> Isn't one of the key ranking factors time users spend on the page?
In the original PageRank? No.
These days, who knows? But Google only knows how long people stay on the page using Google Analytics and many sites (mine, for instance) don't use it.
I am a quick reader and are very good at quickly skimming articles. I hope Google doesn't rank pages based on time spent, since Google would be optimizing their search away from my preferences.
I researched Google's algorithm a while back and a a fairly strong ranking factor IS how long they spend on your page/site (whether you have GA or not) but that can backfire. (I know that Google's exact algorithm is unknown but it's based on experiments by SEO sites)
Here's how I understood it to work:
If a user finds you on a Google search and then clicks the link, Google times how long it takes for you to come back to the search results. So if you immediately backed out of the page they would take that as meaning the result was crap and they took that into account in their algorithm. The thinking being that it was an irrelevant result.
However, that counts against sites that provide the answer you want right at the top of the page so that when you open the page, the answer is staring you in the face... so obviously you are going to back out again as you have your answer.
This was from around two years ago when I was researching building an indexer and search engine for kicks so my memory of how it worked is fuzzy.
If only Google had a technology that would preload the top results (maybe coming from a Google CDN) and show them right on the search results page. I’d be AMP’ed for that.
Users may back out again once they have the answer, but they probably won't click on another search result. This should show Google that the user found their answer on that page.
Every google search result you see is masked to have a boat load of tracking that identifies you, your search query and a variety of other details. When you click out from Google you are first bouncing through Google's instant redirect before you are taken to the target website.
When users click the back button they bounce back through the redirect to the SERP.
You can test the tracking by right clicking and copy/pasting the URL of any Google search result. This can be stripped using add-ons (and I highly recommend it).
Yes, this was my understanding from a couple of years ago when I looked into it.
They know what time you clicked the link in the search results to go to page X. They also know when you hit the back button to come back to the search results.
My understanding was the faster you backed out meant the less-relevant the page was.
It was one of many factors in the ranking algorithm.
How does this track the "open a ton of tabs in the background from search and then cull them" workflow? Can they know which ones I actually spend time on vs eyeball and quickly close?
They can tell if you opened links in new tabs rather than by clicking on them sequentially, and can tell that you kept the search results live in its own tab for at least n seconds after opening tabs (x, y, z) in quick succession. Voodoo, err I mean ML, can deduce a lot from that.
It's really not. Google, famous for never giving users control over UX, has a full-blown option to persistently save your preference to always open search results in a new window/tab.
Google only knows how long people stay on the page using Google Analytics and many sites (mine, for instance) don't use it.
I think they also know if there are any ads hosted by google, or if the user is using chrome.
Depending on how moral they are, they could figure it out if the user is on android at all, or if the site has recaptcha, or if they're using google DNS, or if they have a partnership with the ISP, or if the site is loading fonts or javascript libraries from google servers, and probably some others that I can't think of off the top of my head.
In the original PageRank? No.
These days, who knows? But Google only knows how long people stay on the page using Google Analytics and many sites (mine, for instance) don't use it.
I am a quick reader and are very good at quickly skimming articles. I hope Google doesn't rank pages based on time spent, since Google would be optimizing their search away from my preferences.