Google turns the whole Web into a social network with FriendConnect

It was just a matter of time before Google started to leverage the massive base of Google accounts to connect everyone together. As usual, this new Google project is ambitious – instead of creating a new social network as a destination site, the new FriendConnect service takes the network to where the people already are – on blogs and web sites.

FriendConnect comes in the form of a sidebar for sites and blogs that has a powerful viral invitation tool for getting people to join. You can select from a list of your Gmail contacts or share your site inside other social networks like Facebook, MySpace etc.

There are several gadgets that can be used to enhance the community aspect once people have joined your network – A wall gadget so people can post videos, comments and other bits of social lint onto your site; a review gadget so people can discuss how cool or lame you are depending on their mood and, most importantly, an OpenSocial gadget that allows you to embed OpenSocial applications inside your site.

I can’t say this without sufficient melodrama: FriendConnect turns the whole Web into a social network. Forget the social network as a walled garden, forget social networks as destination sites – yes they will still exist but this is going to be bigger. The scale of it is grand, like the Israelites toppling the walls of Jericho with their trumpets, or the rumble of elephants and infantry descending upon the Rhône.

I have the loyalty of a rented snake: Google Chrome

After using Google Chrome, Google’s new web browser, I set it to be my default browser.  Here are the reasons why:

1. It renders pages incredibly quickly (as does Safari, which uses the same rendering engine) and the browser itself loads fast
2. It does AJAX/CSS as expected
3. It is Open-Source
4. It has this cool application shortcut feature so you can double click on things like Gmail from your desktop (I know you can do this with other browsers but its not as cool)
5. This page works perfectly
6. Hopefully extensions are on their way – see Matt Cutts giant FAQ

Massive – Google offers vertical analytical benchmarking

I have been waiting for this day and it has finally come.  This morning I got an email from Google Analytics describing their new benchmarking service:

Benchmarking is an optional Google Analytics service that shows how your website’s statistics compare against other industry verticals. In the beta version of this service, you are able to compare your site’s Visits, Pageviews, Pages per Visit, Bounce Rate, Average Time on Site, and New Visits data against benchmark data from categories of other participating websites. You can use this data to gain broader context for your site so you can identify additional opportunities to improve your site’s metrics.

Although the announcement is quite low-key, the possibilities are profound. Firstly, it paves the way for Google to create a global web traffic ranking system based on any number of criteria ranging from page impressions, unique visitors, time on site, bounce rate or a combination of all of them. Secondly, it might mean that an API could appear that would enable aggregators like Amatomu to access the relevant statistics which we currently have to gather ourselves. The reason I think this might happen is because there are effectively 4 ways you can choose to share your Google Analytics data:

1. Not at all
2. Only with other Google services
3. With other Google services and anonymously – this is for industry verticals OR
4. Share data without the restrictions of the previous 2 points and, by implication, enable 3rd party access

This final point may be a misinterpretation of the wording but I doubt it and I think a massive centralised system that uses permission-based access to enable site comparisons based on identical criteria is something that is highly necessary.  The pseudo-authority of panel-based systems like Alexa and Compete.com irritate me on a daily basis.  Naturally some site owners would like to keep their performance secret, but take the example of South African publishers:  the Online Publishers Association publishes their members’ page impressions and unique reader figures on a quarterly basis by month.  There is a sense that among the top publishers in the country these analytics are not worth keep a secret and what it means is that we can all compete directly based on real numbers and a single and consistent counting mechanism.

Once benchmarking is enabled, you can immediately see comparisons based on other similar sites for the following metrics:

• Visits
• Bounce Rate
• Page Views
• Avg Time on Site
• Pages/Visit
• Time on Site

This immediately means I can compare, in relation to advertising conversaion and performance, why a site I am tracking might be under-performing and so on, and most of this relates to the bounce rate.  The bounce rate is the percentage of single-page visits.  A high bounce rate is generally caused by either search traffic on individual articles or having once big site referring traffic  – in both cases the page the user ends up may be interesting to them but the rest of the site isn’t, or is not presented in a way that makes subsequent navigation easy.  If the bounce rate is high then the avg time on site is low and so is the pages/visit. So one can see the value of being able to compare your site to other similar ones to see where your winning or losing.

It will probably be another year before Google feel comfortable offering a 3rd party API for this data but the possibilities are profound and I hope they do it.

iGoogle is a dog

A few weeks ago I built a gadget for iGoogle, which is Google’s equivalent of a personalised home page and I have a few things to say about this experience in no special order:

1. Development is ugly.  Google suggest you use an iGoogle gadget to create your gadget – that’s right, you create a gadget using a gadget, and it’s not a particularly good gadget. The hardcore out there will use the gadget editor to create a blank XML file and then paste in code they created in their IDE, but I can just imagine the poor non-techie who beleived the hype about how easy it is.
2. iGoogle is Ugly.  It’s great that iGoogle has an API and a lot of example code, the only problem is you end up with thousands of the same crappy-looking RSS gadgets.  I really wanted to believe iGoogle is cool but after puking on my keyboard for 3 days I turned it off because I just couldn’t take aesthetic punishment anymore.
3. Deployment is messy. Great and all when you finally get your XML file ready and it seems to work well enought to submit into the Google catalog.  In my experience, once you submit an XML file and someone installs it, whatever changes you make don’t seem to take affect.  As a result I have 3 XML files with different names and 3 submissions to iGoogle.  If there is documentation about this it’s not easy to find.
4. Finding anything on iGoogle is LAME. Explain to me why the top search company in the world has such poor search functionality when you’re trying to find gadgets.  Why is it that when I enter Mail & Guardian as a search string, GMail comes up top?  Why is it that when I wrap the phrase in quotes I get the same lame results? Why is it that when I enter the EXACT name of the gadget into the search field I still see GMail? Not even the pager system on the bottom of the search works properly, sometimes it shows more pages than there are results, then tells you there are no results even though you were looking at them a minute ago.
5. Communication is non-existant. After submitting my gadget to the catalog and being told that Google screens the submissions, the least they could have done is emailed me when my gadget was listed.  Though, based on the search, it would’t surprise me if their own mail ends up in the Gmail spam folder.

All in all this is one of the poorest systems Google has deployed to day and should be perceived, internally, as an embarrasment to all involved.  Maybe someone forgot to show them the memo saying that now that Google is big, people expect the appropriate attention to detail.

I have one goof thing to say about iGoogle, which is that you can easily integrate it with Google Analytics to see how many people are using your gadget.  This is the rose among the weeds.

Get the Mail & Guardian Online iGoogle widget here.

Douglas Merril, Google VP of Engineering: Not particularly interested in semantics for search, or competing with Amazon Web Services

Or at least what John Markoff of the New York Times calls Web 3.0 – a trumped up term for the Semantic Web. I was at the South African Google briefing this morning where, in answer to one of my questions, Douglas Merrill, VP of Engineering said Google doesn’t see much value in semantics for its search. According to Merril, the Google search is syntactic rather than semantic and this works well for them. There are, of course, several levels to the application of semantics and I assume restructuring Google’s indexing AI from syntactic to semantic is simply too big a project for arguably less gain.

The kink in Google’s armor however, may well be that their syntactic analysis may not work out for them in the long term if the majority of the Web starts using Semantic relations to express meaning. I suspect Merril was saying that there is little value in semantics as a part of the backbone of the search but this doesn’t mean Google won’t index semantics or output results semantically in the future.

Merril also said that Amazon Web Services have been a great product for Amazon but that Google is “comfortable with search” right now. This could mean one of two things: either Google is comfortable enough with search to branch out into the same space as AWS at some stage, or it means Google plans to stick with its core competency, which is search. Of course one could always ask, of AWS, why a book-seller is getting into the ISP business in the first place. Despite the glitch this weekend, AWS has revolutionized the way web application developer plan and deploy scalable systems but it is still a radical departure from the company’s core business.

Live-blogging the first Google SA event

I am sitting in a darkened auditorium at the Google event at the Forum in Johannesburg watching a video collage of scenes from Google’s early days and media snippets where people like Marge Simpson talk about Google.

This is the first official Google event for the South African office. Stafford Maisie, the SA country manager, is doing the introduction.

Operating on a cabinet level

At this stage they are “looking at the market”, deciding what to do next year. “We’re going to make sure we build the right strategy and infrastructure before we launch”. The launch is tentatively set for February next year.

“Every week I bump into people doing something with Google, it’s very exciting”, says Maisie. In terms of broadband access in South Africa, we want to be participating on a cabinet level… a lot of the statistics you see today might change”. Google wants to be a big pusher of broadband in South Africa.

“We can sit down on a ministerial level and show some very interesting case studies”, says Maisie.

The Internet is coming of age

The country is coming under enormous strain, with events like the Soccer World Cup, to grow the infrastructure and connectivity so that it is ready when the people arrive.

“What we could do for this country is just tremendous”, says Maisie. “We are about to see the dam wall crack, we are about to have pervasive connectivity”.

The message Masie repeats is that Google can play a vital role in the local online economy but don’t want to dominate, they want to facilitate.

The team that Masie is going to put together is going to communicate and interact directly with clients in a selection of vertical markets and the verticals will be a focus initially.

Masie suggested that Google mobile ads will be launched in South Africa before the year-end.

To end it all, Maisie tells us this is all off the record. He should have thought about that earlier.

Analysis

I blogged the intro from Maisie because that was the new bit for me, although the event went on for another 3 hours and was quite interesting for me, given that I don’t work with Adsense or Adwords directly. My overall impression is that Maisie and his team are going to do great work and I am particularly pleased they plan to lobby government to add to the current Internet infrastructure.

Much was said during the day, a lot of it very informative and what I have published here is mostly suggestion and rumours. When Google do officially launch in February I am sure it will be a major event and we will see the fruits of Maisie’s labour. The fact that Google have decided to do some solid groundwork in SA before making a big noise is encouraging, and Maisie’s repeated statement that Google wants to work with rather than against local publishers is a very positive sign. All in all I am very glad I attended.

Google SA – dusty outpost or a revitaliser of industry?

After what feels like ages [and countless rumours] Google has appointed someone to the position of general manager for South Africa. Kimberly Guest’s exclusive on ITWeb reports that Stafford Masie, Novell’s country manager, has been appointed to the job.

There has been much debate about which industry this person would come from – media, advertising, tech – and what sort of attributes they would need to have. Until last week I was thinking it would be a media person, but then someone reminded me of the difficult relationship Google tends to have with traditional media. The pure-play online media people tend to have less business experience/success in South Africa and advertising people are too specialised for a job like this so I think they made the right call.

It’s not that South Africa is a major focus area for Google, though in the EMEA context it probably is big alongside Egypt, it’s that the impact of Google making direct and interpersonal connections with advertisers in this country might well have a significant impact on the online media industry and their margins. The fact is the local online ad industry doesn’t need a revitalisation, it has been going through one for the last two years and the entry of Google on a more serious level may have a negative impact on the local entrepreneurship we’re seeing.

Aside from that, local companies paying offshore companies for local ad-spots seems a bit like we’re a banana republic.

On the other hand, and I know some people are already thinking this way, it might just turn out to be a regional backwater in Google’s colonial sprawl. Either way it’s going to be interesting to find out.

Google Analytics gets a much-needed facelift and the improvements are dramatic

Since its launch Google Analytics has been quite a confusing tool to use for site traffic statistics, especially because of the complicated and obscure menu structure and terminology.

Google launches Web History

Google has often been criticised for the potential risks to personal privacy because of the vast amount of personal information they collect, including your search and browsing history using their adsense network.

They have just launched a new product called Web History that allows you to store and archive your web browsing history and do various things with like, like search it, browse it by date and so on. It generates a Zeitgeist of your searches under the heading Trends and allows you to save bookmarks.

When I saw this I immediately thought about del.icio.us and what the implications are for social bookmarking in general.  With the relatively small user-base at del.icio.us and Google’s massive audience this could just spell the end for other social bookmarking applications.  I for one use my GMail account for Analytics, Blogger, the webmaster tools and Adsense so it would be easy to switch my bookmarking system.

Michael Arrington finally gets off the fence about Google in China

On Sunday Michael Arrington of TechCrunch finally came out and criticized, in a post called Google and China and Evilness, Google’s decision to expand into China, co-operate with the Chinese government and censor their search results for the Chinese public.

This reminded me of a discourse analysis I wrote last year called Harvest Song: Google’s dancing discourse, arguing that the mainstream media swallow Google jibber-jabber uncritically and that their ambiguous position between media and technology creates a critical black hole. Let’s not forget their power to censor their critics if they so choose.

Arrington says:

But if Google wants to stay in the good graces of the smug western crowds, they need to say they regret working with the Chinese government because that government is evil, not because it turned out to be “a net negative” business decision.

This is yet another reductionist American take on the complexity of global politics. Arrington says he didn’t get personally angry with Google when they moved into China but

…in late 2006 I visited Taiwan for a conference and learned a lot more about some of the things the Chinese government continues to do, particularly torture and other reported atrocities perpetrated on Falun Gong members.

Okay two points:

1. There was no ambiguity about how the Chinese government treat opposition and the media BEFORE Google moved in there, and no excuse for not knowing about it in a world where instant information is available on any topic
2. The United States has very very dirty hands right now. In the global village, is there any difference between taking out opposition at home or abroad?

Of course it would be inconsistent to suddenly admit that the wars the US has perpetrated over the past 50 years have been about money, oil, or retaining their position of power. It must, always, be turned into a moral dichotomy of extreme good vs extreme evil.

The damage Google did when they went to China was in showing the world that their search results can be controlled and filtered ideologically and therefore cannot constitute scientific knowledge. The algorithm is an ideological device.

Believe me, Google results in the West seem like a transparent democratic system of merit because they do not contradict the dominant system of power and its consequences.