has said that they do look at page layout and penalize ad heavy pages. Seems to me that page layout is related to UX.

Even if Google is or is not directly measuring UX, I am not going to change. I will always want the best UX because it leads to better conversions and profits. It also helps me when I am recruiting link/traffic partners. Webmasters don't want to link to crappy sites.

[supercyberbob]

Umm, what a bunch of crap. And the Above the Fold algo was what, a Greek myth?This company needs to wake the duck up.

[martinibuster]

Not true at all. There's the Page Layout Algorithm, [searchengineland.com] an algorithm that penalizes sites with too many ads up top. Quantity of ads was also a factor cited by Matt Cutts in the Wired magazine interview about Panda.

There was an engineer who came up with a rigorous set of questions, everything from. “Do you consider this site to be authoritative? Would it be okay if this was in a magazine? Does this site have excessive ads?”

Then there's the whole thing about Google insisting that publishers open up their CSS and JS files so that Google's crawlers can render the web pages for evaluation as a user may see it. That involves user experience factors as well.

[Robert Charlton]

A few posts earlier, in that same twitter thread, Gary Illyes says something which, as I read it in context, keeps content as the base but doesn't necessarily rule out UX as a factor...

Gary Illyes ‏@methode 11h11 hours ago

@jennita @ToriCushing @dr_pete yes, think about the user, but if you don't have content for them, how are you helping the user with UX only?

[Dymero]

Maybe it's not as important on desktop because form factor considerations are not as much an issue, but if a site has poor navigation or categorization (which I'd argue are user experience issues), I certainly wouldn't expect it to rank higher than one that doesn't have these problems, all other things being equal. And, as people above pointed out, if you can't find the content within a sea of ads, that's also poor UX, and we know they have an algorithm for that.

[toidi]

The engineer at a car company working on engine design most likely does not know what new cars the engine will be used in or even what new cars are in the works.

No slight on gary, he might be able to recite the algo in his sleep but by the time the powers that be are done with their manual tweeks, the algo no longer works as designed.

So, everyone's guess is a good as a guess can get.

[martinibuster]

Discussing this on Facebook, Mikkel DeMib Svendsen (a former mod at SEW) noted that Ilyes framed his subsequent statement around the use of GA data. So that Twitter conversation is getting messy. If that's where he's coming from, that his comment is in reference to GA data, then it sort of makes sense, unless it conversely means they use GA data for the mobile algo, which doesn't make sense since GA data is supposed to be used only in aggregate. So is Ilyes talking about apples and oranges? Messy.

[guggi2000]

UX is so important for website visitors that it makes any discussion about being a ranking factor redundant. If G picks it up, it's good for everybody.

[nomis5]

Until Gary defines exactly what UX means, any discussion about it is worthless.

Too many times G say one thing about subject X only for subject X to be redefined by them as and when it suits. It took eons to define what "mobile" meant and to exclude tablets from that definition. Even now it's not 100% clear and neither is his tweet.

[Robert Charlton]

...that Twitter conversation is getting messy...

Twitter not always a great platform for subtlety. We're talking about one additional question and one response, not really a reexamination of the original statements. Here's where that conversation has gone. My emphasis added at the word "analytics", adjacent to "bounce rate"...

Daniele Nespoli ‏@dnespo May 13
@methode Thank you Gary, does UX include serp bounce rate for desktop users?

Gary Illyes ‏@methode May 13
@dnespo we don't use analytics/bounce rate in search ranking

It seems to me that this is a discussion we've had in this forum many times before, and I agree with Mikkel (on FB) that Google would not want to risk its credibility by using Analytics data after saying it doesn't. There are many reasons in any event why Analytics data would not be a useful basis for an algorithm.

Google has also used return-to-serps as a signal, which is not using Analytics data. Early on, in Panda, return to serps was considered to be a signal that was confirmable by multiple sources and could demote a page. I assume that Google is using an analogous signal in evaluating Mobile.


FWIW, the above is not the original meaning I'd attached to the question of "UX as a ranking factor".

The meaning of "UX" has evolved over the past several years to encompass the entire experience of using a site... and certainly measuring it is much more subtle and multi-faceted than return to serps. Page layout and design, ease of navigation, frequency of sharing and bookmarking, etc, all enter into the success of a site and the desirability for Google of sending searchers to that site. Google I'm sure generates heuristics based on aspects of user experience, and these get incorporated into algo evaluations in many ways.

I feel that these "UX" factors are all major considerations in the success of a site, with or without Google.

Regarding the distinction being made between mobile and desktop, though, the question may be what Gary means by "ranking factor", and perhaps also (and I can only guess here) what's the difference between a positive ranking factor and a negative filter?

A page has got to pass some usability tests to get into the mobile index. So, usability is in a sense a "ranking factor" in mobile.

But, on the desktop side, Google initially allows pretty much anything that it can crawl. Gary may be saying that the ranking is determined by content, relevance, link factors, and personalization... and that usability doesn't boost you any higher. But, lack of usability might drop you lower. There's a gray area in there as well of query rewriting and perception of intent, not all exactly "personalization" in the traditional sense, but still a layer of user behavior.

I'm not pushing this theory... just trying to get at what he was saying, to see if there's anything we can learn from it.

[seoskunk]

we don't use analytics/bounce rate in search ranking

Does google use Chrome data though? Visually people use Analytics to monitor bounce rates I would suspect google uses a variety of other factors such as Chrome browser and info from third parties as well as its own user behaviour on search. To get the correct answers from Google you have to ask the exact question. There as slippery as politicians.

[Robert Charlton]

I'm wanting to revisit a comment I made toward the end of my last post above...

Regarding the distinction being made between mobile and desktop, though, the question may be what Gary means by "ranking factor", and perhaps also (and I can only guess here) what's the difference between a positive ranking factor and a negative filter?

In light of the recent "Quality Update" [webmasterworld.com...] which is generally considered a positive ranking factor, as distinct from a negative filter, I'm wondering whether Gary would have made the same statement today.

It could be that this particular April-May 2015 Quality update... which we're assuming has to do with the history of searches involving the site... does not specifically involve UX metrics, and as such is not a "UX factor".

[Spiekerooger]

which we're assuming has to do with the history of searches involving the site..

+1


UX doesn't seem to be much of a factor in desktop search - otherwise google would have killed an ecom site of mine that is broken (users can't order any of the listed products :o ) -but nothing happened to that site in the last updates. And even in mobile search I see a lot of mobile friendly tagged pages that are not mobile friendly regarding UX.

Also I see lots of sites with bad UX ranking higher after the quality update.

[martinibuster]

I see lots of sites with bad UX ranking higher...

Maybe those are better than the alternatives (particularly in a niche with limited alternatives)? If UX is a factor that affects ranking, I don't believe it's a make or break factor, that a poor UX would cause a relevant site to be excluded- and less so in a niche that might not be particularly competitive. As Mr. Charlton noted, that it's probably a bigger factor in mobile.

[Spiekerooger]

Maybe those are better than the alternatives (particularly in a niche with limited alternatives)?

Both not the case.

I rather assume that UX (in desktop search) is more a secondary vector - not calculated directly but felt if user signals decline 'cause of poor UX. So I'll watch if this will happen in the next couple of weeks.