Ammon: Click Through Rate (CTR) in SERPs has Zero Effect on the Scoring of the Site



Rizwan
Does high Click Through Rate (CTR improve position fast in Search Engine Result Page (SERP? How to improve Click Through Rate (CTR in SERP? (Other than changing title and description)
2 πŸ‘πŸ½225 πŸ’¬πŸ—¨

πŸ“°πŸ‘ˆ

Ammon Johns πŸ‘‘
No, Click Through Rate (Click Through Rate (CTR in SERPs has zero effect on the scoring of the site. After all, Google are the ones that decided the site should rank there, based on their algorithm, which was in turn based on how Google interpreted the needs and intention of the query.
Click Through Rate (CTR is used by Google to determine if they have the right algo, and are correctly understanding the intent. To compare one algo flavour (balance of signals used) against another to see which got most searchers to a point of satisfaction faster, usually in less clicks.
In an ever increasing number of cases, Google find ways to create zero-click SERPs by serving answer boxes and knowledge panels, making an answer to a simple query visible at a glance in the Search Engine Result Page (SERP itself without a single result needing to be clicked.
Google are NOT in the ranking sites business. They are in the 'satisfying the search user' business, and Google don't care what sites rank or in what positions, only whether a whole SERP gets a user more quickly and successfully to a state of being satisfied, and being sure to use Google again for future search queries.

Rizwan ✍️ » Ammon Johns
Ok thanks, so it is just a myth.
Ammon Johns πŸ‘‘ Β» Rizwan
Yes. Click Through Rate (CTR does *something*, in that it signals to Google that normal behaviours have changed, and that there might be a change of meaning or context to the whole query, but it does NOT change the rating, rankings, or signals of any one result. Instead it prompts some changes for the entire query matching, and that might prompt it to try signals the site does LESS well on.

πŸ“°πŸ‘ˆ

Andrew
It helps Google My Biz (GMB), but that's different.
If you're just talking traffic, then yes, traffic improves Search Engine Result Page (SERP
I massively prefer a link from a page with traffic than a link with no traffic and high domain stats.

Ammon Johns πŸ‘‘ Β» Andrew
How do you mean that "traffic improves SERP
Andrew Β» Ammon Johns
Traffic is a branch of the popularity signal, therefore, if you have traffic to a backlink or to your own web page, your position on the SERP will be affected.
Ammon Johns πŸ‘‘ Β» Andrew
Ah, I got what you mean.
Sadly though, no.
The traffic your page gets is not a signal that Google have enough access to, universally, to use, and is categorically NOT a ranking signal. It simply *correlates* well with things that are, like how prominent many prominent links you have on important sites that people use a lot, and thus are likely to click through those links.
But I absolutely promise you, traffic itself is NOT a ranking signal, and never has been. If you look at the papers and patents, or if you talk with engineers, or simply listen carefully to statements they've given, you find that the robustness and clarity of a signal are always key. To be good, a signal has to have a clear meaning, and to work consistently as a signal with a set level of accuracy. Google have tested and dismissed literally hundreds of signals that were found to be too variable, too noisy, or simply too inconsistent. Click Through Rate (CTR is one of those that is exceptionally noisy, as is bounce rate, and those both have far greater consistency and accuracy than traffic.
Google never use Analytics data at all in ranking sites, or as a ranking factor, and have outright said so many, many times. Not all sites use Google Analytics (GA), and some of the major ones that don't are some of the biggest and most important sites online, who definitely don't want to be giving away all of their data to Google.
Andrew Β» Ammon Johns
So Google chrome data, Google pixel data, and your analytics pixel isn't enough for Google to understand traffic?
Based on my current understanding of what you're saying, I disagree.
Heck, click through rate is also a way to identify traffic. They can identify if you clicked a result, and can count the clicks. They can also figure out if your search ended at that click.
Did you read something that told you Google doesn't have enough access? From where I'm sitting, they have PLENTY of data on traffic.
P.S. I wouldn't doubt it if Google had data share partnerships that gave it even more access to traffic data.
Andrew Β» Ammon Johns
That being said, I don't care to win an argument here but I do want to make sure that I have a correct understanding. I don't think I've ever disagreed with you so I may be wrong here but I'm hopeful for the day that I'm right and you're wrong. πŸ§ πŸ™Œ
🀭
Ammon Johns πŸ‘‘ Β» Andrew
Chrome data and even their own Domain Name System (DNS server help give them more data, but it is skewed data, representing only a particular type of user. It's not complete enough.
Android devices gave them their best shot, as at least that is the one area where they have a *majority* of all users (somewhere between 80-90% of ALL mobile devices worldwide) but still that 10% who are dedicated Apple fans are too prominent to ignore. Especially since that 10% isn't a flat, worldwide distribution, but rather heavily Western and wealthy. The Android data puts too big a weighting on Asian and African continental use.
Android does get used for their Local search, of course, but Local is a different ballgame entirely, where footfall and localization play the major part. The output is already so tightly geolocated that the input has to match.
Web search and desktop is different though. Nobody wants to get entirely different results at home from at the office. International teams work on a single web property and need to be all getting at least similar results. It means you can't use that fine detail and difference by locale to your advantage and it becomes a weakness.
There was a chance they might have decided differently, when they first announced 'mobile first' indexing. Basing all results purely on mobile feedback data, but I think they found it was too biased, excluded too much of the commercial data that only comes from the millions of workers using hardware supplied and maintained by an IT department at their place of work. An IT department that may have all sorts and levels of caching involved that thwart a lot of the actual visit data.
In all, the various details (and arguments) are complex enough that experts can argue over them for years (and have done), but the upshot is always the same – the data is incomplete and the amount and direction of discrepancy is random, changeable, unpredictable. All words that put a shiver down the spine of a bunch of scientists still trying to make an omniscient knowledge engine.
You know the famous Pareto Principle (aka the 80:20 rule)? That's Chrome in a nutshell. They are the 20% of people making up 80% of all searches kind of thing, but they still only have 20% of the households and 20% of the wallets (even if those households and wallets are bigger and richer pickings).
Many parts of the world are reliant on second hand tech. Huge swathes of land and billions of people still using older windows machines and Internet Explorer. Then there's that other group of people that we largely forget about until we have to sit down with them at Thanksgiving, who still use Windows and the default browser. They are a huge group, those 'non-techies'. It is why Bing still represents 10 times as many actual people as the search volume alone shows.
Andrew Β» Ammon Johns
All very good points. The only thing I'd pick at is the idea that incomplete means that it's not used.
Incomplete data is fine. Ranking has so many factors. It's not like traffic alone is strong enough to brute force anything, but you can use the data you have to provide a signal.
This is easy to prove. Just buy two links. One with 1,000 traffic to the page per month and one with no traffic.
The traffic link will move the needle more. This is only possible if Google uses traffic in its ranking formula.
Have you tested that yet?
Andrew Β» Ammon Johns
Engagement is an amazing metric to use for ranking. We see it in YouTube and it works very well.
Why wouldn't Google value a link with more
Engagement if they can tel it has engagement (site has Google analytics installed for instance like the majority of commercial websites).
It doesn't matter if they can't tell with some links in my opinion.
Ammon Johns πŸ‘‘ Β» Andrew
Engagement on YouTube is used to predict engagement on YouTube. It is like for like data, applicable to the exact place from where it was gathered. In other words, it is direct feedback.
Now consider how different that is from trying to apply click data or visit data, when you don't have it consistently, don't have it at all for ANY of the major ad-driven sites that are competitors to Google, and where you are already fighting hundreds of anti-trust and other lawsuits that often depend on being able to assert the use of impartial data and 'fairness'.
If you look at all of the core features of how Google approach their functionality, it is always, always, about completeness to the best of their ability. The link graph that maps out ALL the known links in existence, even the ones they cannot follow, or cannot index what is on the other side of the link.
The knowledge graph, attempting to map out all recognizable entities and the relationships between them.
Completeness, or as close as they can get to omniscience, has always been their core modus operandi.
Everything for Google is about scale, to have at least the potential from the foundation to have it all.
It is why I was actually shocked in the past year when Gary Illyes finally came out and said they don't want to index everything. Before this, they always seemed to, even spam, to at least know it was spam and mark it as such with a big fat penalty. After all, you cannot apply a penalty in your database fields to a record that doesn't exist.
All of this above, while compelling in its own way, comes within a broader context. Google are not lacking in data and signals. They don't *have* to go out and look for more, as they already ignore hundreds that they have, or could have, that were simply too process intensive, or otherwise just not quite worth the hassle.
So it isn't so much that they would never, under any circumstances use an incomplete dataset, as that they already have access to more data than they can actually process that is both better and more complete. The stuff they could already calculate and add, if only they had unlimited processing power, and no need to return a Search Engine Result Page (SERP, fully calculated and filtered in milliseconds, is absolutely mind-blowing. And they already turn that down because it takes too long, or too much, for the amount of extra end-value it adds.
Ammon Johns πŸ‘‘ Β» Andrew
I'm afraid this one I have to pick on separately:
"This is easy to prove. Just buy two links. One with 1,000 traffic to the page per month and one with no traffic."
Have you ever actually read what PageRank is, and how it works? I mean, it has been widely published and discussed since 1998. Have you actually gotten your head around the math?
Because what PageRank does is Network Theory. It doesn't measure traffic, it *predicts* it, but not in numbers but rather, in relative scale compared to all other links in the entire web.
PageRank *is* the calculation of the probability of the next click on a link anywhere in the world being on any given link. Originally it used the Random Surfer model, which assumed if you just randomly clicked links, one on each page, how likely you'd be to land at any given URL.
The weighting of links by the importance of the page they are on is the model for the more important a page is, the more chance someone is there to click on a link.
They don't need traffic data, they have PageRank – you just haven't understood why PageRank eradicates the need. Fix that.
You NEED to understand PageRank. Even if you have to sit down for a whole week and get your head around it (a bit extreme as most will grasp it in just working on the math).
Andrew Β» Ammon Johns
I have read page rank paper. I have a tattoo of Larry page on my right ass cheek.
I take an annual pilgrimage to the garage of Larry page that birthed Google.
As for the math. Not quite sure. I am not a mathematician. I'm just a dedicated learner.
Are you saying that the only algorithm that is used for ranking websites is the pageRank document from the 90s?
Ammon Johns πŸ‘‘ Β» Andrew
You know what, I've spent a great deal of time on trying to help you out to understand this stuff.
Now you're being a dick about it.
So, believe what you want. I'm done explaining.
Andrew
Google's grown quite a bit since 98.

πŸ“°πŸ‘ˆ

Rizwan ✍️
I think top position improve traffic…

Ammon Johns πŸ‘‘ Β» Rizwan
The top ranked site just about always has the highest bounce back to SERP rate (as people go on to look at other results to compare it to for price, offer, accuracy, etc).
Which, of course, highlights the idiocy inherent in those who persist in spreading the myth, because the same people who will say that Click Through Rate (Click Through Rate (CTR improves position also say that bounce back to SERP drops it… πŸ™ƒ
🀭2
Rizwan ✍️
yes top position may have high bounce rate but it is not bad for ranking. What would you say about Backlinko's 200 ranking factors post? πŸ˜‚


Buth
Bing's guys say that Click Through Rate (CTR may improve ranking, Google's guys says Click Through Rate (CTR doesn't improve ranking.
There are two things that can help you to improve Click Through Rate (CTR: title and description. This is why many SEO users and marketers try manipulate these things.
Well, you don't need to manipulate title and description, Google rewrites that. So, you should focus on the content of page. headings, paragraphs.

Rizwan ✍️ » Buth
Google will rewrite your title and description if it is unable to find perfect match details in your page.

πŸ“°πŸ‘ˆ



Can a Popular Search Engine Distinguish the Intent of a Searcher, For Instance, Tutorial, Product Page



Leave a Reply

Your email address will not be published. Required fields are marked *