Google gets funky in Taiwan

Today's logo in Google Taiwan

Update: Ha-ha! Said image is also now pretty much world wide! Thanks timezones.

Be Social: del.icio.us, Digg, Furl, Reddit

CPC determined by visit history

I seem to be having a t-shirt flavoured Gmail AdWords week. This is Google's targetting system; I clicked on one and so now their system is tempting me with more t-shirt goodness.

This could be my favourite AdWord of all time.


The site in question can be found selling Engrish t-shirts here. Don't go clicking on the poor site's AdWords campaigns.

The ad reads:

Engrish.com can help you with our pleasure. Let's shopping t-shirts!

The t-shirt AdWord which first caught my attention this week was for Torso Pants. I liked the site so much that I bookmarked it using my Google toolbar. I'll buy something later.

Here's a thought; I bookmarked Torso Pants via Google. Google continues to show me AdWords for Torso Pants. If more people actually used Google's bookmarking feature then I, as an advertiser, would want an AdWords option which let me decide whether or not to show my ads to people who had already bookmarked me.

I might decide - they know about my niche store already; I don't want to ad serve them.
I might decide - I want to remind them that I'm here; I do want to ad serve them.

In fact, you can take that concept and simply throw away the Google bookmarking feature. Wouldn't it be a good idea if you could set CPC based on whether the searcher had been on your site recently, some time in the past or never before?

Disclaimer: I'm heavily NDA'd by Google but, at the time of this post, I've never talked to them about CPCs determined by visit history. If I had knowledge of CPC/visit history features then I wouldn't be blogging this!

Be Social: del.icio.us, Digg, Furl, Reddit

Google blocking Yahoo Pipes - again (again)

We've been here before but Google seems to be blocking Yahoo Pipes again.

Here's the catch. I think Google likes Yahoo Pipes. Google's even demonstrated how Yahoo Pipes can be used. When I pointed out that Feedburner was blocking Pipes I even had MC pop over to the blog to make sure everything was okay.

So I don't think Google intends to block Yahoo Pipes. I think Google constantly fights off RSS scrapers and sometimes Yahoo Pipes gets caught in the crossfire. The screen grab above shows how it's the blogsearch results that are being blocked.

Be Social: del.icio.us, Digg, Furl, Reddit

Google's countrty faux pas

I think this is fairly well documented but Google's "Did you mean..." suggestion can really commit some clangers. This screen grab was sent to me by an annoyed Austrain bigmouth!


My personal 'grrr' at Google is when they personalise my results but insist on telling me that they've 'personaliZed' my results. No Google, no. If you personalise my results then you spell in British English.

Be Social: del.icio.us, Digg, Furl, Reddit

Google: With Safesearch On

File this under; "I think this is new".

Google's now telling me when safesearch is on or not for normal text SERPs. I imagine this becomes more of an issue as Universal Search ramps up and the basic web search starts to include more images and video.

alternatively, I was doing some work for a cosmetic surgery client on Friday and flicked the safesearch settings on Google Image search on and off a few times to see the difference (wow; what a difference!) and so the appearance of this safesearch banner may well be due to that!

Be Social: del.icio.us, Digg, Furl, Reddit

Google black for Earth Hour

It's after 12 here in the UK and Google's doing something a little special... it's turned off the lights for Earth Hour.

Google's page on Earth Hour says;

On Saturday, March 29, 2008, Earth Hour invites people around the world to turn off their lights for one hour – from 8:00pm to 9:00pm in their local time zone. On this day, cities around the world, including Copenhagen, Chicago, Melbourne, Dubai, and Tel Aviv, will hold events to acknowledge their commitment to energy conservation.

So I'm not sure they've got their timing exactly right. It's after midnight. Not after 8:00pm.

Be Social: del.icio.us, Digg, Furl, Reddit

What Facebook Germany looks like - and problems

Hopefully this isn't another 'me too' post about the launch of Facebook Germany.

I thought it would be fun to see the choice Germans face when they log into Facebook today. Pick your language.

I'm not actually in either of the bigmouthmedia German offices. I'm in Gatwick, London, England (oh! why am I flying the day the 'storm of the winter' hits the south cost?) and am on a T-Mobile hotspot.

I've already touched on Google's problems with these 'annoying' hotspots with confusing IP geolocation so I thought I should point out that sites like Facebook are equally effected...

... except, well, does Facebook has less of an excuse? I've never had a German ad shown to be at Facebook. My profile clearly says I'm United Kingdom based. Is that language option box a compromise between my profile and my current IP address? I suspect not... I suspect Facebook is simply checking by IP geolocation this week.

Be Social: del.icio.us, Digg, Furl, Reddit

SES London: Click Fraud Issues

I'm just out of the Click Fraud session where we had Shuman Ghosemajumder from Google, Jon Myers from MediaVest and Andrew Goodman from Page Zero Media presenting.

I scribbled notes on Google Docs. The whole conference centre is safely blanketed with wi-fi.

Let's start with Google. While running through the different click fraud strategies (attempts to do it, rather than prevent it) Shuman noted that there were regional fashions.

Automated approachs such as click bots and botnets are more common in Eastern Europe than elsewhere in the world.

Click farms (manual labour) are more popular in India and China.

Pay-per-click sites (in which people are paid to click on ads) are more common in United States and Canada than elsewhere in the world.

Shuman also noted that Google disqualifies just under %10 of all clicks as fraud. That's a huge amount of clicks for Google to refund. Although... Google doesn't actually refund most of those clicks. The vast majority of click fraud, Google says, is caught by their algorithms at the point of click and the advertiser is never charged in the first place.

Google does engage in reactive and manual investigations - these are the type that are set up with a significant client complains. Here they'll use human skill as well as algorithms to examine what had happened (rather than catching something live).

Less than 0.02% of click fraud is found via these reactive manual investigations.

Here's a breakdown of Google's anti-click fraud steps:

• Proactive
• Filters
• Automated algorithms which filter out in real time
• Analyze all clicks
• Accounts for vast majority of invalid click detection
• Offline Analysis
• Automated algorithms and manual analysis
• Focused on the adsense network
• Accounts for a smaller percentage than the filters
• Reactive
• Investigations
• All advisers inquiries are investigated by the traffic quality team
• Invalid click percentage is a negligible slice compared to #1 and #2
• Relatively rare (<0.02%)

Google also pointed out that estimates for the total click fraud volumes are tend to overestimate the problem. If you sign up for a £10 AdWords account - ie, a daily cap of £10 then someone could send £1,000,000 worth of invalid clicks then you'd still only pay £10 (if Google doesn't catch it - which at that volume they would). So, to you it's £10 fraud but some of the "total value of click fraud" estimates would quote the £1,000,000.

Rushing off to the next sessions; updates to follow.

Be Social: del.icio.us, Digg, Furl, Reddit

Google's problems with wi-fi hotspots in the UK

There are three main providers of wi-fi hotspots in the UK;

• BT Openzone
• T-Mobile
• The Cloud

BT Openzone and T-Mobile have an agreement which allows their users to use each other's networks. I'm a BT Openzone user but it is typically a T-Mobile zone I end up logging in through (T-Mobile provide the wi-fi in Starbucks, for example).

Here's the catch. T-Mobile is a German company. The servers that support the wi-fi clearly have an IP address that Google associates with Germany.

If you look at the iPhone / iPod Touch's unique Google Mobile screen there's no easy way to change country either. I use my iPod Touch a lot on wi-fi and frequently have to struggle to de-German my results.

There are some other twists too. National Express is a very large travel company over here in the UK; they run thousands of buses and coaches (think Greyhound) and many important train services too. In fact, National Express recently won the franchise for the important East Coast line (that hooks London into the east coast of the United Kindgom) and brought free wi-fi to all their passengers.

Free wi-fi on long train journeys is great news! Google results in... Swedish are less hot. Google currently associates National Express' wi-fi with Sweden.

It isn't just Google's search results which are effected. Google does a lot of content changes that are aligned with geo-detection. For example, I made a post to this Blogger account yesterday from a National Express train and had a lovely Swedish GUI to content with.

Google's struggles with correctly identifying the location of the searcher/ap user seem particularly noticable right now. This is a problem is only going to increase as more and more people access the web over mobile products. My ideal scenario would allow me to pop out of our office in Munich or Paris, saunder down to the local cafe, pull out my iPod Touch, connect to the wi-fi and get to English language content despite the fact I'm somewhere in mainland Europe.

Be Social: del.icio.us, Digg, Furl, Reddit

Scientology, Googlebombs and the Algorithm

There is a discussion as to whether certain Scientology search results are a Googlebomb or not.

Google announced they'd tweaked their algorithm to cope with the Googlebomb not long before SES London last year. We were lucky enough to have Matt Cutts and Vanessa Fox along for that one. During SES London a number of high profile SEO speakers/bloggers called Google liar, claimed there was no such algorithm and that it was all manually done.

It was one of those times I felt like standing up, shouting over to the speaker and getting them to clairfy that this was their personal opinion and not a known fact.

When I spoke to Google about this - pointing out what was said - they were clearly annoyed/upset. I can understand why they don't like being called liars!

I do think the Googlebomb fix is algorithmic and here's one simple way it could work.

1. The algorithm tends to put media coverage discussing the Googlebomb attempt above the actual Googlebomb when the algorithm is in effect
2. Media coverage is, I theorise, the key.
3. The algorithm picks up on the possible Googlebomb once enough trusted news sources call out a search result as a Googlebomb
4. After all, there are probably thousands of 'invisible' Googlebombs but it only becomes an issue once the results become 'political' or a talking point.
5. The very same trusted news sources which are used to identify the Googlebomb attempt are then easily upgraded, algorithmically, with a bonus weight appropriate to their own influence and the weight of the Googlebomb target page
6. The weighting ensures the news sources are ranked appropriately among themselves but out rank the Googlebomb target page

If this is an algorithm and if enough news sources talk about the dangerous cult search then we can expect the Google results to change fairly quickly.

Be Social: del.icio.us, Digg, Furl, Reddit

Spamming with Google Docs

I use Google Docs a lot. I'm rather alarmed by the fact that anyone can pop a Google Doc into my document list. Here's an example where I send a document into an email address I don't have in my Google Contacts list - and email address which, I assume, Google has no idea I can control.


The first step is easy enough. Create the document, share it and then opt for the "skip invitation part".

Then look what happens.


The document turns up in the documents list for the email address I've targeted.


The document itself could contain any message I wanted. This one has a hyperlink but it could also have an image.

Clearly, this isn't a problem right now. Google's fighting to establish Google Docs as a possible replacement/supplement to Office and that's probably why it is so easy to fire around Google Docs, Spreadsheets and Presentations. It seems more than likely that Google will have to crack down on the invitation process at someone point.

I could draw connections between this and Yahoo Groups. In the old days Yahoo Groups moderators would add just any old email address to their mailing list and then use the list to mass mail spam.

Be Social: del.icio.us, Digg, Furl, Reddit

Kevin Fox leaves Google

Leaving Google is different than any other job I've left. Joining Google in 2003, it was the first time I took a job without knowing at the outset the reason I'd eventually leave the job (even if my employer didn't), and so it's strange to have found success there and yet feel a need for greater fulfillment sufficient to pull you away from what's generally recognized as the best workplace in America. It's even stranger that Google is the first place I've ever worked where I feel that I'm part of the company as opposed to working for the company.

Thoughtful comments from Kevin on his last day at, as he says, the Big G.

Kevin is a user experience designer behind Gmail, Google Calendar and the redesign that was Google Reader 2.0. These are all fantastic applications and I think the change from Google Reader 1.0 to Google Reader 2.0 is particularly noticeable and illustrates just how good Fox is. He's off to join a small start up like so many other ex-Googlers.

Be Social: del.icio.us, Digg, Furl, Reddit

Chris Di Cesare: YouTube's New Marketing Director

Here's a name all gamers will know - Chris Di Cesare. Well, perhaps gamer marketing geeks anyway.

Chris Di Cesare was the director of creative marketing at Microsoft. Simply put he put much of the ooomph into the hype around Halo 3, the Xbox 360 and even the first Xbox.

He's not at Microsoft any more. He's surfaced at Google and will be YouTube's first director of marketing. Yeah, like YouTube needed any more status with the hipsters of the web? :)

Be Social: del.icio.us, Digg, Furl, Reddit

Google blocking Yahoo Pipes - again

Did you know that Google has a bot which ignores robots.txt and does so defiantly? It's true. Google's RSS grabber, Feedfetcher, ignores robots.txt as Google reasons a human decided to publish the feed and a human has decided to request the feed. It's all explained over at the webmaster help center. I actually think Google's made the right call here, although it means you can't slam the brakes on an RSS by slapping up a robots.txt block and I'm beginning with this just to set the precedent.

I like to think one of my real scoops this year was when I noticed that Google seemed to be blocking Yahoo Pipes. Only yesterday I noted that I was disappointed that Sphinn didn't like the story but pleased that Wired writer Betsy Schiffman had.

I do believe that this blockage was temporary and accidental. Google have said complementary things about Yahoo Pipes before and you can use Yahoo Pipes to take data from Google Base. In fact, Yahoo Pipes and Google Base have been a featured project on Google Code.

In a quirk of timing, bigmouthmedia colleague and Wonga World blogger, Chris Cathcart pointed out that Google's Feedburner is also blocking Yahoo Pipes.


This time the blockage is certainly not an accident but is a human controlled decision. Why would Feedburners want to keep their RSS out of Yahoo Pipes? One possible answer is that although the publisher is happy to distribute content (or teasers) in a feed they don't want that content to be sliced, diced and mixed up with other content. One of the ways I use Yahoo Pipes is to monitor dozens of feeds but only alert me when a story is gaining a critical mass, this means I don't need to manually review all those feeds nor even look at any adverts inside them.

Here's the plug for Wonga World! Chris is our Senior Strategist in the Finance vertical. He's years of experience working in banks and digital marketing. In fact, he spoke at SMX London this year. Wonga World is written with that savvy financial sector bias which is why he gave me this 'search only' lead. What a nice man.

Be Social: del.icio.us, Digg, Furl, Reddit

Top 10 List of Things Not To Have Expanded Broadmatch On

We had a spot of good news today which inspired me to put together this top ten list!

• 10. Safety Equipment
  • (Goggles are not an expanded match for safety rope)
• 9. Jobs
  • (Police jobs are not an expanded match for illegal jobs)
• 8. Celebrity news
  • (Simon Pegg is not an expanded match for Johnny Depp)
• 7. Religious terms
  • (Voodoo is not an expanded match for Dunfermline Athletic)
• 6. Books
  • (Lord of the Rings is not an expanded match for Lord of the Flies)
• 5. Places
  • (Edinburgh is not an expanded match for Sydney)
• 4. Dates
  • (1966 is not an expanded match for 2008)
• 3. Names
  • (Robin Richmond is not an expanded match for Batman)
• 2. Dating
  • (Men are not an expanded match for women)
• 1. Google
  • (Money is not an expanded match for quality)

Apologies for the obscure football joke! Expanded Broadmatch is a (fairly) new twist from Google's CPC system. Google now pretty much matches X with Y by default on broadmatch (aka the new extended broadmatch system).

Be Social: del.icio.us, Digg, Furl, Reddit

That was Search in 2007

The talented editor-in-chief of the bigmouthmedia newsletter has put together a great summary of the year's news in Search in 2007: the year as reported by the news hounds at bigmouthmedia.

As you might have come to expect from reading this blog we've included a statistical review too. Here are some of the cherries:

• over twice as many stories on Google as on Yahoo!
• nearly four times as many stories on Google as on MSN
• three times as many stories on MSN as AOL
• and nearly twice as many stories on AOL as on Ask.com
  

Thanks YS!

Be Social: del.icio.us, Digg, Furl, Reddit

Google restores dates to Web Clips

Only two days I go I noticed that Google had removed dates from web clips.

Well. They're back.

New:

Be Social: del.icio.us, Digg, Furl, Reddit

Google drops dates from Web Clips

Back in November I noted that Gmail can be awfully slow with web clips and had a 'headline' which was 6 days old.

Google's solved that problem.

They've simply stopped reporting when the story broke. Hmm. Oh well! I suppose that solves the problem in some ways!

Old:


New:

Be Social: del.icio.us, Digg, Furl, Reddit

Google at the Great B2B Marketing Debate

On Tuesday I was at The Great B2B Marketing Debate hosted by the B2B Marketing magazine and Mardev. The topic was "Is SEO eclipsing pay-per-click for lead generation?"

This was one of the events that I end going to because I must have said "Hey, that looks like a fun debate" and then a few days before the event itself someone in our marketing department tells me that a) I'm going, b) I'm speaking, c) They require a written speech ahead of time and d) I'm for the motion.

It just so happened that a few days before this event I was in New York. Umm. Anyway; I hope that the speech I put together was good enough to interest most of the audience and as it happen - we won the debate.

I wasn't really for the motion. SEO and PPC are different and simply have different strengths and weaknesses. A good blog post which explains this would be Lisa Ditlefsen's SEO is like buying a house whilst PPC is like renting. Lisa Ditlefsen just also happens to be the B2B Marketing Newcomer of the Year.

So when I say we won the debate and then say it was Lisa and myself arguing for the motion then I hope the bigger picture starts to fall into place.

It wasn't an easy debate to win (and not just because we were arguing for the purposes of the debate - not because either one of us thought PPC is loosing traction) as we were up against Simon Norris a co-founder of Periscopix and Stuart Small the Business and Industry Leader from Google UK.

What caught my attention at the start of the evening was the attention that Google was giving it. You don't tend to see Google at many events in the UK - not even the big trade shows. I was surprised to see six other Googlers on the attendee list at the start of the night. In truth I don't think all six turned up (Christmas shopping on Oxford Street was too tempting, I bet!) but Google did sponsor the event and provide quite a lot of reading material.

I jotted down some quotes from Stuart which I'd like to share:

• For every 100 searches - 20 of them are a PPC click
• 20% of searches are unique - that's not seen in the last three months
• Google Checkout buttons increase clickthrough rates

It's not a surprise that the (still relatively rare in the UK) Google Checkout button increases clickthrough rates but perhaps slightly more surprising to hear Google pushing that so directly at business decision makers.

I was interested by the three month time limit on what defines an unique search though. Three months is clearly the time which Google keeps data for analysis on search queries (though Google Trends or the AdWords Traffic Estimator chart further back).

Be Social: del.icio.us, Digg, Furl, Reddit

Has an Iron Mountainview become more attractive to Google?

Just a quickie. Back in July I brain stormed the idea that Google might look at buying Iron Mountain.

This was my attempt at a left field prediction that no one else has made or would be likely to make... without it being a stupid prediction. At the time I listed a few reasons why, maybe, perhaps, it could happen. I talked about Iron Mountain's new data handling system but mainly this was a way for Google to get into the business of organising a whole lot (a mountain) of offline information.

Iron Mountain has just become the holder of some cracking 'online' information. I think this may make this "left field prediction" a little bit more likely. ICANN has appointed Iron Mountain as the data escrow holder of domain information. What does this mean? This means Iron Mountain looks after the details of who owns what domains. It is the third party who keeps the records should the main parties (such as registars) get into a mess, a fight, or some other distraction.

I know about data escrowing from my programming studies. If Company A hires Company B to develop some software - who holds the rights to functions, libraries, ideas or even the source code used in the software? Make sure that's in the agreement somewhere. It's not uncommon for Company A and Company B to agree an escrow solution for the code.

Be Social: del.icio.us, Digg, Furl, Reddit