| This 56 message thread spans 2 pages: 56 (  2 ) > > || |
|Google Should Stop Displaying Toolbar PR - Part 2|
< continued from: [webmasterworld.com...] >
|How it is possible for inaccurate information to accurately indicating the importance of a page (or anything)? |
Again, most inaccuracies will be invisible because the toolbar gauge has such broad increments. To use the fuel-tank analogy again, if the gauge shows the tank being a quarter-full, it doesn't matter if the actual amount is .20, .25, or .29 full, because the gauge is meant to show (and does show) only an approximate amount.
The only time the toolbar gauge's possible inaccuracy comes into play is when the true PageRank crosses the line from 6.00 to 5.99, from 3.9 to 4.1, etc. and the toolbar is delayed in catching up. And that's likely to be an infrequent event that doesn't affect most pages most of the time.
[edited by: tedster at 7:52 pm (utc) on Aug. 14, 2007]
There's also the inaccurate PR zero reading for home pages that have recently had a PR6 or whatever. When rankings don't fall, and traffic doesn't fall, then that new PR0 is not an accurate report, as Adam_Lasnik has acknowledged.
We're not publishing the majority of the pots that are coming in about this - we are referring those posters to the existing thread in the Hot Topics [webmasterworld.com], which is always pinned to the top of this forum's index page.
Looking back, such reports were part of my motivation for beginning this thread. This kind of thing has been going on for months, and I will be relieved if that false PR0 bug gets fixed.
|To use the fuel-tank analogy again, if the gauge shows the tank being a quarter-full, it doesn't matter if the actual amount is .20, .25, or .29 full, because the gauge is meant to show (and does show) only an approximate amount. |
That example implies the only problem is a lack of precision. But most of the complaints being raised in this thread are about situations where the TBPR is not only imprecise, but it is also inaccurate, and that the lack of accuracy is not readily apparent to toolbar users.
It's as if you were used to driving your car, and 99% of the time the fuel gauge provides you with a rough feel for how soon you need to get gas. But every once in a while, and without any warning, it suddenly is completely misleading. The gauge says your tank is roughly half-full, and moments later your tank runs dry. Or, it says your tank is empty, and yet you can drive for hours.
That's annoying enough if it happens despite the gauge manufacturer's best intentions. It's downright irritating if the random inaccuracy is introduced intentionally, as some posters suspect (e.g. to keep spammers off balance).
I understand the argument for why it would be more ethical to completely eliminate display of TBPR than to knowingly publish data that is intentionally made inaccurate. But personally I'd rather have a few convenient, always imprecise, sometimes wildly inaccurate clues to how Google views a page, or a website, than to not have any clues at all.
One of the main benefits of TBPR from my perspective is that it helps reduce the risk of accidentally linking into a "bad neighborhood."
|I'd love to know what feature you [i]would[/b] like to see. |
- Would have to provide actionable info for webmasters
- Would need to be useful and interesting for the ~99.9% of users who aren't webmasters
In terms of actionable info for webmasters, the most useful information provides an indication of whether Google "trusts" a site, or whether it considers it to be part of a "bad neighborhood".
Given the huge downside risks associated with linking into a "bad neighborhood" or inadvertently participating in a "linking scheme", webmasters need any clues we can get to help us stay out of those traps.
TBPR currently provides some clues in this regard. But, Google could more directly report data that suggests how well established and trusted a site is -- perhaps using a High Medium Low schema. If Google is reluctant to be quite that explicit in telling us what it thinks about our sites. ;), perhaps it could at least provide us with some hints, so that we can better judge the situation for ourselves.
For instance, Google could report whether or not a site is signed up with Google's Webmaster Tools, and it could establish multiple levels of certification that webmasters could opt into.
That would create an incentive for sites to provide proof of ownership, disclosure of affiliated sites, and other information that is useful to Google and helpful in proving their good faith intent to abide by Google's guidelines. Properly constructed, a multi-level voluntary certification process would be hard for spammers to game, yet allow normal sites to pass with flying colors and minimal effort.
This would also provide useful and interesting information for the 99.9% of users who aren't webmasters -- in the same way consumers sometimes find it useful to know whether or not a particular business chooses to be a member of the chamber of commerce, local retail federation, or local better business bureau. And, unlike TBPR, it wouldn't encourage people to chase after PR via link buying.
I bowed out, so this is only a pseudo post:
|To use the fuel-tank analogy again, if the gauge shows the tank being a quarter-full, it doesn't matter if the actual amount is .20, .25, or .29 full, because the gauge is meant to show (and does show) only an approximate amount. |
The gas gauge is giving you an approximate reading from 3 months ago.
You don't know who's used the car, how long it's been driven, who's been driving it, etc. All you know is 3 months ago the tank was approx. .25 full... How much gas do you have today? There is no way of knowing.
And, the false-empty-indicator, as Tedster pointed out, could very easily skew the perception of visitors into thinking the car was misused and 'run out of gas' months ago, which could make people much more skeptical about driving it.
[edited by: tedster at 7:59 pm (utc) on Aug. 14, 2007]
[edit reason] by request [/edit]
|Am I alone in believing that most users don't know or care about toolbar PR? Is there anyone who wouldn't order HARRY POTTER AND THE DEATHLY HALLOWS from Amazon because the Amazon page for that book isn't displaying any PR in the Google toolbar? |
Most users know and trust Amazon so no, I don't think it would deter them from buying or to feel insecure about providing personal information.
|The issue isn't with Google or Knowledgeable Webmasters... It's the 99.9% of people who install it and have no idea it's not accurate, and have no idea they need to look to see if it's accurate or not. |
Precisely, and I don't think they have a clue as to what it means. It would be naive, IMO, to believe that users don't hover over the green bar and wonder what this means: "PageRank is Google's measure of the importance of this page." If the all-powerful Google doesn't believe a page is important, why should the average user? Is the site not to be trusted, or what?
|- Would need to be useful and interesting for the ~99.9% of users who aren't webmasters |
I don't understand how it is useful and interesting to these users in its current state. It doesn't disclose to users what Google deems "important". I see it as possibly casting a negative light and a certain amount of distrust for lesser-known and newer sites. Ecommerce sites have a particularly difficult task with instilling confidence in their users and anything Google can do to clarify the "importance" statement would be a step in the right direction.
Conversion is the responsibility of the ecommerce webmaster but why unnecessarily present what might be a negative factor, or a blown-out-of-proportion positive factor from a highly-regarded source such as Google that these users don't or can't understand or glean any useful information from?
I don't care if the green bar stays or goes, but if it stays, more needs to be done for the ~99.9% looking for useful information.
|But, Google could more directly report data that suggests how well established and trusted a site is -- perhaps using a High Medium Low schema. |
Interesting idea, though a "trust rating" indicator could be more confusing and potentially misleading than the toolbar gauge, especially when a user is looking at a new site that doesn't have a track record.
|Google could report whether or not a site is signed up with Google's Webmaster Tools, and it could establish multiple levels of certification that webmasters could opt into. |
I doubt if Google has the time or inclination to get into the certification game; as for whether a site is signed up with the Webmaster Tools, I can see two problem--one minor, one major:
MINOR PROBLEM: Webmaster World's server would crash under the weight of "Google is a monopoly that's trying to dominate the world by forcing everyone to use its Webmaster tools" threads.
MAJOR PROBLEM: Many (most?) site owners have never heard of Google's Webmaster Tools and wouldn't care about signing up for them if they had.
Sites that don't submit any information to Google, or sign up for their Webmaster Tools can still be "verified" to a limited degree -- for instance Google can evaluate how long the site has been on the web, it can check to see if the Whois Data matches a real person or business listed in the phone book, and so forth.
There are plenty of other bits of data it could use, as well -- such as frequency of spam reports against the site -- but only considering spam reports that were filed by trusted sources via Webmaster Tools, or those that were subsequently confirmed using automated benchmarking, etc.
(I'm assuming Google is already doing these sorts of evaluations -- my only point is that it could use some of that internal data to "seed" the certification process even for sites that don't voluntarily submit to additional scrutiny).
But, the real point is that this provides a "win win" mechanism for non-spamming sites and Google. It's the same principle behind the Webmaster Tools initiative -- Google provides sites with some valuable information, and the participating sites provide Google with some valuable information (e.g. identifying common ownership across multiple sites).
I can't say whether Google would be interested in a "certification" process, but it isn't a completely far-fetched idea. Microsoft and Sun have profitably pursued that concept with respect to training and certifying software professionals -- its a "win win" for the participating professionals, and for the firm that performs the certifications.
The positive thing about this approach is that it is more carrot and less stick -- encouraging the "good guys" to help Google improve their SERPs, rather than only focusing on punishing the "bad guys."
As for accuracy, I dare raise the question whether this is solely googles part.
I think it was Adam, who - a year ago or so - pointed out about some very basic problems google is faced with during the crawling process: broken links, illformed websites down to imparseability, timeouts, infinite 30x-loops and so on. All of us, who have ever made some attempts to automatically crawl or analyse unknown websites for whatever purpose, will know what this means. It's a bit like with employees: any mistake that might possibly be made, will be made.
I believe quite a number of the webmasers contributing here would at least in principle have the capacity to crawl the web and calculate PR the way google does (and if none of you, some industries definitely would). So, apart from copyright issues: Why isn't there any tool out there showing accurate pagerank? Anyone could make millions with such a tool?
Because already in the crawling process there are a large number of ad-hoc-decisions to be made, concerning what of all that crap out there is really worth being swallowed. And the fine-tuning of these decisions is a secret born from googles very own experiences. It is as imperfect as the underlying data.
Given that any such irregularity might have huge effects in the iteration-process of the PR-calculation-algo, it is rather a miracle that the toolbar shows roughly consistent results at all.
Yeah but most professional webmasters are curious of *those decisions* made ( not just the ones on spot, the ones afterwards too ).
And to the public, the few users with PR on, the many that have the Toolbar on, to the inexperienced, to the newbie webmasters too, such info is far more valuable than trying to guess actual PR down to .00001 .
It's a -1 / 0 /? / 1 issue.
WARNING / none / don't know / OK
The rest I don't care for, and neither should anyone.
Not because the PR bar is lagging, but because PR shows volume, not quality. A low PR site can outrank a high PR site, everyone knows that.
If Google showed the difference between *banned* and new sites, and didn't show the PR 0 warning for completely OK sites for months... and for no good reason, I don't think I had anything else to say.
PR 4 or 5, 7 or 8, such differences rarely matter in my opinion. But why can't there be some kind of an indication, either on the PR bar or perhaps completely separate to offset its disinfo, that a site is NOT banned, it's not SPAM, malware or spyware... it's just *new* or the Google Toolbar Query servers somehow messed up its data and it's actually pretty old and trusted.
Perhaps a new, simplified - additional - gauge that displays such information, nothing more or less that anyone could query with site: info: cache: and link: ( yeah even with them more or less broken )... is this site OK? Banned? New? Does anyone, even a single site link to it? A true/false indicator, 4 bits. No, make that 2.
That's all that's really important, the rest can be kept 3 months lagging behind, right? This information, unless it'd be used not only for bans but also penalties, could not be used to reverse engineer anything. But it would be useful for all users, webmasters and the public alike. And dismiss any misunderstandings a faulty PR indication could cause.
Perhaps PageRank should be available only as an add-on and not a standard feature of the toolbar...along with a brief explanation as to what it means in the event non-webmasters stumble upon it and add it (e.g., it has nothing to do with a site's reputation for doing business, etc.)
Useful and entertaining things that should (IMO) be standard on the toolbar:
An icon indicating the site accepts Google Checkout, along with an easy way for users to sign up.
A button/icon to add to Google Bookmarks and a quick and easy way to add a site's feed to Google Reader. IOW the ability to add the feed to Reader without leaving the site.
A separate field for a general product search, and possibly another for blog search, or at least icons to easily access these general searches.
I realize that similar things can already be added to the toolbar but I don't believe the average user is aware of it, or would think to look for it. Better marketing/awareness needed if these basic useful and entertaining features cannot be standard. Yahoo has gained a lead in user satisfaction as a portal and I think it is because too few users are aware of Google features and gadgets and they are currently too cumbersome to access and use.
Oh wow, I was just looking at the Google web site here: [google.com...]
and saw these statements:
|PageRank Display |
Wondering whether a new website is worth your time? Use the Toolbar's PageRankô display to tell you how Google assesses the importance of the page you're viewing.
Is the site worth your time? What?
For those who don't believe PR may discourage or turn away some of your users, you may want to re-evaluate.
How did that language get past Google's legal department?
|Wondering whether a new website is worth your time? Use the Toolbar's PageRankô display to tell you how Google assesses the importance of the page you're viewing. |
Well, that pretty much seals it.
Again I suggest, move the PR gauge off the user toolbar as a default feature (make it available as a special download to people who do want it, where its relevancy could be adequately explained); and add a section to Webmaster Tools where siteowners/webmasters could check the PR of any external page, if they were considering a link agreement or had some other reason for needing to know its status.
The gauge remains, but would be mostly utilized by people who knew what it meant, and who could keep it's numeric values in perspective.
That's not a bad idea:
- If Google regards the PageRank gauge more as a Webmaster tool than as a tool for users;
- If Google is confident that the toolbar will continue to attract users without the PR gauge.
Ultimately, any change has to work to Google's benefit (or at least not to work against Google's interests).
It's optional as it is.
Futhermore it's not turned on by default.
At least it wasn't the last time I installed it.
I find it incomprehensible that Google is telling users that the PR display determines if a page is worth their time. There is nothing useful or entertaining about this claim. Plus it just gives webmasters another reason to manipulate PageRank.
|Plus it just gives webmasters another reason to manipulate PageRank. |
I doubt if many of the manipulators need another reason to manipulate PageRank.
|I find it incomprehensible that Google is telling users that the PR display determines if a page is worth their time. |
It's an extremely bad choice of words, and it sounds more like breezy marketing copy than anything that someone who'd thought much about PageRank or the Google algo would put up. That said, it's not a reason to dump the PR display... but it should serve as a strong indicator to Google that parts of their website and documentation need some work.
I'd like to see the Toolbar display updated monthly, but I can understand why Google may not want to do that. I do think that if Google doesn't keep it updated, Google should provide some guidelines to users which contain some caveats. Ditto with the link: operator. It's just plain silly to have discussions about the omissions of these tools go on for years... and, as some have pointed out, it's potentially dangerous to have this incomplete data guide some major decisions.
The documentation should be clearly referenced, btw... not buried in the Google site somewhere. In the case of the link operator, some clarification that these are only a sampling of backlinks should come up with the link operator results on the serps page.
I imagine that Google can find an equally prominent spot for a few important comments about Toolbar PageRank, how fresh it is, and how it's not an indication by itself of how a page might rank for a query. That word "rank" confuses a lot of people.
FWIW, my thoughts on TBPR are on this thread, and I think it's pointless to repeat them again...
Toolbar PR. I'm confused?
Is it or Isn't it of any value?
PageRank is an vital measure of page importance. Without this you are blind to a page's popularity which is a central aspect to Google's social interaction (social feature). It is a way to not only judge publishers but also the validity of retailers.
I would not buy from PR0 sites as they could not be trusted.
If Google take this away they are also taking away an important measure of their search engine for us users.
This will cause an immediate shift as to where people will look for importance. Away from Google that is.
This is also a very lazy way of getting rid of link buying instead of continuing to work on the algo to improve quality.
If the intention is to kill 'natural search' then I suggest go ahead guys. Cannot believe you are even considering it.
|This will cause an immediate shift as to where people will look for importance. Away from Google that is. |
I think the question remains about how many of the "99.9%" that Adam referenced actually use the PR bar and check it when they enter a new page. If there have been any studies about that, then I'm not familiar with the results.
It would be interesting if an icon for the PageRank gauge loaded with the toolbar, rather than the current green one. If the person wanted to see the numeric PR value, they would click the icon one time and the number would display. In this manner, Google could keep an exact record as to how many "requests for PR" they got on a daily basis, which would help explain its true popularity amongst the web surfing public.
|I doubt if many of the manipulators need another reason to manipulate PageRank. |
It is additional incentive for webmasters with a need to convert visitors into users and buyers. If the goal is to get visitors to click on ads (which does not equate to real users), low to mid-range PR is unlikely to deter them from doing that.
|It is a way to not only judge publishers but also the validity of retailers. |
I would not buy from PR0 sites as they could not be trusted.
OK, so the fact that other sites don't link to it means it is an untrustworthy and invalid business. I guess a PR 1-7 wouldn't be good enough either using PageRank to determine trustworthiness. If PR equals trust, I wouldn't risk buying from less than an 8. A PR5 would be out of the question as I'd have a 50/50 chance of getting ripped off. So if the new PR0 webmaster manages to buy enough links to raise the PR to at least an 8, this makes it a trustworthy and valid business?
It is mind-boggling that users determine whether a business is trustworthy and legitimate based on the number of sites linking to it. But why wouldn't they...that's what Google is telling them. Never mind wasting your time looking for real signals of quality and legitimacy before coming to that conclusion.
This thread has been a real eye opener and I personally never imagined that what Google considers entertainment is yet one more hurdle in converting visitors into users.
|It would be interesting if an icon for the PageRank gauge loaded with the toolbar, rather than the current green one. If the person wanted to see the numeric PR value, they would click the icon one time and the number would display. In this manner, Google could keep an exact record as to how many "requests for PR" they got on a daily basis, which would help explain its true popularity amongst the web surfing public. |
That's a great idea!
By the way, the last few posts have talked about e-commerce businesses, but it's worth noting that providing credibility for e-commerce businesses isn't part of Google Search's stated mission--or the toolbar's stated mission, for that matter, which is to serve as an indicator of "the importance of the page" (presumably as a source of information), not of whether a seller is trustworthy. If some users think a low-PR page is a sign of an unreliable seller, that's bad for the seller, but it doesn't necessarily mean that displaying PageRank is a bad idea.
(Disclaimer: I don't especially care whether Google displays PageRank or not; that's for Google to decide, although the toolbar would be a more useful measure of a page's importance if it were kept up to date and included a broader quality indicator based on "signals of quality" in the Google Search algorithm.)
|By the way, the last few posts have talked about e-commerce businesses, but it's worth noting that providing credibility for e-commerce businesses isn't part of Google Search's stated mission |
Exactly! So why is Google telling users which pages are worth their time? Why lower a site's credibility with such a nonsense statement.
If you are referring to my posts, they are not only about ecommerce. Other types of sites have goals of converting visitors into users such as forums and blogs.
See Stakaman's post about trust. Clearly some users equate PR with trust and credibility.
|which is to serve as an indicator of "the importance of the page" (presumably as a source of information), not of whether a seller is trustworthy. |
Yes, and that is fine except when you state that it is an indicator as to which pages are worth your time.
|If some users think a low-PR page is a sign of an unreliable seller, that's bad for the seller... |
Well that's what Google is telling them, or at least insinuating, by stating the PR display will tell them which pages are worth their time.
|...although the toolbar would be a more useful measure of a page's importance if it were kept up to date and included a broader quality indicator based on "signals of quality" in the Google Search algorithm |
I would agree with that IF Google's quality indicator doesn't mislead users into equating it with trust and credibility. But at the end of the day I don't think Google or any other search engine should be advising users as to the credibility of any site based on an algo. That is not what determines credibility.
|So why is Google telling users which pages are worth their time? |
How much of the general public knows that PR changes for each PAGE opn a website? Who, except for another webmaster, would bother to check to domain root to see PR, if they already landed on an internal page?
|How much of the general public knows that PR changes for each PAGE opn a website? |
I know Webmasters who don't understand that. Some of think they're pretty clever at SEO, but then they say things like "I've got a PR4 site."
|How much of the general public knows that PR changes for each PAGE opn a website? Who, except for another webmaster, would bother to check to domain root to see PR, if they already landed on an internal page? |
Right, we don't know if they bother to check. Since internal pages more commonly have lower PR, the display immediately indicates those pages are not worth a user's time and, in that case, there's no need to click through to the home page or any other page.
|I know Webmasters who don't understand that. |
Imagine what the remaining 99.9% must believe with documentation stating it is a way to determine whether a page is worth your time.
|Imagine what the remaining 99.9% must believe with documentation stating it is a way to determine whether a page is worth your time. |
The remaining 99.9% don't look at the documentation, so they haven't seen that statement.
There are people (including Webmasters) who think PageRank is a measure of traffic. They just see the green bar and make their own assumptions.
I like Reno's idea of displaying an icon for the gauge and showing the gauge only to people who want to see it. That should keep everybody happy: Clueless mopes would have less likelihood of being confused, and Webmasters, SEOs, or obsessive users would still have access to the green bar.
|The remaining 99.9% don't look at the documentation, so they haven't seen that statement. |
I don't know why they wouldn't see that statement when they select to add the PR display. It's right there in black and white. At any rate it sounds as though you are saying if users haven't seen it, it's OK to state it. Even Matt Cutts is more objective and critical of Google than that.
|At any rate it sounds as though you are saying if users haven't seen it, it's OK to state it. Even Matt Cutts is more objective and critical of Google than that. |
It has nothing to do with being critical or non-critical; it's a matter of being pragmatic, realistic and--yes--objective. In the overall scheme of things, how Google maintains and describes the toolbar PR gauge just isn't that important (especially to users, who aren't likely to digest and certainly aren't likely to remember whatever Google says about it in a FAQ or on a download page).
I can think of far more important issues to get upset about than whether Google says that toolbar PR is a measure of a page's importance (or whatever the definition du jour may be).
I am surprised to see that people do not use PR more often to judge validity and relevance. Google's algo is based on these factors. If you are an online retailer, have been around for some years and your website PR is 0-2 to me it says that not enough people find your web design, products or services exciting enough to link to.
And of course the way to judge is not by saying the more the pagerank the better the online shop, but a PR4 and above would give enough indication of a healthy site (in terms of popularity and trust) and therefore more likely 'not to be' dodgy. There are enough cons and spammers (of any sort) out there..
I am pretty sure that by following Matt Cutts and things that Larry Page has said in the past this is on par with all what the Google algo is about...
|... it says that not enough people find your web design, products or services exciting enough to link to. |
When I started on the web in '96 it seemed more like a "grand experiment" in cooperation and information exchange, so in that sense there was a more universal willingness to point to other websites and say "hey, there is some pretty good stuff over here also".
But I'm beginning to wonder whether that attitude has been shifting over the past few years, as the WWW becomes more and more a hugely important commercial vehicle, so people are less inclined to send their visitors offsite for any reason. I'm wondering if for many businesses that sell something -- as opposed to providing pure information or opinion -- if the attitude isn't becoming more of "keep them inside the store", much as it has been in the bricks & mortar world.
That statement is not meant to be a sweeping generalization of every webmaster, but rather, a potential trend that I see as becoming more and more the norm.
Add to that Google's downplaying of traditional reciprocal linkage ("you link to me and I'll link to you"), and what we may see in the coming years is many millions of isolated boxes, with few if any external doors. If my perception is correct and that happens, where does it leave PR (with its emphasis on "natural linking") as a truly meaningingful measuring tool? Time will tell.
[edited by: Reno at 12:11 pm (utc) on Aug. 17, 2007]
| This 56 message thread spans 2 pages: 56 (  2 ) > > |