Showing posts with label Google. Show all posts
Showing posts with label Google. Show all posts

Current Search Share 2010/2011




This is the current organic search share for the end of 2010. You can click on the image for a larger version.

Organic Search Engine Market Share - Global 2010/2011
(Rounded data in brackets)
  1. Google/Ask 86.21% (87%)
  2. Bing/Yahoo 9.06% (9%)
  3. Baidu 2.97% (3%)
  4. Others 0.68% (1%)
2010 was a busy year for search, due to Google pulling out of China and merging with Ask, and the Yahoo/Bing merger. In order for the above chart to be useful, I've merged the Bing/Yahoo and Google/Ask data, even though for much of the year these were separate search engines.

Yes, you can use this in your presentations, as long as credit is given.

Raw sources: Hitslink, Comscore, E-Consultancy, eMarketer, Hitwise, Nielson Media Research, McAnerin International Inc internal analytics data.

Redirect Issue with IIS

I'm not sure if this is just an IIS 6.0 issue, but that's where I just ran into it.

A user of the SEO Browser contacted us today because the browser couldn't load his site. That wouldn't have been so bad if it wasn't for Google ALSO not apparently being able to load his site (it was unindexed). However, browsers loaded it just fine.

It took some investigating, but I finally found the issue. The site was set to redirect visitors to the domain root to an internal page. No problem. The redirect was the default 302 that IIS uses. Ok, a bit of a problem, but not enough to stop Google or the SEO Browser from loading it.

But buried in the HTTP headers ( I used SEO Consultants "Check Server Headers Tool") I found this directive:

FollowRedirects=False; Server requested redirection, but sent no new location.

Oops. If you set a redirect (301 or 302) but then say that you should not follow any redirects, strange things happen to your site as the two incompatible directives play out.

In this case, browsers (Firefox and IE) followed the redirect and ignored the server directive saying they should not.

Spiders (Googlebot and the SEO Browser) acknowledged that a redirect existed, but also acknowledged that the server was telling them not to follow it. Therefore they did what they were told and stopped.

Naturally, the fix is a very simple checkbox telling IIS to set the FollowRedirects=False to FollowRedirects=True.

I've never encountered this issue before, but I thought I'd share, since it might help someone out there in the future. I'll see if I can have this issue detected by the SEO Browser, as well.

Ian

PS I need to blog more often. I feel a New Years Resolution coming on...

Google International PPC Ad Targeting Twist

After some testing and a confirmation from Google, I've discovered a new (to me, anyway) twist to Google's PPC Ad targeting for international sites.

In SEO, Googe decides on the country your site is relevant for first by ccTLD (Country Code Top Level Domain - like .ca, .uk, .mx, and so on) and then by IP and other methods.

Turns out this thinking also applies to PPC.

If you target a specific country, like in my case, Mexico, Google will only show ads to surfers within Mexico, right?

Wrong. They show ads to surfers in Mexico, yes. But they also show ads on Mexican tagged sites, such as sites with a Mexican IP address or sites that use the ccTLD of .mx, even if they are hosted in the US, etc.

It makes sense if you think about it. Of course, you may be thinking that this would only apply if you have chosen to advertise on the content network. You'd be wrong.

See, even if you don't advertise on the content network, there is a site that is a .mx and is not on the content network - www.google.com.mx

That's right. Even if you target only Mexico, your ads can show up to anyone anywhere in the world who happens to use google.com.mx. Anyone who needs to look at or verify geo-targeted ads knows it's sometimes hit-and-miss, but they do show up.

I first really put it all together when I was looking at traffic patterns for my Mexican campaign and noticed a bunch of traffic from the US and even India. This is normal, and is not necessarily an issue, as long as you are aware of it.

But let's say you really don't like it. There is a fix.

Instead of targeting Mexico (or wherever) you simply target physical locations within Mexico, such as cities, and lat/long areas. This forces a geolocation check and gets rid of your Japanese traffic.

Just remember that the traffic from Japan may be traveling Mexican business-people, since they would have to be using google.com.mx, and you are now excluding them.

But if that's what you need to do, then that's how you do it.

Ian

New Use for Canonical Tag - Geolocation.

I've been messing around with it, and I believe that I've found a new use for the Canonical tag - geolocation of gTLD's.

Simply park (not redirect) a ccTLD on your site, upload an HTML sitemap that points to all your pages with the ccTLD, then place the canonical tag with the gTLD on your pages.

Viola - the search engine "tags" your pages as geolocated via the ccTLD, but only displays the gTLD.

You keep your .com, but have now geolocated your site without having to host it locally.

This is a variation on a technique I've been using for a while, but much cleaner due to the addition of the canonical tag.

Ian

PPC Text Ad Formats for the Major Search Engines

I'm currently working on a very, very cool tool that I can't tell you about (yet). However, part of it (an extremely minor part) allows the creation and export of PPC text ads for all major search engines.

Although I can't share the full tool yet, it took me some work to collect this information in one spot, so I figured I'd share it with you.

Where bulk upload is allowed, the listings use the exact terms that the search engine in question uses (i.e. the heading you would use for the CSV).

"SB" is single byte (i.e. English) and "DB" is double byte (i.e. Chinese). Enjoy!

Google AdWords Character Limitations and Format

Headline: 25 SB, 12 DB
Description Line 1: 35 SB, 17 DB
Description Line 2: 35 SB, 17 DB
Display URL: 35 SB (SB Only)
Destination URL: 1024 SB (SB Only)

Yahoo YSM Character Limitations and Format


Ad Title: 40 SB, 20 DB
Ad Short Description: 70 SB, 35 DB
Ad Long Description: 190 SB, 95 DB
Display URL: 25 SB (SB only)
Destination URL: 1024 SB (SB Only)

MS AdCenter Character Limitations and Format

Headline: 25 SB, 12 DB
Description Line 1: 35 SB, 17 DB
Description Line 2: 35 SB, 17 DB
Display URL: 35 SB (SB Only)
Destination URL: 1022 SB (SB only)

ASK ASL Character Limitations and Format*

Ad Title: 25 SB
Ad Description: 70 SB
Display URL: 35 SB
Destination URL: 1000 SB

*Ask is English Language (SB) ads only.

Baidu Huobao Character Limitations and Format*

Ad Title: 13 DB
Ad Description: 32 DB
Display URL: 35 SB
Destination URL: 1000 SB

*Baidu is Chinese/Japanese (DB) only

Naver added to Robots.txt Generator

Naver

As you may know, my company offers the most popular robots.txt generator in the world free to visitors, and we are continually adding functionality.

In response to a recent request, we've added Naver to the robots.txt tool. Naver is the most popular search engine/portal in Korea. I've been to Korea several times, and (because it's my job) I always ask everyone I meet what search engine they use. It's almost always Naver, and there is good reason for it.

There is one thing I've learned doing international SEO - the Koreans are really, really good at creating useful and engaging web content. Some of the best free online games, tools and services are Korean. Google is getting their proverbial butt kicked there, and it's because they simply are not good enough. I still honestly think that Google got the idea of universal search from Naver - Naver was doing it long before Google was, and is still doing it better.

As a result, I'm kind of surprised I hadn't added Naver to the robots.txt generator already, but I guess it's better late than naver.... ;)

Ian

Google and God

I came across this sign during a recent trip to Banff, and thought I would share. Google is everywhere...


Update: Google Releases Google Branded IE7

This is a followup on a previous post complaining about my beloved Google Toolbar being lost during the upgrade to IE7 and Googles own Chrome not even allowing me the option.

I guess Google gave up on getting their toolbar to install on IE7, so they appear to have used the Admin Kit provided by MS to create a version of IE7 that has the toolbar included.

http://www.google.com/toolbar/ie7/

Works for me.

Search Engine Share 2008

I just made these for a presentation, based on the latest information I have for this year. Enjoy.



Organic Search Engine Share 2008 for North America


Organic Search Engine Share 2008 for North America





PPC (Pay Per Click) Search Engine Share 2008 for North America

PPC (Pay Per Click) Search Engine Share 2008 for North America

New Browsers and Google Toolbar

I like the Google toolbar. There, I said it.

Today, I decided to take a chance and upgrade my IE 7 to IE 8 Beta. I use Firefox for SEO purposes and IE for normal browsing, so this was a big deal for me, since I use IE a lot.

Upon installation, IE 8 promptly informed me that the Google Toolbar isn't compatible with it. A quick check on the web showed that it can cause all sorts of freezing and problems. Darn. Otherwise, it's a really, really nice browser. I was very pleased. Except for my Google Toolbar. I use this for checking my spelling. I spel reel bad. Eye knead ah spel chequer...

However, in a coincidence, I heard several people talking about the new Google Chrome browser. HAH! If anything should work with the Google Toolbar, it should be the Google browser, right? So I installed it, and went to the Google Toolbar page.

In it's infinite wisdom, Googles toolbar page detected IT'S OWN BROWSER wrongly as Firefox, but then informed me that to use the Google Toolbar I needed to download Firefox!!




Huh? That's a marketing and branding nightmare if I ever saw one. I think there needs to be a leeeetle more communication over at the 'plex. And maybe some testing, too...

Just dumb, guys.

Ian

Google Rolls out New Invoice System, and Immediately Bills Wrong People!

So, I have a client, all nice and tidy in my AdWords Pro Account. This client has a very large spend (500K+) and frankly, my credit cards simply don't have the space on them. So I set up invoicing with Google.

Google, to their credit, work quickly and are very helpful. Even so, there are delays involved simply due to processing times, etc. During this period, which is rather nerve-racking, I actually have to take out a second mortgage on my home in order to pay Google. The second mortgage also takes time to put into place, so in the meantime, I'm frantically paying Google by rotating through every credit card I have, and I even borrow a card from my brother in law.

Eventually, everything gets straightened out, invoicing kicks in, my client pays their bills, the credit cards are paid off, and my company is now stronger for the experience. All told, although it almost gave me a nervous breakdown, it was an overall positive thing.

(The reason I couldn't bill the client up front and pre-pay is another story, related to contracts and government regulations. Suffice to say that if it had been an option, I would have used it, but it wasn't).

But for now, everything is going fine. Once per month, I'm sent a snail-mail version of the invoice, but since I'm in Canada and the invoice is coming from the US, it arrives later than I'd like. Instead, I log into AdWords, download a PDF of my invoice and pay it. No problem.

Now For The Problem

At this point, life is good, I get paid, Google gets paid and everyone is happy. Until yesterday.

Yesterday, Google sends an email version of the invoice. Sounds good, right? Much faster than snail mail.

Except:

  1. The invoice was sent to my client's email address, not mine. Big mistake number one. I'm surprised it even got through their spam filters, much less got forwarded to me. They certainly had no idea what to do with it.
  2. The invoice was billed to my brother in law. You know, the guy who loaned me his credit card once on this account. He's certainly not going to pay it, and my client had no idea who this guy was.
  3. The invoice was sent yesterday - not at the time when the invoices should have been sent. By this point, I had not only received the proper invoice (when I download them, they are billed properly to my company, not to my brother in law), but I had paid it already. Why the invoice was sent AFTER THE DUE DATE and after being paid is totally beyond me.
  4. Finally, it was sent on behalf of Google Collections Canada. Now, we don't work with Google Canada due to some very bad service we had received last year. When I contacted the person listed in the email, she said that her name had been added automatically and she had not directly sent the email. So the email was sent FROM the wrong person (and wrong country)!

So, Google invoiced the proper amount, but did so:

  • To the wrong email address
  • To the wrong billing contact
  • At the wrong time
  • After it had been paid
  • From the wrong country and collections department

Not exactly a stellar start to the new system. If you are an AdWords customer, be very, very careful about this new system, and I strongly recommend you check everything. Twice.

Just like Google should have.

Ian

The Biggest Information Integration Mistake

Has this ever happened to you?

You phone the customer service line of a large company (bank, telephone, Google AdWords, etc) and get an automated voicemail system.

This voicemail system then asks you for information "to help you get to the right party". You then dutifully punch in your telephone number, bank account number, or client number, whereupon you are transferred to "the appropriate party".

The "appropriate party" finally answers the phone and then they ask you a question. You know what question they ask, don't you?

They then ask you for the SAME DAMN INFORMATION you just gave them.

Sometimes, they then transfer you to one or more other "appropriate parties", who all, without fail, ALSO ask for the same information.

Why?!!!!

In this day and age of analytics and customer tracking, why do companies consistently drop the ball with one the the most important visitors they can possibly have - the current client who needs help or is unsatisfied?

Although privacy laws make it clear that you should not be handing out your customer information to just anyone, they also acknowledge that you are entitled to (and must) collect, pass on and use the information necessary to complete the transaction requested by the visitor. Otherwise there is no point.

It is the height of rudeness to ask someone to give up some of their privacy to you and then for you to treat that information as unimportant and disposable, in the process inconveniencing them.

It's kind of like asking someone for their business card, looking at it, then throwing it away in front of them. Followed immediately by asking them to give you their phone number again. It's stupid, wasteful, inefficient and rude.

This also applies to online forms - deleting a customers form data and forcing them to start over just because they had one typo is bad usability, bad form, and bad programming.

Rules for collecting personal information:

1. Create and enforce a privacy policy that is designed to both protect your visitors and to allow you to do the job your visitors want you to do, safely and securely.

2. Never collect personal information that you do not need to do the job (any additional information should be anonymous and aggregate), and make sure that it's with permission.

3. After you get permission to collect personal information - use it to help the visitor. That's why they gave it to you in the first place, and is the basis of the permission.

Ian

Google Agrees - Hostile Takeover Attempt

I note with some vindication that Google apparently agrees with my assessment that Microsoft's bid for Yahoo is hostile.

Admittedly, Google has a vested interest in promoting the idea that a combination between MS and Y! would be unfair, but, honestly, tough noogies. Competition is good, and I can't see Googles viewpoint as anything other than a cynical attempt to keep and increase their market share by pretending to actually care about poor little old Yahoo. This is the first time they've ever publicly shown an interest in the search market outside of themselves that I can remember.

That doesn't mean they are wrong about the move being hostile, though.

http://googleblog.blogspot.com/2008/02/yahoo-and-future-of-internet.html

It's rather annoying that no one else in the industry seems to care or know about this issue ,and that my thoughts on the subject were ignored in favor of idle posts talking about trying to name the new potential search engine instead.

I'm really starting to wonder why I bother...

Ian

Adwords Editor, Vista, and Error 2739

I've been having problems with the installation of Google Adwords on my Vista machine for a while now. I've been putting off fixing it because I hire people to do PPC for me on a day to day basis, so it's not a huge issue. But today I needed to fix something for a client ASAP, and it was enough of a job that I needed to do it personally, and I needed to use the Editor.

Well, it turns out that Googles installation package doesn't work on Vista. It won't install, it won't UNinstall, it won't repair. Nothing. It's very annoying. I'd blame Vista except nothing else gave me this problem, so it's Googles fault. Oh, heck, I'll blame Vista as well.

Anyway, courtesy of Brent Crouch, I found the fix:

  1. You need to make sure you are logged into Vista as an administrator.
  2. Open Command Prompt and change into the C:\Windows\System32 directory.
  3. Type: regsvr32 vbscript.dll
  4. NOW PRESS Enter (I got an error after this command, but it still worked.)
  5. Type: regsvr32 jscript.dll
  6. NOW PRESS Enter (I got another error after this command, but it still worked.)
  7. Now try the install again. It should work this time.
I didn't have to change directories - I just clicked on the "Run" choice in the start menu and typed in the 2 commands. It worked fine. I was able to uninstall a previous version of Adwords and install the latest with no problems.

Thanks Brent. And a big raspberry to both Vista and Google. Mostly Google.

Note to Google: the "eternal beta" trick only works online - you might want to test downloadable software from now on - just a thought. It makes you look silly when you can't even get your software to install properly and every other piece of software I've installed has no problem...

Ian

Google Webmaster Tools Geotargeting Added

Google just unveiled a new tool in their Webmaster Tools - the ability to set geotargeting for sites, even down to the street level if necessary. Naturally this is very cool, and interesting to those of us that deal with geolocation issues all the time.

The tool will not allow you to override the ccTLD, so you can't declare your .ca site to be from the US, for example, but if you have a gTLD like .com, .net and so forth, then you can. For many people, this is an excellent method for accomplishing what they want.

If you want your site to have multiple countries, you would just create sub-domains - france.domain.com, for example, and then get geolocation for that sub-domain to France. You can do this with multiple sub-domains each to multiple countries. No site can be geolocated to more than one country, however.



Naturally, this will bring up the next question - "Hey Ian, since Google has this now, why would someone need your IPGeoTarget tool™ "?

Well, as long as you only cared about Google and used the Webmaster Tools, then you would not, and I would not in good faith recommend paying for any system as long as Google has a free tool that does the same thing.

However, since this only works for Google, you would not be able to geolocate for Yahoo, MSN, Ask, or other search engines, so that's certainly a limiting factor.

Second, this currently only works for entire sites (including sub-domains), but not directories or pages. Many companies have set up their sites like this: domain.com/canada/ and Googles system would not help them in this case. This can cause big issues if your CMS doesn't support cross-site editing, or if you are looking at the possibility of 301'ing thousands of indexed pages (along with the related drop in traffic/rankings during the switchover).


Finally, not every company uses or likes to use the Webmaster Tools (though I admit they are pretty useful myself).

I admit the timing is a little annoying - I have no idea if it's a pure coincidence or if someone decided to speed up the announcement because of the interest in IPGeoLocate. Either way, it was just a matter of time before Google did this, since webmasters have been clamoring for it for some time, so I'm not worried.

My recommendation would be to use it right now if your site structure allows - Google does drive a fair amount of traffic, so it's not like it's a waste of time - and then add IPGeoLocate to the mix once it's available for your target country - hopefully next month.

Ian

A Patent Rant.

Andrew R H Girdwood is a very smart fellow. He's already posited that my new IPGeoTarget system is likely based on proxying, and of course he's correct - it really can't be done any other way.

Well, there is one other way - one could alter the geographic tag for an IP, either directly at the IP Mapping provider or after the fact, like a private database. Both of these have issues - mainly the problem of shared IP addresses. The real answer to the whole geolocation mess is to identify *domains* (or better yet, pages and directories) as geographically located using something additional to a ccTLD. I'm leaning towards either a metatag or an entry in a robots.txt file, myself.

Registering with Google would only help for Google, and not everyone wants to register everything they do with Google, particularly since they have been acting less and less like idealists, and more like a shareholder-owned corporation (unsurprisingly).

In particular, companies located outside of the US are hesitant to give additional information to companies (like Google) that are easily targeted by US laws that may not have their best privacy interests at heart. The feeling sometimes outside the US is that anything the Chinese government can order Google to do, so can the US government, and better, since that's where the head office is. It's not that they actively distrust them, it's just that international companies tend to not get to where they are by being blindly trusting with their data.

So until the search engines get together on this issue, it's going to continue to be an issue. Even afterward, it would still be nice to speed up connection times to visitors without having to physically move a site - there are reasons other than geolocation to use this type of technology.

Anyway, Andrew also posts a worry that I may be trying to patent a technique that's already been done, or that tries to lock down common internet technology. I'll directly address that, since it's a legitimate concern and he's right to bring it up.

1) To the best of my knowledge, it's not covered/prevented by prior art (though of course almost everything on the internet has some sort of prior art connection simply by being on the internet), and

2) I'm not trying to patent the concept of a proxy, IP address, "Click to buy" button or anything that basic or obvious. Though you'd be surprised what can be patented nowadays.

At least, I hope so on the second item - every patent applicant has either nagging doubts or is delusionally self-important. I think my teen-aged delusions of infallibility have been quashed out of me after years of being in flame wars on forums, working with non-profit organizations, and having a family. I guess it's up to the patent office to ultimately decide, and for now I'm leaving it to them. The point is that I'm acting in good faith and trying to make things better, not prevent competition or cash in on anyone else's hard work.

Beginning of Patent Rant

I'm changing topics now - this has nothing to do with Andrews post. I'm just on a role and am too lazy to start a new post. Besides, if you read this blog you are probably used to really long, wandering posts by now. It's because I type exactly like I talk.

I've been asked several times what I would do if the patent office said no to my poor pending patent proposal for pinpointing positioning (how's that for an alliteration?), and I'm drawing upon my previous experience as the patent manager for a company with 72 patents worldwide for the answer.

The answer is that it doesn't matter. Surprised? Then you don't know as much about patents as you think you do. Experienced patent lawyers would not be surprised by my attitude (though they may be dismayed at the thought of losing all that money made during the process), and I'll tell you why.

At the end of the day, a patent is simply protection for a business idea, so if you can't make a business work from it, you've wasted your time on what is basically an ego trip. So patents don't matter, business concepts do. There, I said it.

It's more important to have a legitimate and profitable business than a patent, and some people (notably inventors, dreamers and narcissists) never seem to really get that, which is too bad, because then they get screwed by businesses that may not be as creative, but have a stronger drive to succeed and profit.

It happened to the original inventor in my previous company - he's broke now and doesn't even own any shares in his own company anymore, which is still going strong. The last I heard he was in hiding. That's what happens when you trust venture capitalists to run your company for you while you hope to rake in the royalties.

I was part of the "cleanup crew" hired after his original company imploded when the VC's exercised their "exit strategy", and I learned a lot from the experience.

In particular, I learned 2 very important lessons from the whole mess:

1) You don't have control over the patent process - other people do. Lawyers, competing companies with their own patents, owners of prior art patents who think they also own everything even slightly related to their own patents,, law firms that buy vague patents and then make money suing people at the drop of a hat, "free spirits" who don't think anything should be patented/copyrighted/trademarked, naysayers who think everyone else's ideas are always wrong, friends who are worried you might be hurt, and, of course, the patent office. And not just the patent office, the particular patent examiner you get. Then the whole thing starts over in every single country in the world that you try to patent in. I'm surprised anyone bothers to even try anymore!

2) You DO have control over your business - unless you give it up. Too many people think that if they get a patent then they can sit back and let the royalties roll in while others do all the work. Well, it's not that easy in the real world, which is why usually the only patent holders you meet that are rich are those that are astute business people, and it was their business dealings that made them rich, not the patent. IBM makes tons of money on royalties from it's patents, but it's not because they sit around waiting for people to send them money - they work the angles and earn the royalties actively. The patent process itself can make you broke very quickly. Therefore, forget the patent, and focus on the idea. Is it a good idea? Great! Go make it work. Who cares if you don't have the patent yet? The fact that you are at "Patent Pending" generally scares off those that care about such things, and for those that don't, they don't care about whether the patent is granted or not.

I am personally aware of a well-known person in the SEO world with a patent that Google and Yahoo are both flagrantly in violation of. It didn't seem to stop them at all. This person knows if they sue it will likely be more trouble than it's worth. So really, what is the patent worth? Once again, it's not the patent, it's the business. Learn that lesson well before you decide to patent anything.

(I don't think this person is trying to keep this a secret, since otherwise they would not have done something as public as a patent, but I'll let them identify themselves or not as a courtesy, just in case.)

In the meantime, I'm going to proceed on the basis that even if the patent office disagrees with me, the usefulness of a company being able to open an account at IPGeoTarget.com, type in their URL or domain, choose a target country, and then be geolocated to that country with little other fuss or muss, will be a viable business model.

As a matter of fact, I'm about to spend a whole bunch of money on exactly that. The patent is icing on the cake and a nice angle in a sales pitch or press release. But a patent is not a business. Work on what you can actually control, and for the rest, do your best to set things up so they end up in your favor, then forget about it and deal with things as they come.

Bottom line: Thinking of applying for a patent? Make a detailed business plan first, because that's what it's really all about.

Ian

AdWords PPC Geotargetting / Language Setup

"When geo-targeting ads by a client's country, what is the best practice for language targeting?"

As a general rule of thumb, your ad should be in the same language as the SERP the searcher is looking at.

Let's say you target Korea, to use an example. If you are just beginning, then it would be best to geotarget Korea and also target the Korean language. Although many Koreans read/speak English, it' s jarring to see an English ad when the rest of the SERP is in Korean. It makes it stand out, but in the wrong way. Usually they decide that the company is clueless and "doesn' t understand Koreans". I've had many discussions with Koreans on this very topic. The same also applies to Chinese, and especially to Japanese.

If you wanted to be more accurate and do a really thorough job in the market, you could do the following (though it's more work and for some markets isn't worth it):

  1. Geotarget Korea (or whatever country you are looking at)
  2. Create a KeyWord list. Separate out the keywords that are the same in English and Korean ( i.e. " Samsung") from the pure Korean words.
  3. Anything that is pure Korean, target Korean language only.
  4. Anything that could be both (and would result in a SERP with both English and Korean in it) you would use as two different groups – one targeting English with English ads, and one targeting Korean with Korean ads.

Some other observations:

  1. A single English word ("Samsung" ) could be equally in either language, but multiple English words "Samsung office in Seoul " is usually (though not always) an indication that the target language should be English. This also works the other way – one English word in a Korean phrase is probably Korean.
  2. If in doubt, use the official national language of whatever country you are in, or the most common language of the region if there is more than one. For example, in Canada, you would default to English for western Canada and French for Quebec, unless someone indicates that they are looking for a language specific Keyword.
  3. Due to the different character sets between Asian languages and English, this might seem more complicated than it needs to be (you are normally safe in assuming any keyword written in Chinese characters has a preference for Chinese ads, for example) but as a best practice it 's a good idea to language target as well as geotarget, especially when you begin to work with multiple languages that share characters (English/Spanish/French or Chinese/Japanese/Korean).

Practical Final Answer: Start off with Korean language ads geotargeted to Korea and the Korean language keywords, including the dual-language Korean list. See how that goes. If it goes badly, you are unlikely to fix it by adding English to the mix , and it will just complicate things.

If it does well, then add the dual-language English list to the mix. In this case, you would just create a second campaign, but this one geotargetting Korea but only the English language , then use the dual-language English words. In this case, Google (for example) would not treat that as a duplicate, but would trigger the English ads for searchers that had indicated a preference for English, and Korean for those who indicated a preference for Korean.

This type of system is especially useful when you have products and numbers involved – for example, the "SGH-L760 " from Samsung is the same search term in any language – Korean, English, Chinese, Japanese, etc. You simply can' t just geotarget it – you have to also target the language in order to trigger the correct ad.

I hope that helps,

Ian

Google Proxy Hack - Part 3

Well, my client is now ranking again for negotiation training and the proxy has disappeared. The funny thing is that it was fixed before we were even able to start blocking IP's.

If you are from Google and brought this to the attention of those who fixed it - thanks :) Hopefully this is the start of this entire issue being fixed for everyone.

Ian

Google Proxy Hack - Part 2

As I mentioned in a previous post, Google is susceptible to a proxy hack that can wipe sites off the SERPs, and do all sorts of other nasty things, as well.

Naturally, my client is not the only one with this problem - it's a well known issue with Google and has been for almost a year. Based on how long it took them to address the 302 Hijack, I'm not holding my breath for a quick fix on Proxy Hijacking, either.

Dan Thies posted a great article on this issue, so I'll link to it rather than repeat the information in it. The takeaway is that it's a real issue, and it's not that easy to solve from a victims standpoint.

The easy fix is to block the IP of the proxy. This only works if the issue was accidental and only one proxy.

If it's a case of someone deliberately gunning for you, then don't expect any help from Google. In this case, you need to get a bit more fancy. There are several different methods, all of which could be countered, but Jaimie Sirovich wrote a nice script to do what is basically reverse cloaking, feeding the normal pages to search engines and putting noindex,nofollow on pages sent to everyone else (including, hopefully, proxies).

I hate proxy hackers. Proxies are a useful part of the internet, and people who abuse them in this way are NOT "pointing out flaws in Google to make them better", they are simply people who don't understand that pissing in your own well water is stupid, and they deserve nothing but contempt.

Ian

Exploiting Googles Proxy Weakness

I have a client. He does negotiation training and has clients like IBM. His website is http://www.negotiationdynamics.com/. He does the actual SEO on his site and phones me for help when needed, which is a business model I rather like.

Today, I got a phone call from him, and he told me he'd disappeared off of Google. Worse, his site seemed to have been hijacked. When you type in the search negotiation training, the result that used to be his now looks like this:
You will see that instead of the URL being negotiationdynamics.com, it's now a search result from a known Open Proxy called firewalldown.net. Naturally, efforts to contact these guys and tell them to put a damn robots.txt file in that would exclude the spidering of results has not been successful.

But wait! There's more! If you sign up now for the Google Proxy Hack we'll throw in the following for FREE:

First, Google will cache (and rank) your website (or that of a competitor you don't like):


BUT, when hapless searchers attempt to connect to the site, they are redirected to a domain reseller! Cool! Google R0X! (rolls eyes):



It's not that I blame Google for being gamed - that can happen to any company approaching the size and influence of a public resource. But I *do* blame Google 100% for indexing obvious and PUBLIC proxy results. It's sloppy programming and poor usability.

Proxies are all over the internet. Many are used for perfectly benign purposes. It's simple to identify them. The answer is not to pretend they don't exist, or to attempt to ban all of them. It's certainly not to try to find and kill proxy results one at a time by hand as they are reported, which apparently is Googles current method. The answer is much simpler.

Here is a thought: don't index URLs that have other URLs in them as a variable. Or is that too complicated? This could be done in like 20 minutes. But it hasn't been. For shame.

Ian