Forum Moderators: open
How do sites get top rankings with a low PR?
Is Google not using PR now as a prominent factor for SE rankings...just mostly good SEO of web pages?
Thanks
[edited by: engine at 1:30 pm (utc) on June 8, 2004]
[edit reason] specifics [/edit]
That would be a gold mine if it was easy :)
Write some content, design your pages wisely, connect them together even better, and soon you will have links to your site, and all would be better ( PR, rankings, visits, everything ) :)
"How do sites get top rankings with a low PR?"
Plenty of anchor text inbounds (keyword and related semantics) from PR1 and PR2 sites is often enough to do the job.
TJ
Now that I do, I realise the only importance of PR is when asking other webmasters for links on all my new sites that are currently taking ages to acquire their first PR at all.
GranPops
It is very much a question of "never mind the quantity feel the quality."
Have managed to get 762 non reciprocal backlinks on one site, with only one outgoing, and it has achieved a magnificent PR4.
As all the search terms were No.1 without backlinks, it matters not, but it illustrates the point made by many of the calibre of the link.
Thanks
Now I see some sites have 18 backlink and quoted PR7!
1 decent PR8 backlink will do that on it's own.
TJ
1 decent PR8 backlink will do that on it's own.
I have looked at it.
The site I am talking has 2 outbound links with PR4 and 14 inbound links.
Is it still possible PR7?
I have looked at it.
Stop looking at it and worry about your own site.
The site I am talking has ........ 14 inbound links.
As far as you can tell, or as far as google is showing you.
You could do with some reading here through the google forums. Spend a day or two reading some of the posts, I'm sure all this will become clearer to you.
Backlinks and PR are generally out of sync. They update to their own schedules. Of the 14 inbound links, if one is a PR8, there's your answer. 14 high PR7 inbounds would also do the trick.
Do some reading though, I would suggest you could do with a primer on the importance of ignoring PR ;-)
TJ
The site I am talking has ........ 14 inbound links
The page you are talking about has 14 inbound links, not the site.
TJ
One issue is, when linking to your own pages internally, should you use the same anchor text or vary it?
The other issue is, when requesting links from other webmasters, should you try to convince them to use varied anchor text or not?
On the first issue, I'm with steveb on this one. When designing a site, the natural thing to do is to make a navbar or menu or something which goes on all pages, or at least all pages of one category, such that all internal links to a given page have the same text. To do otherwise would be extra work, as you would have to modify each page individually, instead of just reusing the same navbar code. So why would you do it? The only reason would be SEO. It's not natural. Also, from a UI perspective, it would be confusing if the anchor text in a navbar kept changing. That would be just silly.
(edit: for example, on this site, the links to the google news forum use the AT "google news", amazingly enough, as well they should, since to call it anything else would be confusing to everyone)
On the second issue I really don't know. On the one hand, clearly most natural links will vary. But, as has been asked, what if your page title is your keyword phrase and is the most natural phrase to use to link to you? What if it is the name of your organization? If you are called "fuzzy blue widgets," wouldn't any sane person link to you with the text "fuzzy blue widgets?" For other cases, of course, the text of inbounds should obviously vary unless you are feeding all your link partners the same text deliberately.
Would you seniors who have commented on this mind clarifying your positions with respect to this distinction?
People have different opinions, but natural linking is the only thing that makes sense to me, and it works great. Artifically varying your links has to be seen as a signal of spam. It's not useful to users and it isn't even helpful with engines. Your internal navigation should be consistent, helpful and accurate for your users. If a page is about multiple things, you will end up with multiple link text. But if your page is only about red widgets, there is no value at all in naming some links "page three". Don't fall in the spammers trap. Construct your site sensibly and descriptively. Users and bots will be glad you did.
Some folks here are trying to put a dress and nylons on a pig. The problem isn't the dress. If a bunch of links all with the same text is a problem, its part of an equation not a cause and effect. Don't send out spammy signals. In some situations all links being the same could be a spammy signal, in other cases artifically varied links could be a spammy signal.
Personally I think it all about going to the root of the matter. Make pages about stuff that will naturally be linked to by helpful words. Don't try and force a pig into a garter. It isn't natural, and if a bot comes upon that pig in a garter, it isn't going to like what it sees.
Some of us (including me) had suspicions of Google looking at 'too similar internal anchor text' for a while. The idea doesn't seem to make so much sense now (keep in mind those were fraught times for some people!), and Matt Cutts seemed pretty clear that is wasn't an issue a couple of weeks ago when he was at PubConference and SES.
> what if your page title is your keyword phrase and is the most natural phrase to use to link to you?
I thought the same when some clever people talked about this some time ago. On looking at random-ish samples of backlinks from major brands I was astounded. If you or I were to link, say to the BBC, we'd almost certainly do it like this: BBC [bbc.co.uk]. Yet many people link to click here [bbc.co.uk], here [bbc.co.uk], site [bbc.co.uk], Web site [bbc.co.uk], website [bbc.co.uk], etc.
If you put up a new site and get 50 back links straight away, all with "fuzzy blue widgets" as the anchor text I think you'd stick out like a sore thumb.
steveb, I like the sentiment but search engines aren't all-knowing beings coming to intuitively rational conclusions. Their spam filters may consist of highly complex rules, but they're still just rules.
It is useful to think about search engines, and how they respond to what we do. Remember the famous "PR0" penalty for heavy cross-linking? Many spammy groups of sites were caught, but some entirely sensible groups of sites were hammered and would not have been had the owners seen what was coming...
Matt may have been thinking of "normal" non-spammy keyword usage in global navigation at the time, but lets take it a step further, step by step:.
1. Side navigation column includes link to homepage (which it should), which is a graphic with the main keyword-phrase in the alt attribute.
2. Bottom text navigation for the convenience of users when they're scrolled down includes "keyword-phrase home" in the homepage link.
3. The top header graphic links back to the homepage globally (many people expect that graphic to be a link) with the keyword-phrase in the alt attribute.
4. For good measure, throw in some inline links with the keyword back to the homepage within body text in several places across the site. Maybe a ton of them if someone is getting really zealous in their quest for the top spot for allinanchor:
5. The domain name and site name include the main keyword-phrase, so of course inbound links and directory listings will include it.
That is how a site can get to be in the top 5 for allinanchor: out of many thousands of pages returned for the search with maybe 4 or 5 inbound links altogether.
Many spammy groups of sites were caught, but some entirely sensible groups of sites were hammered and would not have been had the owners seen what was coming...
Exactly the point, and it goes along with the territory. Even though some may think it shouldn't be that way because cfertain things are *normal* in their view, that's the way it is - things are how they are whether they agree with it or not and whether they like it or not.
We can't ever tell people something is OK because we think it's OK unless we're willing to bear responsibility for misleading them into thinking they'll be walking on safe ground even though there are land mines out there just waiting to be stepped on.
"Natural looking" implants are NOT natural.
When Googlebot does "hands on" inspection/crawls there is a very good chance the deception will be noticed as not natural, which will make one step back and wonder "gee, besides implants, what other deceptive plastic surgery has been done here?"
There is a very fundemental logical flaw being presented here. Pretending to be natural is not natural, and is in fact a major warning sign that there is more un-naturalness around (even if there isn't). Artifically varying anchor text is an enormous red flag for spam, and obviously, an accurate one too.
If you want to look natural don't get implants. If you do get implants, don't pretend you don't have them. Once you are discovered the rest of you will lose credibility.
When Googlebot does "hands on" inspection/crawl
Right, so Google look at the page by hand and say "hey, he's trying to get to the top by making his site more attractive to our algo - lets ban him"
Nope
This is an automated issue.
Pretending to be natural is not natural
If you do it right it is undetectable - depends how clever you are.
How many external and inbound links are there pointing to the entirety of Google forum that are the exact same as that; and what is the percentage of all put together that are identical?
It's all balance and percentages.
>natural
It's a matter of bot intelligence. If it runs into a 40DD, it makes little difference whether it's falsies or natural - the result will be the same.
The most serious and dangerous logical flaw I can think of would be considering it a service to encourage and be letting people think because something is natural it will escape being penalized if it's over the top or runs into a filter.
It is a much, much greater risk to be deliberately deceptive, especially from a fear that be genuine will be seen as ungenuine. Advocating this unatural linking is many times riskier, in a reckless way.
If you are cleverly deceptive, you might fool the bots for awhile... just like all the whiners here who complain about penalties and bans that they deliberately solicited through unecessary practices.
If you are out to cover one bit of deception with another bit of deception, then you have gone done that road and it doesn't matter. If you are doing something over the top, THAT is the problem. If you have a clean site, you have nothing whatever to worry about, beyond the engines sometimes making a mistake, which they can do in any case no matter what you do.
Not the point. The linking to this forum is natural, and it ranks well. There aren't a bunch of deliberately false links out there trying to trick the bot by "varying" the link text. Changing 10% of the links on this domain to "News about Google" would be a clear signal of weirdness, and would be absolutely useless in any positive way.
Balance and percentages come natural. Trying to pretend something is legitimate when it is not is just that, pretending. Just do it right. Don't pretend to do it right.
You don't get it
All SEO is 'wrong' because it seeks to manipulate the rankings
No whitehat or blackhat, just shades of grey and degrees of risk.
Your 'right' might be 'wrong' according to someone else.
If you swallow the 'right and wrong' dished out by Google you are just jumping aboard the Google Cluetrain.
You can follow Google's guidelines exactly and still trip filters.
Good SEOs don't whine, they manage the risks and play the numbers game.
It is the point. In fact, it quite emphatically proves the point.
The issue breaks into two distinct camps. One is a concept apparently important to people with something to hide. News flash, the problem is what you are attempting to hide, not a tangent.
For people with nothing to hide, there is no reason at all to use deception. Sites with natural linking rule the serps, even the pretenders aknowledge that.
The point here is that natural linking is good. Pretending to have natural linking is simply pretending. And dangerous. If you want to do that, or think you need to, fine, go ahead. But it is totally wrong to suggest that people need to do something deceptive instead of what is natural, effective and user friendly.
And most important, if anyone reading this has less than 100 links, don't waste one second of your life making your links unnatural.
"In fact, it quite emphatically proves the point."
Well, it proves my point, so cool.
Um, sorry but I do get it. I do get that people who don't have the goods want to pretend they do.
That just confirms how little of 'it' that you 'get'.
There are members here who could 'pretend' to get natural links without (even) you being able to spot it.
Its about getting results.
If you are any good at SEO, you do what is required to get results for your client within the risk limits discussed and agreed in advance. If you don't, you are letting your client down.
If you like playing Golf with only one club, thats up to you, but the professionals use a full set.
..or maybe you want to re-define what you mean by 'pretend natural' linking.
Sites with natural linking rule the serps, even the pretenders aknowledge that.
Ummm... no.
Maybe in some industries they do. But in some sectors artifical linking does work. The question is, can you make a buck between now and the time that Google better filters out these artificial links?
In many cases, the answer is YES!
Imagine plotting a graph of the keywords/phrases used as anchor text for a site, as a "cloud" surrounding the primary theme of the site. So "racing", "drag racing" "racing reports" "NASCAR reports" would be dots in a cloud around the primary cluster representing "car racing" or "auto racing". A website on-theme for auto racing may link to multiple clouds including this one and related neighbor keyword clouds (ticketmaster, Fox sports network, etc. will be related, but in different clouds than core car racing keywords).
Now plot your own mini-net that uses the same anchor text from 500 sites. It appears as nearly dot, with an area of near unity or impulse.
Your site is not natural.
With statistical analysis of weblink data, numbers are applied to things like this to form a score, and thresholds are applied to the score to separate out suspect sites (or weight them based on classifications).
I have no knowledge that thi sis being done, but if I had the time to put onto the task of weighting natural backlinks more than obligatory or mandatory backlinks that result from hit counters, awards programs, and mini-nets, this is what I would do.
I suspect very true. However, where there is a will there is a way, and there is a way to detect automated structure no matter how cleverly done.
I doubt very much Google does this, but if you research cyclostationarity you will see that a machine/algorithm used to generate cleverly varying anything actually imparts a statistical signature upon the variations it inserts. They vary according to a pattern, and if there are enough data points to analyze cyclostationary analysis will reveal them to be programmed (and therefore, not natural) :-)
Used on military radio systems to figure out which brand of radio generated a detected signal, for example.
(Maybe this post should be in foo)
You seem confused about the topic. Results of effectively tricking the bots isn't the issue. Do whatever junk you want. If we are going to talk about results then we need to talk about first building quality websites.
The principal issue that I'm addressing, and again, you can talk about whatever you want, is that natural linking is good and works, and is less risky. A secondary issue is that articicially varying anchor text or doing anything else unnatural (including in some cases having all the exact same links) is a negative signal. If you manage to hide that, fine, but the point is, you are hiding it. Don't delude yourself about that. Take whatever risks you want. That is your business. But it is better to understand what you are doing and the risks you are taking.
Linking to your pages naturally and appropriately is the least risky way to go, despite their being risks in every aspect of life and webmastering. Whether there are ways to be more effective sometimes while taking greater risks, that could make a different thread (or different off-topic offshoot to a thread like this).
How likely do you think it is that 500 websites will link to one page/site with the same anchor text.
Whew. Someone better go tell all those CMS's this. Because almost every site using one has a box called
LINK TO US
Which GIVES the user the link to highlight, cut and paste.
Which means they all use identical anchor text.
And its not manipulation.
Of course. Many people hide for many reasons. That is one reason why some people wouldn't dream of putting a site in their profile and others don't give a hoot about staying in hiding.
It doesn't have to be deliberate manipulation. Unfortunately, if you are not a duck but you put on a duck suit, waddle like a duck and start quacking like a duck, if they are out there shooting ducks you will get shot.
Ordinary, non-SEO type people do "natural" things every day of the week that are exactly what spammers get bombed out for. And they often get hit just the same for it. Mom & pop are spammers, and that's the truth.
The first and most important job in ethical SEO is to determine what they're out their hunting for with loaded rifles and make sure to stay out of shooting range.
