Forum Moderators: open
ive recently come across a number of websites which are using the above techniques, shoudlnt google be picking this up?
-Humpingdan
There are so many valid reasons to use some of these techniques that any automatic filter presents a risk to the quality of their SERPs
Of course, a hand inspection would reveal all.
Is it really worth it though?
The only hidden text I have out there is on a test page on a disposable domain which I then spam reported myself. I got fed up of all the 'hidden text gets you banned' preaching and thought a real-life test would reduce the volume a little.
It is has not been removed yet after a good few months.
(Of course, you only have my word for all this;))
Google really struggles with css based hidden text
Google struglles with ALL hidden text.
Absolutely, a lot of good, legit sites could get knocked out.
>>background colour the same
Among valid reasons: Body background white, navigation table data cells navy background, white text in them. Not hidden at all, done so the text IS clearly visible.
One way of combatting is devaluing keyword density relative to other factors, but if anyone thinks they're getting beaten by other sites because of keywords in hidden text it's either time for them to do likewise or take out the heavy artillery to beat those sites - or it's time for them to re-examine their grasp of Google optimization.
Where it could be a factor worth considering is with links, but a good question to ask is whether that will have an effect on rankings for the page the links are on, or the pages that are linked to using hidden text links.
- SE's will ALWAYS be abused
- SE's will always fight the abuse
- SE's will ALWAYS try to get the great web sites (as of user perception) highest up
Now do you want to spend your time to build great web sites for your users (and get rewarded by the SE's in the long run along the way) or adjusting to your competitor's scams and the SE's counter measures?
Just a simple question to me, ignore the "extra" dollar today, go fish big for the long haul.
Reward: great stability and consistency, way better efficiency, less trouble, focus on the end user = not always the biggest bucks but by far less downturns...
PLUS: Nobody else wastes MY time!
Cheers, Jens
good call .. but very few that i have come accrossed to in comparison with hidden texts.
Using invisible GIFs as links, or even sneaking them in with individual characters or punctuation are tricks that are older than dirt. And let's face it - if there are enough hidden links out there to sites chances are there's got to be a lot of them to make a difference, and could be a clue about large networks of sites belonging to the same people. If they're doing that, chances are it's for highly competitive areas with not a few competing sites doing the same or similar thngs, and that's probably not all they're doing.
Leaving those alone deliberately might not be a bad way to catch on to whatever else the latest tricks are that those folks are using, and I can think of worse ways to backtrack and compare and find out eventually exactly who those folks are and what new sites they're putting up.
I've never been convinced that Google doesn't know exactly what's happening. There's a lot of investigative value with some of those sites staying out there. Leaving a bit of bait dangling on the hook can sometimes lead to catching some nice, big fish to fry.
Studying up on those sites is one of the best uses of time I can think of. It's like a free SEO course to see what's working and how long it stays around.
So let the whiners continue to whine and the spam reporters continue to report what Google probably already knows, while others appreciate the educational value that's out there just for the taking.
The best defense is a good offense - build a killer site with killer content.
Considering that source code for web browsers is available to look at, and that google strives to be a high end search engine, it makes little sense to me why they do not yet have the capabilities to have the crawler render the webpages it indexes at least to some degree.
I say, do NOT use this method to find what webpages contain hidden text. That is thinking backwards.
I propose that google uses html rendering to see for itself what text on a page it should use for indexing. Ignoring text it can't render (hidden text). If a webpage isn't written with proper html and doesn't render properly, that will not be google's fault, nor something google would need to worry about.
Basically, stay away from "find N, X, Y loopholes in html that allow the search engine to see hidden text that users don't see" method, and instead go with "render only normal html text that meets html standards that a web browser should see". So white listing rather than black listing :)
There are some issues with what I propose, but I don't see any of them as being a deal breaker for my proposal.
Creating a web browser is by no means a small task, but it is by no means something that can only be done with rocket scientists.
Exactly.
Render page into memory following all the rules as if you were a browser, permit for indexing only that text which is rendered with sufficient contrast between FG and BG color.
I'll bet Googlebot has just such a "Virtual Browser", rendering pages into memory, following all CSS and JavaScript instructions, and then studying the bitmap.
So in a way, it is not really checking for hidden text; it is more a new way of extracting indexable text from a set of source files comprising HTML, CSS and JavaScript.
Note also that since the web has moved on from plain vanilla HTML; it is in the search engine's interest to be able to index the computed visible text following "virtual" processing of JavaScript etc; since this allows pages with dynamic inserts and various other document.write hacks to be indexed successfully.
It's win win - and would explain why G'bot has been busy requesting all your .js includes... :)
Surely this is a little optimistic? Not many browsers follow "all the rules", even if you can work out *which* rules apply to pages in HTML 3, HTML 4, HTML 4.01, XHTML, and which rules apply to those many sites that serve pages that don't actually declare what language they're using...
The day that all browsers work properly is the day that I think there might be some reason to try to render pages in a search engine, but even if one renders the whole darn lot, what about such tricks as placing a single link about 20 feet below the rest of the content. Rendered or indexed as lynx-like text, the link is there, but I suspect Joe Public may well not find it... Search engine tricked again...
Best regards
DerekH
permit for indexing only that text which is rendered with sufficient contrast between FG and BG color.
That's the hard bit - i've got competitors where the first thing on the page is a keyword targeted href to an internal page, slightly different color to the background, but not enough for the human skim reader - ranking top 10 for a top 30 query according to wordtracker.
I guess it's a moot topic: "sufficient contrast" (only used "moot" because of mooter ;))
Surely this is a little optimistic? Not many browsers follow "all the rules", even if you can work out *which* rules apply to pages in HTML 3, HTML 4, HTML 4.01, XHTML, and which rules apply to those many sites that serve pages that don't actually declare what language they're using...
I agree that it is optimistic; but it is also the only practical way to tackle the hidden text problem.
JavaScript is a programming language, and brings with it an infinite number of ways to do an infinite number of things; so there is no way that you approach the hidden text problem from the angle of "how can you hide text with JavaScript"!
The browsers may have given us no reason to try and adhere to standards; but the search engines may just have...
I'll bet Googlebot has just such a "Virtual Browser", rendering pages into memory, following all CSS and JavaScript instructions, and then studying the bitmap.
I'll take some of that action if we can establish a way to verify who's right.
It all comes down to (best use of) CPU power. Rendering the pages will not be the best use of CPU power for some considerable time yet.
Kaled.
PS
With respect to Google not being able to automatically detect hidden text, I can honestly say told you so.
Google struglles with ALL hidden text.
Apply a 80:20 rule (only 20% of all factors make up 80% of ranking) and you'll probably arrive at word density being in the 80% of factors that only determine 20% of ranking.
What am I trying to say is that Google probably can proove statistically that detecting and penalizing all those hidden text will improve results only 1% while costing 100% (meaning: requiring double the CPU power than now). This simply is not efficient.
Better invest in strategies that WILL make results better.
