|Bounce Rate/Time On Site Manipulation & Well Ranking Sites|
| 5:34 pm on Jan 29, 2013 (gmt 0)|
I don't know if others are seeing this, but I am seeing a lot of sites wiggling up to the top spots that have a very similar strategy. They "promise" certain info with their title tag and SEO, but when you get to the site, the information that you actually need, you have to 'click' for more information - which displays part of what is actually promised. Then, to get the important information like a phone number, you have to click again, but often times, that isn't actually what is delivered - it's just a referral form. This is doing 2 things for the site - keeping people on the site (time on site) and lowering bounce rate - making people go around in circles to get information that should be displayed immediately. These are almost across the board the sites that are 'winning' in my niche - they don't even give the information promised...just keeps people trying to get to information that never really happens, or takes incredible digging to get to.
From a bounce rate/time-on-site perspective to a non-human, this probably looks like an engaging, sticky site - which, is anything but the truth.
Anyone else see this in their niche.? It's really frustrating to see so many sites rewarded handsomely for this.
| 10:28 pm on Jan 29, 2013 (gmt 0)|
I see this in my particular niche, we present the information on the landing page - which may cause a user to get what they want & bounce. Competitors often will hide the information & need an action to reveal it - sometimes opening in a new window (to avoid users going back to SERPS in the old one).
Not convinced Google treats this how I think they do, but it would make sense to say:
"If user x goes to site a then comes back to Search in some form & refines their query to go to site b then does no other queries" Site b could be more valuable for that given query.
I don't think Google would look at time on site or pages per visit as this can be easily gamed (unlike user intent).
| 11:09 pm on Jan 29, 2013 (gmt 0)|
Gyppo, thanks for your input - I appreciate your comment about going back to search results, and user intent. I hope that bounce rate and time on site for the types of sites I mentioned is not what they are being scored on - if that is the case, then we're all in trouble..
I am not really sure how bounce rate and time on site play into everything yet..
I have to wonder why so many sites are doing this - needing to do an action to reveal something like a phone number..which is basic stuff...what do those site think this will do for them - it's a horrible user experience in my view...especially when every single click to get promised information ends up at a referral form of some sort, without ever getting the information needed. Those are the sites that are driving me nuts, and I can't imagine why anyone would consider them 'good'
Thanks for your input!
| 12:10 am on Jan 30, 2013 (gmt 0)|
@besnette: it is a very interesting observation! I have been seeing this in the niches I work in pretty much since the beginning of the Panda era. All the biggest sites in the niche are doing it, and most of the top results in Google SERPs are landing on what appears to be just another set of search results, only this time all results belonging to one site.
Very often (I'd say more often than not) the site that gets #1-#3 in Google SERP does NOT have the answer that they promise in the deceptively worded title and meta description. And even then, it's been years since Matt Cutts proclaimed that Google does not want someone else's search results in theirs and yet this is all over the SERPs I see every day - not only what amounts to either doorway pages or search result pages, depending on how you want to see them, but even after you go through all the doorways, you still won't get the info that Google is sending you to find. And just like you said, you're almost guaranteed to click 2-3 times on various links on that site before you realize that it does not contain the info you were looking for. You're also almost guaranteed to click your "back" button two or three times (which gives the offending site two or three more ad views) before you come back to Google results.
This seems to work for everybody - the site gets 4-6 ad views out of thin air, and Google does not lose the visitor because they have to come back to SERPs anyway, since they could not find what they were looking for. This has become so prevalent among my competitors that I am seriously considering joining the party, just looking for a way to do it safe. There's just no other way: I cannot create content pages at the same rate they can create doorway pages. It's like 2004 all over again!
| 6:54 am on Jan 30, 2013 (gmt 0)|
Eric Lancheres (not sure if I'm allowed to quote names) said that reducing bounce rate down to 55% and below will most of the time see a recovery. It is nearly impossible to get close to that number for informational websites, unless you implement the strategy mentioned by the OP which is annoying.
| 8:41 am on Jan 30, 2013 (gmt 0)|
I'm seeing the tactic mentioned in the opening post working for rankings even if the pages with the promised content are not on the same domain.
There are 5 specific and independent tools being used in my niche, created by 5 different companies and hosted on 5 different domains. About a year ago one guy wrote an article about the subject of the tools and he links to all five of them within the article. His title is misleading, you end up thinking the page will have the tools you want, and he is undoubtedly bouncing visitors to these other sites BUT he is also ranking #1 for the keyword term people use to find these tools, ahead of all five companies.
My take is that linking to related content in general is a good idea, on your domain or not. If you need more incentive to link out I don't know what to tell you but in the bigger scale this topic is tightly knit with the one we're discussing here. Which comes first?
| 3:53 pm on Jan 30, 2013 (gmt 0)|
@1script - thank you for your insights...I am glad to see I am not the only person seeing this. Like you said, exactly - the top 3 or so sites are doing exactly this in my niche. Another tactic is "click here for the important information" like a phone number - but instead, a huge referral form comes up blocking everything else, and in tiny print at the bottom you can click out of it to finally be taken to the phone number. @raymondcc - 55% is a really tough task for most sites - I have a lot of informational pages with a high bounce rate, but people are on the page for 6-10 minutes, so I hope that is a ranking factor, not just bounce rate.
@sgt_kickaxe - I link out to trusted state resources from my site, just to help my visitors...and to provide a trusted option for information - and I hope it helps, and I know I bounce some visitors to these sites- but this is in addition to my own content, that does display the information promised, without going around in circles. My site still ranks decently, but seeing sites like we are talking about in the first places really makes me wonder that the hell a site owner is to do.
Some other common things I see with these types of sites is that they often do no social stuff...nowhere to be found on g+, very little FB activity, etc. I also see that, in my niche, several sites are .org, but they are not 'non-profit' but they are trying to play on that...which, I think makes them seem trustworthy. I don't know if .orgs still get more favor, but I do see many sites working this angle.
I also see on many of these types of site absolutely no authorship information, no hint as to what actual human being is behind the site, and no hint of what 'expert' is behind the content. Contact forms are generic, if they exist at all.
Another item that I have discovered is that a few of the top sites are owned by the same person or company, so it is all going into the same pocket.
These sites are often very pretty and 'well done' but if you look 'under the hood of intent' - it's often a dark neighborhood.
I'm not against doing referral activity or ads, but when that is the only obvious option - i.e. all clicks and activity funnel into referral stuff, and it takes 5 minutes and a dozen clicks to do it...I think it is a horrible user experience, and a bad representation for the top spots in the SERPS.
Thanks everyone for your input. I am trying to just stay the course, keep working on my site, producing original, straightforward stuff, as I've done for many years....hopefully in the long-term it will be the right strategy.
| 4:00 pm on Feb 10, 2013 (gmt 0)|
besnette, thank you for bringing this up. I'm seeing this too, as a searcher, and it frustrates me so much I've started just ignoring the top several results in searches to try to avoid these types of sites.
1script, is really IS like 2004 all over again. In fact, I've started talking to black hatters lately because it seems really clear to me that spam is still working.
| 4:50 pm on Feb 10, 2013 (gmt 0)|
Keyword stuffing, doorway pages, duplicate content; all of it is working for my competitors.
|it seems really clear to me that spam is still working |
Matt Cutts made a point of showing results for phone number searches [plus.google.com...] saying "it's not really useful and a bad user experience" - we search for phone numbers daily and saw a slight fluctuation in those results but they returned to normal a short time later with only the exact type of sites in his example in his post missing most of the time. The remaining sites aren't any different, just laid out differently and with larger budgets; the site that belongs to the phone number rarely shows up even when surrounded by quotation marks "like this".
We aren't in the phone number niche but we see the same techniques in ours. The example relates to the original post in that these are all useless sites unless you're willing to pay for the information but it will take average users a few clicks to find out they need their wallet.
| 4:54 pm on Feb 10, 2013 (gmt 0)|
I think this manipulation is a wrongful tactic that will not last for long.
Because search engines can easily measure "time on page" and "behavior in page", if they don't do it already.
Clinking on links shouldn't give any credit to the site if time on each page isn't long enough.
Think about that.
| 6:52 pm on Feb 10, 2013 (gmt 0)|
I think it does matter but I don't think it's nearly as simple as most people think and it's impact probably depends on the signals contained within the query.
Take three examples: "play space invaders", "distance to the moon" and "things to do in London". If you think about how long each of those might typically engage a user, they are very different. Combine the first and third with return visit stats, the second with follow-up actions on Google, the third with page views, etc, etc and maybe there are clear signals there.
So, going back to the OP, time on site and page visit stats may or may not provide additional info to Google and where they do, it may only be useful when combined with other signals.
The problem of the sites not providing the correct info rising to the top may be because the terms you chose to search on aren't condusive to user behaviour ranking signals, or maybe there is a temporary exploit available or it's simply that page views aren't influencing rankings on that query.
But whatever way, from what I see in my niche, I am inclined to believe user stats are in play for certaian queries myself.
| 7:27 pm on Feb 10, 2013 (gmt 0)|
"Manufacturing" this user experience metric is incredibly easy: make the next page as cumbersome and convoluted as the previous one and put anything that resembles a link to the resource the visitor is looking for at the very bottom (or below the fold, anyway). That'll get you a second a two extra to fool any automated filter. A human being would have spotted this instantly but Google never had and never will have enough human beings to check on everything. Besides, some of the landing pages are (or purport themselves to be) about very complex subject in which you actually have to have some kind of understanding to realize that you are being fooled.
|Clinking on links shouldn't give any credit to the site if time on each page isn't long enough. |
@Simsi: regarding the vague context of the original search term that may be playing a role in this. In my niche the most blatant offenders are sites that contain collections of datasheets on various parts and devices. The initial queries are very simple and cannot even be taken two ways: when you type "blue widget datasheet", I guarantee you 99% of the time you want to download PDF that contain the said datasheet.
What happens next is very indicative of the user metric abuse going on, as described in this thread. Many of the datasheets are not so easily available and pretty much noone has them - the manufacturer probably never made them available online. The more rare the datasheet, the more determined the searcher to find it and so you can have at least two or three pages where you will have lists of OTHER datasheets, sometimes related, sometimes not, and on the bottom you'd see a link saying something like "see other blue widget datasheets". Since you haven't gotten what you came for yet, you'd click there thinking their sorting is off and the one that you need is surely not far away, but no such luck. I can imagine some people looking for information on rare parts, could click much more than 2-3 pages, simply out of desperation.
So, there you have it: the bounce rate goes down simply by definition - they are still on the site. Number of pages opened is high and time on page is usually still decent - they are normally filled with pretty long lists of complicated part numbers - you will need at least several seconds to realize that what you are looking for is not there, and several seconds is a VERY decent on-page time by today's standards.
| 9:22 pm on Feb 10, 2013 (gmt 0)|
@1script: It's hard to guess just how complex Google's algo is obviously but many signals must combine with, and rely on, other signals to be triggered or to formulate.
It's possible that for your particular datasheet search example, Google doesn't regard page views as a logical signal and by extension, time on site increased by moving through pages might therefore become irrelevant. Speculation of course but in which case, the good rankings might be based on other factors.
I mean, if you say it's possible no-one has the answer to a query, then you could argue that no site in that niche is good and Google might - for all we know - realise there is no answer and just leave it to it's own devices! OK so that is extreme, but either way, someone has to rank top and if every site is useless, what does it matter to Google who does?
But that might be different in another niche/vertical. Which I guess is my point in relation to the opening question. These user behaviour signals might only be in use where Google knows they are strong indicators of quality.
Look, we all know that Google's algos are far from perfect. They know that too I'm sure. But they are constantly moving forward, even if that means the odd step back every now and again.
But one thing we can be sure of is that any signal that can be manipulated will be targetted sooner or later. SO if people are trying to manipulate page views and time on site at the detriment of user experience, it's a risky strategy that has a shelf-life and may even lead to bigger issues.
If my competitors were doing that, I might even be quietly optimistic, although I said that about link building 6 years ago and it's only now that the Penguin seems to have rewarded my patience lol.
| 9:54 am on Feb 11, 2013 (gmt 0)|
Average time and bouncing rate are not ranking factors but affect the relevance of a page.
For some sites the bouncing rate is more important (informational sites for example) and for other Google consider more important the average time (entertainment sites for example).
Depending on what niche you're in, manipulating this values with the right strategy can actually give you an edge and that's why you see many sites doing that.
I don't believe is the right approach for long term business but I've implemented it with success on one of my sites and if I had an informational blog -like some of you- I'll definitely split the post in two or more parts in order to reduce the bouncing rate.
| 10:27 am on Feb 11, 2013 (gmt 0)|
besnette - Yes, this is one of the most annoying types of sites to encounter... probably tricky to detect algorithmically... and yes, I've had the same thought you've had that they might be faking engagement.
I think they dropped for a while after the first Panda and above-the-fold roll-outs, as they were greedy about AdSense as well... but recently I've seen them come back.
The sites probably calibrated the ad threshold level enough that the above-the-fold algo is no longer hitting them as much. Probably a combination of signals is involved, and it may take some trial and error for Google to get rid of them. AI using heuristics requires trial and error.
I keep thinking that Google does not want to cut too deeply all at once, as that's more likely to produce collateral damage.
| 3:48 pm on Feb 11, 2013 (gmt 0)|
I'm not sure it's that easy to 'fake' engagement. From personal experience people only click on things if they really want to, so even if you try to 'fake' improved engagement by making people jump through more hoops, you'll only be successful if people really want what you appear to offer. If you tease people too much they'll lose interest and if you don't tease them well enough they won't be tempted to click.
In some cases attempting to fake engagement might actually inadvertently create a better user experience. Breaking up a blog article into several pages could actually make the whole article more digestible and less daunting, which is an improvement for the user.
I'm trying to improve engagement on my ecommerce site, I've tried adding competitions and other distractions to encourage more clicks but it's only things that help people achieve what they came to the site for which encourage more engagement. In other words, I couldn't fake it, I had to provide something useful to successfully get people more engaged.
| 5:52 pm on Feb 11, 2013 (gmt 0)|
|I'm not sure it's that easy to 'fake' engagement. From personal experience people only click on things if they really want to, so even if you try to 'fake' improved engagement by making people jump through more hoops, you'll only be successful if people really want what you appear to offer. If you tease people too much they'll lose interest and if you don't tease them well enough they won't be tempted to click. |
claarky, here here....haven't you noticed how a so called "new revenue model of the future" thread had drawn many people to engage on that thread...you can come out with similar stuff in most things...you just need to know the trick to control the herd.
| 9:17 pm on Feb 11, 2013 (gmt 0)|
@Simsi - well put - I do think these sites have a shelf life...I've just been noticing more of it in my niche lately..
My strategy will always be long-term, and it is definitely risky to model these types of sites.
@claaarky - I found myself again on one of these sites this morning..leading me on a 5+ click adventure to get to promised information that was to never be. I'm no dummy, but if a site promises more information, I would like to trust that if they are well ranked, I'll get the info.
You mention "Breaking up a blog article into several pages could actually make the whole article more digestible and less daunting, which is an improvement for the user. " - in some cases, yes - but to obscure a 10 digit phone number and force someone to click on it to 'get it' but instead get blasted by a huge referral form, and a barely visible link at the bottom of it to 'close out of it' to actually get the number is manipulative and unnecessary - and becoming a common practice. Or, listing other business names and addresses, but nothing else all over your page to 'fish' people into your site, and then 'click for more info' again coming up with 'submit your information' to find out more...is another manipulative tactic..
I think in a tech/internet/saavy niche, or website (like this one) - most of the time we can spot a questionable site. However, there are millions of people out there who are not tech/seo/internet superstars (seniors...for example) who often times place too much trust in what they come across on the web...
@Robert Charlton - yes, I think it is very tricky to detect these sites. I think we all agree that it is a risky thing to do. My hopes by posting this was to see if others were seeing this too - it's good to see I'm not alone or just 'seeing things' that aren't there.
I appreciate everyone's input and feedback. Interesting comments from all...!
| 3:31 am on Feb 12, 2013 (gmt 0)|
|but to obscure a 10 digit phone number |
There are several good reasons for which this might have been done. I could think of a couple of reasons on the fly.
1) To prevent bots stealing those numbers.
2) For some businesses, the revenue model might depend on such phone numbers. Take for example a real estate broker's website.
So there are also valid reasons for which one might want to obscure the phon numbers.
|and yes, I've had the same thought you've had that they might be faking engagement. |
There is a subtle difference between saying "faking engagement" and "faking things to draw real engagement".
Engagement might in most cases be real. But some people fake things to get people engaged before getting the content they want. So a site might still have useful content a.k.a content for which the user might have landed on it. It is just that they get people to play around before getting the needed info. But this isn't easy. One need to draw the user's attention and make him/her act. It shouldn't end up making them bounce back.
So there are two types of sites.
1) sites which have the useful information but fake things to get people engaged.
2 Sites which don't have any useful information but fake things to get people engaged.
What is really annoying is the latter type of sites ranking high.Yes, for many, even the former is annoying but it might be that Google does not really care about it.
| 9:44 am on Feb 12, 2013 (gmt 0)|
I think Google would be able to identify many of the number 2 cases. On-site engagement is not the end of the story.
If the visitor goes back to Google and visits another site for the same thing, it suggests the previous site didn't deliver. I think the sites that engage AND deliver will eventually win in the rankings and be able to maintain rankings over the long term.
I can see that a site that doesn't deliver but is able to engage visitors, say, 4 times more than any competing site might win in the rankings purely on engagement, but any site that can do that must have something major going for it which can't be faked (like a great brand name). You can't fool all the people all the time.
| 5:26 pm on Feb 13, 2013 (gmt 0)|
I agree @indyrank that not all sites are 'bad' that use some of these tactics. It all comes down to intent, and to me, looking at a site that does all of this, but has no 'authorship' associated with it, no expertise, no hint of who is behind the site, no real way to contact the site owners or figure out who they are - just a purchased database of information that's found 100 times over elsewhere, with little original content - that's what becomes a bummer to the rest of us who work on being a 'real' authority in our niches.