Ever Feel that Usage is More Important Than Backlinks

[goodroi]

Last year I separately launched two new sites that I didn't actively try to rank in the search engines. My primary goal was to better serve existing users who were notified of the new sites via email. I added a very small handful of links and they weren't keyword rich seo links. These links were just to help the existing users more easily navigate to the new sites. After a few months the new sites started to rank for some decent keywords. They aren't high volume or very competitive but they are legit keywords (costs me more than $5/click in adwords) and the rankings have remained since last year.

When I encounter things like this it makes me think that traffic & usage is becoming more important than the backlinks when Google decides rankings. I would not go so far as to say that this is direct or indirect impact but there does seem to be some connection with usage & rankings.

[netmeg]

I don't have much in the way of legitimate backlinks (except for a few municipal .govs), and most of the ones I have from newspaper and tv station websites disappear after a few months, but I have a ton of traffic and usage. So ... I can't say you're wrong on that.

[martinibuster]

Yeah... I have a similar thing happen with 100% original non-researched articles, just original content, useful information. No link building because I forgot about it to focus attention elsewhere.

[Planet13]

I added a very small handful of links and they weren't keyword rich seo links.

Could you elaborate on those links?

Would google know that you had control of those links?

Were they on your own sites?

~~~~

I would say that we would be quite foolish to think that google is NOT using usage information as an increasingly significant part of the algorithim (or as part of the algorithm that makes up Panda or Penguin).

On the other hand...

Google seems to like one of my sites more and more as the bounce rate and pages per session get worse and worse.

And I have seen instances where pages that had an overly high click-through rate for their position (and were equal to the site's average metrics) LOSE ranking position, while some pages that had an underperforming click-through rate (and below average user metrics) rose in the SERPs.

So I would guess that there is some complex relationship or google is still in the early testing states or... something else?

[brotherhood of LAN]

I'd say if it isn't a big factor already, then common sense dictates that it eventually will be in most cases. A search engine being able to discern "is the user happy with the result of their query" with a high degree of confidence is job done, for them.

Of course it can be 'gamed' but it's obviously something that doesn't make much sense to game just to achieve better ranks, if it means the user isn't converting on your site. Piling in with fake positive user signals ends up swimming against the tide.

Recently Google showed their ability to measure human/bot differences in UI behaviour, meaning they already have ways to mitigate automating positive signals here. [googleonlinesecurity.blogspot.ca...]

It's an interesting topic, one that will result in a lot of new ghost profiles surfing the web IMHO.

[EditorialGuy]

I wonder how granular Google might get with usage tracking. We know that Google is moving toward having separate results for desktop and mobile, for example, but about tablets, which combine characteristics of desktop and mobile.

Take our site: In 2014, tablet users showed the lowest bounce rate, highest number of pages per session, and longest time on site. (Desktop was second, mobile was third.) If user satisfaction is a ranking factor, wouldn't it make sense for Google to rank us higher for tablet searches than for desktop or mobile searches? (I don't believe Google has ever said anything about "tablet search," but I could be wrong.)

[goodroi]

#1 - Usage and user satisfaction are totally different things

#2 - User satisfaction is not always defined by lowest bounce rate or higher pages per session or long time on site. Imagine someone searching Google for "what time is it in London right now?" If it takes a user 5 minutes and 10 pages to find the answer on your site, I doubt they had a satisfying experience.

[aristotle]

Years ago Google realized that it needed a better way to collect information on user behavior. That's why they expended a lot of money and effort to create their Chrome browser. By now, with tens of millions of Chrome users, they must have collected a gold mine of data.

I get a little annoyed every time someone starts a discussion about bounce rates, because it's been said here numerous times that bounce rate is an unreliable signal. I wish we could spend less time discussing bounce rate and more time discussing what I consider to be much better signals. For example:

-- The user scrolls slowly down the page.
-- The user bookmarks the page as a favorite.
-- The user prints out a copy of the page.
-- The user visits other pages on the site.
-- The user makes repeat visits to the site.

I've listed these signals several times in the past on other threads here, but no one seems to pay any attention. Instead they prefer to talk about bounce rate.

[brotherhood of LAN]

> #1 - Usage and user satisfaction are totally different things

Where do you draw the distinction? Are you just talking about page views, links clicked, and length of time on each page?

[Planet13]

I've listed these signals several times in the past on other threads here, but no one seems to pay any attention. Instead they prefer to talk about bounce rate.

Well, it may be that some of us are already taking these into account when calculating bounce rates since we use event tracking when someone scrolls down past a certain part of the page, or clicks on a print button on the page, or clicks on a popup to enlarge a photo on the page, or click on an affiliate link, or has other interactions which can be tracked using various events.

ASIDE: Anyone know how to measure time on page in GA by measuring how long a tab is in focus, or measure when the mouse leaves the tab and scrollbar area? I know there are all sorts of email subscription forms that popup once the mouse moves up to the various tabs or navigation bar.

[Whitey]

it makes me think that traffic & usage is becoming more important than the backlinks

No doubt about it IMO. Add to that some social signals and brand name searches and that's where it's headed.

Links cannot be a major long term support for Google's technology, with the predominance of mobile.

I'm intrigued, because you have identified your hunch more easily in a low volume / low competitive area, which gives your statement much credence.

[Planet13]

I would still like to know what kinds of links WERE built though...

[Zivush]

I can 'guess' that having tons of backlinks isn't important as having good traffic coming from many types of backlinks, which is a strong signal that a website worth something to users.

[Whitey]

I would still like to know what kinds of links WERE built though...

@Planet13 - are you kinda saying if they were "good" links , they should still be "gold"?

Agreed, if so. But links are increasingly hard to obtain, and Google has been experimenting with SERP's without links. Matt Cutts said the quality was terrible, but the drift of intention is clear. They have to find alternative solutions that better reflect popularity on the go forward.

[martinibuster]

Aristotle, I like those metrics a lot more than bounce rate. Thanks for sharing them. :)

[roshaoar]

Ditto. Another one I like to use myself:

- The user, who comes from from non-English speaking country, translates site using google translate

(yes I know this isn't universally important because sites have multilingual versions, but imo still an indicator).

And btw I've also noticed this phenomenon. Site gets a popularity run and rankings go up. Interesting article at [goralewicz.co...]

[edited by: brotherhood_of_LAN at 3:28 pm (utc) on Jan 6, 2015]
[edit reason] fixed link [/edit]

[goodroi]

The links were on a handful of relevant sites that I controlled. They weren't keyword rich or on very powerful sites but they were relevant sites and did have some link juice.

[aristotle]

Whitey wrote:
Google has been experimenting with SERP's without links. Matt Cutts said the quality was terrible

Well Google has been optimizing its algorithm around backlinks for more than 15 years. Obviously if you suddenly take backlinks out of the calculations, what's left of the algorithm can't be expected to give good results. The only way to reach a valid conclusion as to which is better, is to work on the new algorithm (without backlinks) for 15 years before making the comparison.

[EditorialGuy]

Bounce rate might be useful as one of many "user satisfaction" signals in a few cases. For example, if a site had a higher bounce rate for mobile users than for desktop users--in comparison to sites of the same type, and in combination with other signals--that might say something about its mobile usability. (In other words, it might help to confirm assumptions that Google has already made about the site's usefulness to mobile searchers).

The obvious problem with bounce rate as a metric is that there are many reasons why a user might bounce. Even behavior after bouncing back to search is of questionable value in some scenarios: Let's say that Joe User is searching e-commerce sites for a specific item such as "widgetco foie gras 3tb sata hard drive." If he's simply looking to compare prices and check whether the item is in stock on half a dozen sites, his bouncing back to search and going on to other results says nothing about the quality of the sites that he's bouncing from.

Metrics that I personally find intriguing (and which could be useful to a search engine) are the "engagement" numbers in Google Analytics. If a site has a decent number of users for 20+ pages or 30+ minutes in a session, that might be a positive signal, because at least some users are finding the site worthy of an extended visit. Again, such metrics have to be viewed in context: The aforementioned user who's looking for "widgetco fie gras 3tb sata hard drive" has no reason to look at more than a few pages on a site, no matter how great the site is, while the person who's looking for "neckties" or "history of the roman empire" could be expected to stick around a while if the site is any good.

[Jez123]

With this in mind has anyone tried "Crowdsearching"? It's a service whereby people search for your product or brand and spend time on your site. If you are correct goodroi then "crowdsearching" could be the next SEO! Doubtful I suppose. Has anyone tried it?

[toidi]

Why is it that when usage is discussed, the conversation alwsys turns to bounce rate?

Usage and bounce rate are 2 different metrics.