| This 62 message thread spans 3 pages: < < 62 ( 1  3 ) > > || |
|What Wikipedia Teaches|
There's no magic; you can do what they do.
Content-driven AdSensers sometimes bemoan Wikipedia. They seem to imagine that it ranks well for so many terms because it just has so many pages, and so many contributors, and is so well known. The wiki-mystique blinds people to the fact you can rank higher for terms by just doing some of the same things Wikipedia does.
- Have definition pages.
If your site is about widget repair, and some of that involves widget lock washers, why don't you have a page devoted to explaining what "widget lock washers" are, where to buy them, and so on?
Why does that matter?
It gives Googlebot pages whose primary topic is easy to grasp. It produces SERP listings that do a good job of communicating to searchers exactly what they'll get if they click on your link. That, in turn, means folks who click are highly likely to find what they need, so Google will give that page more ranking love.
It also gives you a nice landing page that can "sell" lots more content to the visitor. Once you know they arrived looking for something so specific, you should be able to tick off all the other pages that are relevant and link to them from there.
- Be coherent When people see wikipedia in the SERPs, they know what they're going to get if they click on it -- some encyclopedia-like explanation of plausible quality. Your website isn't as well known as Wikipedia, so you have to work hard to make sure your SERP entry gives a good clue as to what they're going to get. You don't want a SERP that claims to be all about "mastering pool shots" that actually links to a page describing drinking games for your swimming pool -- when Google sees all the folks who click on that are coming back to look elsewhere, your ranking goes in the crapper. This is where a well-thought-out domain name and carefully chosen URL text pays off. If my SERP's URL looks like www.widgetrepair.com/using-widget-lock-washers, and my SERP synopsis repeats the "how to use widget lock washers" message, and the linked-to page really is about that, then I'm clear about what I'm delivering and Google will reward that (assuming anybody is actually interested in that topic).
If you see people are arriving at your "widget lock washers" page with a search term like "widget painting", figure out why that is. Too many not-on-target SE arrivals probably implies you need to split some content off and give it its own page (with lots of internal links to it, of course). Remember: people finding what they were looking for explains much of Wikipedia's ranking success.
In case you missed it, I'll reiterate an important sub-point here. Your URL is the one thing on the Google SERP that you have 100% control over. Stop focusing on keyword stuffing your URLs and focus instead on using a URL that accurately conveys to the searcher what they'll get if they click on that link.
- Use intra-site links everywhere Lots of people waste enormous time and energy trying to get someone to link to them -- and don't lift a finger to link to themselves. Relevant intra-site links are gold! You don't have to beg anybody for them, you get to control the anchor text, and you don't have to add links back to anybody else to get them. Some people think it's a magic power that Wikipedia has to quickly pop to the top of any long-tail search. It's not. It's the fact that any new page invariably is linked to (with the key term in the anchor text) from one (usually more) other Wikipedia pages.
- Cover up your nav bars Do you feel like you already do use inline, intra-site links a lot? Baloney. Put your hand over the nav bars on your site. Now try to navigate. Can you get to all the pages on your site? How long before you hit a page that's a total dead end. The wiki structure essentially forces Wikipedia content to do things (like extensive intra-links with relevant anchor text) that are great for SEO. Ever get annoyed with how sometimes it seems like every other damn word on Wikipedia is a hotlink to another page? That shows how little you use intra-site, inline links in your own pages.
- Cover up your nav bar -- really! Somebody is saying "but my navigation is really logical and comprehensive, and I don't like the look of inline links." Ever hear of ad-blindness? Well people get just as blind to the navbar. I can put a big, red "CONTACT INFO" link in the navbar on every page -- and every day I will get an email saying "what's your address/phone# over there?" People cannot see your nav bar. Use inline links in the text with relevant anchor text. Santa knows if you're sleeping, but Google knows if a visitor didn't find what he wanted on your site -- and complaining that the visitor should have found the information won't alter Google's ranking one bit.
- Write, dammit, write! Sometimes the most obvious Wikipedia lesson is the most overlooked. I see people sometimes with a product to sell moaning over how they just can't get the SEO tricks right to bring in the free Google traffic. Then I go check out their site and they may literally have less than 10 pages of content. Which part of "Google likes content" did they not understand when they went to SEO school? Often people have extensive product manuals that they haven't even bothered to put online (or almost as bad, they put the whole manual up as a single PDF file), and yet they'll spend hours trying to tweak keyword density, or beg for links. Wikipedia shoves the message of "make some content someone wants to read" in your face, yet some people still don't get it.
- Edit as needed You can't be on WebmasterWorld long before you see people talking about being afraid of changing something on their website for fear it will drop their rankings. Sure, that's why Wikipedia never lets anyone change their pages... Change pages as needed, but most especially(!) to add in links to new relevant pages you've added, so those new pages will get immediate link love. If the last time you added a new page of content, you didn't have to go add inline links to it from at least a couple of other pages, then you haven't learned the Wikipedia Way.
- Make an Authority Page that Attracts Links OK, it is true that wikipedia also benefits from external links. You can compete on that front, too. People will often just link to them because it lets them avoid a lengthy explanation of a term. So supply those pedia-like pages that would be useful for someone who just wants to link to a term definition.
In fact, if Wikipedia is overlapping into your Long Tail content, then why not jump on their back to higher rankings? Look at their page devoted to "widget wuffling". Make your own page devoted to defining that content. But make yours better/different. Supply something they don't. OK, now you've got a ready-made link campaign that will work a whole lot better than begging for cross-links. Use search tools to find all the pages hot-linking to Wikipedia's "widget wuffling" page. Track down those author email addresses and make your pitch. "I noticed your 'widget wuffling' link to Wikipedia and just wanted to point out that it says wuffling was first used in 1857. However, it was actually first used in 1856, as I've documented on my page at [...]".
If you're staying awake, you've just realized that tracking down links to Wikipedia pages can be a real useful research tool for AdSensers in the content business.
Wikipedia dominates rankings for a lot of different reasons, but a lot of those reasons are things you can do as well or better than Wikipedia. And you're not handicapped by a lack of consistent style, a prohibition against commercial or controversial slants, or a horde of random visitors who keep patching your content when you're not looking! Don't hate Wikipedia -- learn from the free lessons they provide.
Great tips here, I've learned a few new things and will be looking again to re-read. I'm so glad I stumbled on Webmaster World with such nice people! :-)
Thank you ronburk, thank you all who contribute.
Printing this one for weekend re-reading.
Anyone know how the internal Wikipedia links are generated? Automatic suggestion, manual searching?
Just looked at the science portal page and noticed:
- inline links to external sites
- inline links to the Widget page (filename:widget, anchortext:widget)
- inline links to the Widget page (filename:widget, anchortext:widget-related term)
Who/ what determines the related term in the anchortext?
I would like to imagine a crawler that goes through my site, classifies my content and provides related terms (thesaurus) suggestions.
ronburk, thank you for the very thoughtful response to a few of my reservations about following the Wikipedia model too closely in terms of inner links.
As one who thinks associatively by nature, and knows the thousands of pages of my sites almost too well, I do apply these links where I think they are appropriate, but not all sites are encyclopedias. Articles, for example, can become annoying to read with too many hyperlinks, from my personal experience, unless the hyperlink addresses -- as you say -- more detail for that particular reference. I think it is a common-sense to seek a balance, and I often hyperlink OUT of my site for information my site does not have. I don't worry about stickiness at all, as the reputation of my sites are based on helping the user find the best information available no matter where it may lie (including Wikipedia). I always liked the Google model of leading users elsewhere and thereby enhancing their reputation as a point of departure.
Having said that, I am not sure that I want the user who has found a particular page via a long-tail search to be distracted from the narrative of an article though excessive hyperlinks to related articles. But, as you point out, this is all a judgment call and really requires experimentation and creative thought in addition to analysis of users' activity on your site. Very good point of emphasis you have made.
I really appreciate the distinction that you made between "sticky" sites and those that are found via search, as an examination from that angle does change the way you think about the quantity of internal links you wish to use, though since this is an Adsense forum, the next question becomes whether users are more likely to click on ads if they find everything they want on your site moving from page to page! The Wikipedia experience would certainly not lead me to click on an ad unless I hit a "dead end" or "stub page" perhaps.
To me, it comes again to that paradox of advertising. If your content is too rich in links, no one will click on your ads of any kind. If it is not, they may well click, but you are not doing your job.
Thanks again for your thoughtful and rich remarks ronburk. This is the kind of thread that is enjoyable and educational to read, without requiring participants or the audience to be overly technical and using arcane references which tend to alienate those who do not live in the world of jargon.
|To me, it comes again to that paradox of advertising. If your content is too rich in links, no one will click on your ads of any kind. If it is not, they may well click, but you are not doing your job. |
I have the same conflict.
I have one small site that provides “widget” information. I provide a link to the manufactures, and approximately 30% of visitors go straight to the manufacture.
When I first set the site up, it immediately began receiving traffic and I was very happy. However, no one was clicking any of the ads. I was disappointed and stopped developing the site. The site is still getting traffic but very little income.
I am thinking about removing the manufacture’s link and see if that works. I am somewhat reluctant to do that because it just does not feel “right.”
This is a project that I am interested in but I cannot spare the time to work for free right now.
Maybe a little less information will save the site.
|Articles, for example, can become annoying to read with too many hyperlinks |
Maybe, but I don't think many of us have to worry that we'll be tempted to link from every other word a la Wikipedia. Why? Because most of us don't have millions of pages on every conceivable topic.
Also, not every link to an internal page has to be an embedded link in body text. A 5-page article on chocolate, for example, might consist of:
2. Origins of chocolate
3. How chocolate is made
4. Types of chocolate
5. Related resources
Each page could have a link to the next page ("Origins of chocolate," "Types of chocolate," etc.) at the bottom, along with an internal navigation table that shows all five pages. With this kind of structure, there wouldn't be a need to as many embedded links in the body text as there might be in a Wikipedia articles. The embedded links would be for links to pages outside the article, such as a page about chocolate desserts, a page about chocolate in Mexican cooking, a page about chocolate drinks, etc.
I use this approach on my own site (which isn't about chocolate, by the way), and I like to think that it offers the best of both worlds: structured navigation for multi-page articles, directories, etc. plus the use of embedded Wikipedia-style links as citations where appropriate.
Yes, signor_john, many of us use this methodology where appropriate, but some of us DO have an enormous amount of related content--perhaps not millions of pages, but certainly tens of thousands (which is my case), not to mention the millions and millions of external pages which are just as valid and important to any site with its own notions of integrity (including the notion that key information and background is important to get to users, no matter where it lies, on your site or that of others.)
Internal navigation methods are many and varied, and should be used where appropriate or most successful, including breaking up one's content, as you have cited. That is just one of many, many scenarios depending upon your site structure or sub-structures.
I absolutely hate the way many many major and minor newspapers, for example, break up article pages such that I have to click next, next, next in order that they receive more pageviews and display more ads... That is just a very cheap and off-putting gimmick, IMO. That, and the AJAX-based hyperlinking of every word such that my mouse provokes endless undesired popups as I move about. Gimmicks that you would not expect of some major newspapers, no matter their financial situation.
|I absolutely hate the way many many major and minor newspapers, for example, break up article pages such that I have to click next, next, next in order that they receive more pageviews and display more ads... That is just a very cheap and off-putting gimmick, IMO. |
I hate to wander off-topic, but a number of years ago, WIRED reported on a usability study at a university (Stanford, maybe? ) where Web users were shown two versions of an article: one that required a lot of scrolling to read, and another longer version that was broken into multiple pages. A majority of the users preferred the multi-page version, and get this: The users thought the multi-page article was shorter than the scrolling article, even though it was in fact longer. (As ROLLING STONE used to say in its trade ads, "Perception is reality.")
Do you have that link? I would be interested, as I have read many usability experts speak of the converse effect. But this does have to do with hard copy thinking being applied to the Web. I find that having to click page 1,2,3,4,5,6 to be idiotic if there is no option to see the whole article at one fell swoop.
Oh, and Wikipedia pages do tend to be long scrolling pages and they are rather successful...which I think was where the wise poster initially started the topic.
No, I don't. It's been a few years.
|But this does have to do with hard copy thinking being applied to the Web. I find that having to click page 1,2,3,4,5,6 to be idiotic if there is no option to see the whole article at one fell swoop. |
It can make a lot of sense if the article breaks down logically by subtopic or if you want to include photos of reasonable size without breaking the flow of the article. Also, many newspaper sites (and other sites) offer a "printer-friendly" option that places the entire article on one page for those who'd rather scroll than click.
This is getting a bit off-topic, since most newspaper articles do not break their articles into topics and subtopics (which would be nice) insofar are they are generally created for the print world, but the original poster's far larger point relates to internal linking within one's site and he has brought out some very important perspectives regarding the many possibilities in terms of internal linking.
The question is how these lessons can be applied to Adsense publishers, and there are clearly many more possibilities than there are publishers. He is absolutely right on one of his central points in that it boils down to being creative with your content and intra-site navigation, whether you are selling widgets or information, and Wikipedia's information-intensive model and structure clearly provides a proven template for creating or enhancing content when it is customized to one's particular sector and audience. I thank him for adding a useful and helpful perspective.
|Oh, and Wikipedia pages do tend to be long scrolling pages and they are rather successful... |
They are just successful in Google search but they DO NOT SELL ADEVERTISING. If they were a commercial site, they would to change some things, probably a lot of things.
Thit has a simple proof: there's not commercial sites like Wikipedia; maybe About.com is in the same range but you can see the difference.
Lexur raises a good point. This is the AdSense forum after all, and content presentation, including linking schemes, needs to include our commercial objectives.
It's NOT just about readership.
For the past 13 months I have been developing a wiki on the mediawiki platform, ronburk is on the money here, it's suprising how much extra pages you can create when you re-read your newly created page. This can easily be 'overdone' though, I try to keep to a formula, can I write more text on the New page/word/topic than the template has, because a page with 3 lines of text isn't much use to anyone!
The good thing about wiki software is the ease at which you can link, for example:
If you were to link to the "Widgets" page all you do is [[Widgets]] now you can also easily link to that page with different anchor text like so [[Widgets¦widget repairs]] or [[Widgets¦buy your widget now]] if a word appears as a link but red, it means there's no page for that word yet, the red link tells you it's a wanted page.
After you create a new page, you can search your wiki for other pages that contain that word and just put square brackets around it : [[word]]
For external links it's the same method
[http://www.example.com Buy widgets direct]
Again, "buy widgets direct" is the anchor text, wich you can setup differently on any link.
Don't forget the pipe character gets broken on WW
On another of my sites I have a customer feedback/letters page (this is a good nugget) all feedback letters are added to a DB and when called out to their own page I have a find and replace function, to change specific words to links, you decide where those links go, example:
"I really love my new widget, it has made widgeting so much more fun"
The word widgeting can now become a link to my widget page/site.
|They are just successful in Google search but they DO NOT SELL ADEVERTISING. If they were a commercial site, they would to change some things, probably a lot of things. |
Damn right they would.
|This has a simple proof: there's not commercial sites like Wikipedia; maybe About.com is in the same range but you can see the difference. |
About.com is an example of the opposite extreme: commercialization to the point where the content becomes filler. I think that, for an information or oontent site, there's a happy medium or "sweet spot." And while there are things to be learned from Wikipedia, we can't simply emulate it unless we're not-for-profit entitities with foundations to support us.
"Lexur raises a good point. This is the AdSense forum after all, and content presentation, including linking schemes, needs to include our commercial objectives.
It's NOT just about readership."
Yes, this forum is about commercial viability (which sometimes takes the form of producing lousy content, I suppose, as the frustrated user will then click on ads), but the points he brought up
specifically about creative intra-site navigation
make for a better user experience which often translates into income and growth in reputation which brings in more traffic via the power of high-ranking external links, in my ten years of experience doing this for a living for both Adsense and traditional advertising revenue. Content remains King, and therefore Wikepedia offers a lot of pointers for even commercial sites.
|Content remains King, and therefore Wikepedia offers a lot of pointers for even commercial sites. |
Sure, just as long as you remember that Wikipedia may have reasons for doing things that have less to do with the reader experience than with the editorial process. Long scrolling pages are one example; stubs (which few readers are likely to appreciate) are another.
IMHO, the biggest lesson to be learned from Wikipedia is simply that content does matter, and making that content easy to read (and to easy to find via links) will pay off for the publisher.
[edited by: signor_john at 3:58 pm (utc) on Feb. 13, 2009]
|see your site the way fast-moving, nav-blind, informavores who come to your site see it. |
Most top navigation elements are in the "banner zone" anyway. ;)
ronburk, how do you assist the visitor with getting back and forth between the internal linking mechanisms?
I've been making ample use of a "Back to Previous" link at the bottom of each section of content. I work with "blocks" of semantic content so I can easily insert those Back to Previous links after the last line of that block. It makes for a very user friendly navigational experience and keeps them focused on the subject at hand. I also have ample inline links to specific sections using Fragment IDs. All blocks of content are given ID's so they can be linked to directly. None of this scrolling about looking for information, we're going to put you right there at the heading for that block of content.
All those eye tracking studies sure do reveal quite a bit about the value of that top navigation. It's a dead zone for many. That's why I think an OmniBox filling that dead zone is the way to go. Do something different, break the mold. If visitors see something up there they are not used to seeing, it may just be the trick to add some value up there. A really big dynamically sized OminBox. Onsite search is an invaluable tool!
|That's why I think an OmniBox filling that dead zone is the way to go. Do something different, break the mold. If visitors see something up there they are not used to seeing, it may just be the trick to add some value up there. A really big dynamically sized OminBox. Onsite search is an invaluable tool! |
Yes, when readers are willing to use it. On my site, very few of them do, even though they can't miss the search box. (I witnessed the same phenomenon when I ran sites for 4-1/2 years at About.com: Most visitors may have arrived on About.com from search engines, but once they got there, they nearly always relied on navigation menus, not on search.)
I think ronburk's emphasis on links within the page makes a lot of sense--not just because that approach has worked for Wikipedia, but also because clicking on a link is intuitive and doesn't require much effort by readers (especially by readers who lack the techie or librarian gene and are locked into the mindset that says "Search is for getting to a site, and links are for getting around the site after you get there").
Over the past days I have implemented some ideas and emphases generated by ronburk's suggestions on some high-traffic pages in order to test theory in reality in terms of garnering ideas from the Wikipedia model, and -- though this is a small time-frame -- have seen remarkable results in terms of increased intra-site activity, and, most importantly, in the form of revenue generation this forum addresses.
Hmmm. It would be hard to do that at times. What is in the navigation might not be the natural phrase that would be in the text.
|NEVER double link with different anchor text, it just confuses G |
Ahh, a solution.
|rel="nofollow" all but one reference and it is usually one of the inline versions |
I don't like it when Wikipedia links so many words on a page. It makes sense to link words or phrases that will truly give more information. I do think it improves usability as well as helping with Google.
I'm not too concerned about losing potential clickers to other pages on my site. The more of my site people see the more likely they will bookmark or link from their site. If it nudges the page up a bit in Google all the better. The long range value is far more than what a click on an ad would be.
<A majority of the users preferred the multi-page version>
Not me! It p#$ses me off to no end when I want to print the entire article to give to someone and I have to click to and hit print on every page and most of them are just long enough so that they print one page and two lines on the next page. So you wind up with 6 pages which look a mess instead of 2-1/2 neat dense pages if it were all on one webpage!
Also P's me off is all the inline links on wiki which were obviously set up by a moron who either did not read the context of the word on the current page or who never clicked on their own link to realize the other page had no content or totally unrelated content. It's gotten to the point that I just stopped clicking on these altogether. It would be a lot more useful if the links were far fewer between and actually useful to the point of adding value.
|There's a profusion of woes over in the penalty discussions at Google Search |
All of which is nothing but speculation as far as I can see.
|It p#$ses me off to no end when I want to print the entire article to give to someone and I have to click to and hit print on every page and most of them are just long enough so that they print one page and two lines on the next page. So you wind up with 6 pages which look a mess instead of 2-1/2 neat dense pages if it were all on one webpage! |
The Web isn't a print medium, so optimizing pages for printing is likely to be a low priority for most Web publishers. Still, when a printer-friendly version is available, you can eat your cake in pieces of manageable size and take an uncut cake home to share with your friends.
I have an (niche) encyclopaedia site and competing with Wikipedia is tough (they cover all niches these days!).
A Wikipedia stub can beat a properly written on another site - and not just on mine either.
|Over the past days I have implemented some ideas ... have seen remarkable results |
Would you care to elaborate a bit?
|I also have ample inline links to specific sections using Fragment IDs. |
Good tip - this is what you mean about writing in "blocks"?
|That's why I think an OmniBox filling that dead zone is the way to go. |
Intuitively I would have also thought this was the way to go - a kind of query reinforcement - but it appears not be used?
Thanks for the wiki internal linking info wheelie34
|Would you care to elaborate a bit? |
Three times the thousands of daily page views for this subset of pages (based on drilldown statistics) and approximately twice the CTR on pages which were already fairly complete with links, but in this case the links have been more strategically linked to yet more related content on the site. Basically, the richer the intra-site linking, the better results I have seen thus far -- but again, time will tell.
But the experimentation goes on, and the original post has provided plenty of ideas (many of which I was aware of, but had not implemented to the extent elaborated).
|Three times ... daily page views ... twice the CTR |
Gosh - that is impressive
|links have been more strategically linked to yet more related content |
So you had a number of related links already? And you have not only increased the number of related links but have also changed the format/ placement of said links (is this the strategic bit)?
I wonder how far one could do this and increase page views - obviously not to the point of alienating users where every second word is a "related link". How to find this sweet spot?
Hmmm - excellent food for thought.
|So you had a number of related links already? And you have not only increased the number of related links but have also changed the format/ placement of said links (is this the strategic bit)? |
Yes, both placement and quantity and, as always, looking for the "sweet spot" which likely differs for every page, and will differ on every site. But without experimentation, pages get stale anyway (even for those who have created them), and experiments in internal linking serve many different purposes--though the user experience should remain to the fore, and I would in no way advocate going to the extent of Wikipedia in terms of internal linking for most sites. But getting people to see the richness of your site other than via standard navigation is, to me, a fascinating way to look at some of the common drill-down and Web 2.0 metaphors or templates in a new way in order to try to get inside the visitor's mind at a deeper level.
This might help answer the question as to why some pages which are rich in links do well in terms of conversions, while others do not. The usual paradox, IMO, is that a page is a failure if the user needs to click an ad, but then there are a few pages very rich in well-researched and good content where ads are still clicked. That is fascinating to me.
|The usual paradox, IMO, is that a page is a failure if the user needs to click an ad, but then there are a few pages very rich in well-researched and good content where ads are still clicked. That is fascinating to me. |
I've never understood the notion that "a page is a failure if the user needs to click on an ad." That might be true of an e-commerce site, but for an information site, it makes so sense at all. If I review a camera (to use the classic Google "smart pricing" example) and readers click on a dealer's AdSense ad because my review has convinced them to buy that camera, how can that be described as a failure? It's a win for the reader, a win for the advertiser, and a win for me.
Sure, but you have not provided the links needed to buy the camera, then you are simply providing a tease. In my view--in the world of free-flowing information--that is simply a sell-out. An information site should be honest enough to provide the necessary information if it has it when providing a review. If the site does not, then it is just a made for advertising site. If I review a site or product, I always make recommendations in terms of where to buy the product or alternative products. Always.
But I know that many "offer" made for advertising sites and "reviews" these days in thin disguise, which only brings down everyone's income for the benefit of those who provide the teasers, since there are so many such "reviews" which are SPAM-like in nature.
But this is really, really off-topic relative to the thread, which is the application of Wikipedia ideas in the development of intra-site navigation and other related ideas useful for commercial websites. I have found these ideas extremely useful thus far in reality, as opposed to those who oppose them for the sake of opposing them...
|Sure, but you have not provided the links needed to buy the camera, then you are simply providing a tease. In my view--in the world of free-flowing information--that is simply a sell-out. |
I'm afraid you've got a warped view of what information or editorial sites are supposed to do. Next you'll be suggesting that Wikipedia should include a list of widget dealers in its article about widgets.
As for applying Wikipedia's ideas, I think it's worth pointing out that Wikipedia's navigation methods, use of citations, etc. are simply classic Web 1.0 techniques that were common back in the early 1990s before the existence of the commercial Web. In an era of computer-generated junk sites, it's refreshing to see Wikipedia demonstrating that traditional Web virtues (human writing and editing, internal linking to related information, and citing external resources) are still relevant--and that such traditional virtues can attract a huge and loyal audience.
| This 62 message thread spans 3 pages: < < 62 ( 1  3 ) > > |