Google Killing Keyword Tracking

If you’ve ever done “keyword research” for your websites, you know it’s a minefield. Unless you have a Ph.D. in keywordology, you’ll quickly be bamboozled by comparing ”visitor volume”, “seasonal trends”, “exact versus phrase match” and all the other related variables that go into finding “good” keywords to target and write articles for.

One of the easiest ways to shortcut the entire process is to search your webserver logs for the keywords that people actually USE when searching Google (or other search engines) before visiting your site. I even created a product called “Niche Treasure Hunter” around this keyword research topic. All you have to do is trawl your weblogs for these “Bonus” keywords. You can do it manually, or use software to do it automatically. I do a bit of both, with Google Analytics doing the automation for me.

When you discover a keyword that someone used to find and visit your site, it’s a great feeling. If it’s a keyword you wrote an article around, it’s proof that your SEO is working. If it’s a keyword you DIDN’T write an article around, that’s even better because it shows that Google trusts your site enough to send you traffic for a Golden keyword you didn’t actually optimise for… and now you can write an article around it to get more visits!

For example, say your main keyword is “toy electric train”, but, one day, you notice someone visited your site for the keyword, “electric model train”. When you check at Google, you’re only on page two for that keyword. How hard would it be to get more visitors for “electric model train”, when you know that, without trying, you’ve generated traffic for it from page two of Google? Pretty easy, right? That’s why I call them bonus keywords!

But now Google is starting to hide the data containing the keywords from webmasters. When people do a Google search while logged in to their Google account, the search is done at a secure page, an https request. The results of the secure search are show to the Google user as normal, and the click on the chosen search result takes them to the website they want to visit… BUT the search keyword is NOT passed to the destination web-server.

It’s estimated that about 10-12% of searches done at Google are secure ones, which doesn’t seem like a problem, until you consider what would happen if Google switched ALL searches to the secure system. All referral keyword information would be hidden resulting in webmasters losing “Bonus” keyword data, third party analytics software being unable to track “keywords to sales” and advertising platforms that display ads based on the incoming keyword searches would stop working.

So why would Google make the switch to secure searches? Well, it would be an obvious way to let everyone know (including lawmakers and regulators) that Google cares about “user privacy”. Not only cares, but is actively switching people to secure searches. Who could argue that making the switch would be a good thing? Of course, the fact that it kills “bonus keywords” and “keyword to sale tracking” for webmasters as well as certain advertising methods would be just coincidental. Interesting, Adwords ads still reveals the search query to the advertisers! I guess Google is assuming that, 1: people can tell the ads from the organic results and, 2: are happy to give up a measure of their privacy to advertisers that they’re not happy to give up to the websites appearing in the organic results. That’s quite a big assumption…

Of course, Google says that webmasters need not fear! If you use Google’s Webmaster Tools (GWT) you can, “receive an aggregated list of the top 1,000 search queries that drove traffic to their site for each of the past 30 days”. Woohoo. I don’t know about anyone else, but I don’t like GWT at all. I’d much rather do my own digging through my weblogs or via Analytics than trust the “aggregate list” from GWT.

What do you think? Is this a deliberate ploy by Google to move webmasters to Adwords, or just a privacy -related action ahead of a regulatory requirement? Will it affect keyword tracking? Will it affect you as a user of Google, or as a webmaster?

[comments are now closed. When the "spam to genuine" ratio exceeds 50:1 it's time to close the comments!]

Google Bowling

This is an update to the widely-read post, “Google Breaks Its Own Golden Rule” which you may want to read first to get up to speed.

I’ve tested this myself with new sites… when I built them using only unique, quality content and linked to them in ways recommended by some Internet Marketing “gurus”… BAM.. the sites get dropped by Google. Sites that were built the same way without the links retained their rankings, so it’s not the Google Honeymoon effect in action, but a real negative effect from getting links Google considered “spammy”.

So much for the Google Golden Rule that links from external websites couldn’t hurt your rankings. Perhaps that was true in the past, but it’s not true now. When I wrote about my suspicions back in May I thought that Google would only penalise sites for bad links if they were new sites. I expected the penalty to be lifted as the site aged… perhaps gradually so as to avoid a new Internet Marketing craze such as “Buy Websites Over 18 Months Old To Avoid The New-Site Penalty”. You can almost see the “guru” emails promoting the new ebook, right?

Well, in Dr Andy’s latest EzSEO Newsletter (which is excellent and you should subscribe to it) he released details of a test he performed on an old website which also lost ranking positions in Google when he built some “spammy” backlinks. In his test, the website was five years old, PageRank 2 with 80+ phrases in the top 100 at Google and within a month of building “spammy links”, all the rankings were lost. Dr Andy controlled the source of the spammy links and deleted them as a further test. Within three months 68 of the 80+ phrases are back in the top 100. For those who are interested, in the comments section of the blog post describing the test, Dr Andy says the “spammy links” were, “mass links on WordPress MU networks that I built. I had 25+ links to the site PER MU site”.

So, what conclusions can we draw?

Well, it now seems certain that the sudden appearance of spammy links can hurt sites, so Google has most definitely thrown out its “Golden Rule” in favour of tighter controls on backlink profiles to identify potential spam websites.

Furthermore, the profiles of sites at risk reaches well beyond the parameters I expected meaning many more, if not all, sites are susceptible. With both PageRank and aging, I would’ve expected Dr Andy’s site to be safe from any kind of link-profile penalty, but it wasn’t. So you can’t throw spammy links at aged sites with PageRank and expect them to be at best helpful or at worst ignored. It’s not clear if there are levels of PageRank or years of aging (or a combination of both) where sites do become immune to spammy links in their profiles.

It’s more important than ever to be a good guardian of the link-profile of any sites you own. It’s best to assume that building spammy links to any websites can actually wipe your site from Google’s search engine results… and if you don’t OWN the sites you put the links on, (which could be because you use services providing anonymous links), you won’t be able to remove them like Dr Andy did… so your site will almost certainly suffer a long period of ranking at the back of Google.

Double Six

The goal of SEO is to get to the top of the search engine results pages (SERPs) for any specific keyword phrase. As Google accounts for the lion’s share of the market (about 65% at the moment), most efforts are expended in trying to rank at the top of Google’s SERPs.

In the past, capturing the number one spot was the only game in town. Since then we saw “indented listings”, which were coverted for a while due the the fact that if you had a top ranking, you could get a second, indented, listing quite easily and more than double your traffic for that keyword. Indented listings fell by the wayside as Google evolved… and now, behold, the power of SITELINKS and BRANDLINKS!

When Google thinks you’re searching for a specific website, it may return “sitelinks” for that particular website. They look like this…

Google Sitelinks

But, Sitelinks aren’t the only games in town… there are also “brandlinks”, used when Google thinks you’re searching for a specific brand, in which case it may return several pages from the chosen website as full listings in the results. Here’s an example for Hitwise…

Google BrandLinks

But… wait for it… websites can get an immensely powerful combination of Sitelinks and Brandlinks. For example… Amazon and Nike…

Google Amazon SERP

Google Nike SERP

OK, I can hear the wailing from here, “but I’m not Amazon or Nike… how does this apply to me”?

Well, you don’t have to be Amazon or Nike to benefit from the sitelinks and brandlinks combo. Check out the results for this search…

Google PHPBB SERP

Through branding itself around “PHPBB Hacks”, that one website has no less than TWELVE links to it before the next website gets a look in!

So, the message here is to become the 500lb gorilla for any niche where you run a website. As you build a brand around the name of the website (note, the domain name exactly matches the search phrase) you’ll increase your chances for both sitelinks, brandlinks and the massive 12-link combo!

Google has started to support the “author” markup tag so that when you create a link, you can tell Google that you’re the author of the content you’re linking from.

We now support markup that enables websites to publicly link within their site from content to author pages. For example, if an author at The New York Times has written dozens of articles, using this markup, the webmaster can connect these articles with a New York Times author page. An author page describes and identifies the author, and can include things like the author’s bio, photo, articles and other links.

But, as many people have pointed out in forum threads and blog comments, the current implementation only works within a website. So you can link from one page on your site to another page on your site saying you’re the author of the content by using a rel=”author” tag. Well, big deal. You can already do that by the mere existence of the link itself saying something like, “click here for my author page”, or “click here for my other articles”.

Google, figure out a way for content creators to claim ownership and “authorship” so that scrapers can’t outrank them in the SERPS, please. How about using rel=”unique” and your timestamp on discovery of the document? I’m sure that would catch on pretty fast… and I know you index my pages almost before I’ve finished writing them! WordPress, Blogger and other CMS’s could add the tag by default, but for duplicate content the timestamp would be the deciding factor for who gets authorship credit.

Blogging-Underground

I’ve just completed a video review of Blogging Underground by Mike Liebner. In the video I make a few points that I want to expand on…

When you use Blogging Underground you get to create links inside blog posts and have them appear on individual blogs across the Internet, some of which are general, and some are topic-specific. It’s important to know that the BU system is very powerful. The in-content links you build are some of the most powerful links your site can get, especially if you choose rich anchor text.

You should use BU to build a few links at a time, slowly, because having many in-content rich-anchor text links appearing all over the Internet pointing at your website, especially if it’s new (see my blog post on Google Bowling), is asking for a Google “over-optimisation” penalty. So, my recommendation is to use BU links in moderation, along with other link-building systems (press releases, article submissions, RSS directory submissions, link directory submissions, blog commenting, social bookmarking etc) so that your site has a natural link profile.

Having given you that “heads up”, I highly recommend getting access to BU because it’s one of the few ways you can get the most desirable links (in content, rich anchor text) which can seriously help your ranking as part of a good link profile. Watch my video review here, and take Mike up on the special offer.

What did you think of the video review and offer? I’ll leave comments open for a few days…

Google Bowling

[Update: I've managed to get a FREE copy of Dr Andy's excellent guide to creating affiliate websites that Google loves. Just click here to download it. No squeeze page, just open the zip file and enjoy the 155-page PDF!]

I’ve been building new affiliate websites to diversify away from the Internet Marketing niche for the last few years.

In the past, getting Google to index and rank a new site wasn’t difficult. In fact, my last two IM products were to do with using datafeed content (going back several years!) and, more recently, using the power of a “future niche” to plant your website at the top of Google and stay there.

But, I’ve noticed that Google is using its penalty filters more strictly now than it used to. It seems that if you build a few “suspicious” links, with optimised anchor text, you’ll get slapped with a penalty filter before you can blink.

What interesting is that this activity breaks one of the fundamental rules of Google… a website owner is NOT responsible for links pointing to their website. The reason for this Golden Rule is simple… if it didn’t exist, the world of Black Hat SEO would spend their time and energy building BAD links to competitors to push them out of the search engine results pages (known as Google Bowling).

OK, so how can Google justify breaking its own Golden Rule? Well, it seems that the penalty is only applied to brand new sites… those whose age is measured in months rather than years. And here’s the super-smart bit… Google’s algorithm mostly prevents new sites ranking for competitive phrases… so there won’t be much “Google Bowling” going on because, by definition, trying to knock a competitor out of the rankings only makes sense if they have a valuable number one ranking.

It’s genius. You let webmasters build new websites and if they do anything suspicious in the first year (my guess for how long probation lasts), they’re on the “naughty step”. You give a free pass to any website older than a year to prevent Google bowling. You apply penalties algorithmically to avoid any manual overheads.

My guess is that you can avoid the 12-month probation by getting a link or two from older sites who have some authority and can “vouch” for your new website. If you can’t get those links, be prepared for a long wait before you can do any “real” link building.

What do you think? Am I right, or wrong? I’ll leave comments open for a few days. Also, please share this blog post if you like it… thanks!

[Thanks for all your feedback. Comments are now closed]

Google Places Results

SEO Changes For Local Searches

Have you noticed a drop in Google organic SEO traffic for any “local” search phrases, for example, “holiday cottage cornwall”?

Most likely, if you run any sites that got traffic from “local” search, you have.

It’s been a while since Google rolled out their new “Google Places” listings based on Google Maps locations in the search engine results (it was tested in Aug 2010 and rolled out in Oct 2010). At the time it caused much wailing and gnashing of teeth for anyone doing SEO on the 20% of daily searches that Google says have a “local component” because the Google Places listings jumped in ahead of the organic seo results accompanied by a huge map with pinpoints for local businesses that Google thought you may be interest in visiting, meaning a number one Google organic listing suddenly fell to roughly position five or six.

Now Google has rolled out an even more aggressive version of Google Places (see above image, click to enlarge) which positions the map ahead of the Adwords ads on the right-hand side of the page and fills almost the whole of the left side of the page with what looks like organic results but are actually enlarged listings from Places, meaning the first organic result fall “beneath the fold”.

Why are Google doing this? For a number of reasons…

First, by showing actual, verified local businesses in response to “local” searches, Google thinks it’s giving the user a better experience.

Second, the organic results which are unverified and potentially SEO’d to high rankings are pushed out and therefore the Google user avoids a potentially negative experience from a visiting an organic result.

Third, Google also sees this move as its way to compete with social media such as Facebook Pages and the new “hotpot” will take this one step further by allowing you to add friends, rate the local businesses and get their recommendations, all on your mobile phone (available as an app on the iphone and, of course, Google owns Android)

Fourth, after businesses have claimed their local listing, Google offers them the opportunity to add photos, videos, coupons, menus to their listings using their “tags” service from which Google makes money. Google is also testing the ability to automatically set up Adwords ads for the local businesses using Google Boost based on the information it has about them which would be another money spinner.

SEO Changes For “Branded Phrases”

Nike Brand Results

We see a similar “crowding out” effect when searching Google for company names, or closely related searches… the top three or four results will all be from the same website.

See the search results for “Nike”, left. On the first page of results, different pages from Nike.com command the first four places.

Google is effectively suppressing most of its search index in favour of the one website deemed “official” when your search relates to a company name, brand or slogan.

SEO Changes For “Shopping Phrases”

Shopping Results

Finally, there are the “shopping results” when you search for product names, which are fairly unobtrusive at the moment, but will they grow into another way to crowd out organic results? My guess is yes.

Google Checkout

Currently uploading to Google Base is free, but the whole trend at Google seems to be moving toward verification with charges for enhanced listings. Of course, one existing way to enhance your listing in Google Product Search or on Adwords ads is to use Google Checkout.

What These SEO Changes Mean For You

Here’s fair warning to anyone aiming for top Google organic rankings… beware of Google’s trend towards serving results from established, “verified” businesses to “crowd out” any results from new and unverified websites. In one area after another Google is showing that it wants verified, genuine businesses to be listed in preference to unverified websites. We see it from the local, shopping and brand name results down to requiring privacy policies for Adsense websites. It’s only going to expand, so take a hard look at your web-presence and see what you can do to project a corporate image.

[comments are now closed]

Over at associateprograms.com, there’s a discussion about linkwheels, so I thought I’d add my two pence.

What’s a link wheel? Well, traditionally a linkwheel is when you take multiple pages on Web 2.0 properties and link them back to your main site as well as interlinking them so that each Web 2.0 property points to the next “spoke” in the wheel. When you look at it as a diagram, the link structure looks like a wheel.

The idea behind linkwheels is that Google will index the Web2.0 properties and notice that they link to your main site, which will gain positions in the search engine rankings because of the extra backlinks from popular Web 2.0 properties.

However, linkwheels are also popular with black hat SEO people who use the fact that the web 2.0 properties have millions of backlinks to fire spammy profile links and comment links at their pages on the web2.0 sites. They think that the spam links won’t be a problem for the web2.0 site whereas they would almost certainly get a new site penalised. So, Google launched a crackdown on such link schemes and can now almost certainly find and devalue any “traditional” link wheels.

So I don’t recommend using a standard link wheel structure. What I do is take web2.0 pages, add some original content, then link them to my main pages. I don’t link the web2.0 pages to other web2.0 pages, so I’m not forming a wheel. With my system, it looks more like some people have decided to echo my news/thoughts on separate web 2.0 sites rather than me engineering a linkwheel backlinking system.

I’ve just been running a few of my articles through Web Content Studio. It’s a very cool piece of software. I’ve already written a review of it, so don’t want to go over the same ground again. Just check out the review, watch the tutorial videos and see what the software can do for you.

Hint: if you’re missing any “theme words” from your articles, WCS will help you work out what they are, then rank your page for theme completeness. It’s just one (important) facet of getting top search engine rankings.

It seems that Michael Campbell has started a new series of free videos on SEO advice. In the first one he talks about the importance of the title tag for SEO ranking and clickthrough rates. As you’d expect from Michael, it’s top quality advice…