Every time I open the newspapers or browse popular social media blogs, I see articles by so-called “analysts” (unfailingly with zero years experience in computer programming or social network design), “predicting” the future of my industry. With this week’s reports of Facebook topping Google in web traffic, the journalists have gotten their knickers wound up in nasty twist, with a few cowboys getting so excited they proclaimed that Facebook is on par with Google. Browsing through the entries in the national news and then the blogosphere I saw that the populist “analysis” centered on three tenets:
- Facebook have more traffic than Google and this continues to increase
- Facebook are creating a new killer OS (a delusion based on a recent quip by young Zuckerberg)
- Facebook’s targeted advertising is more effective than Google’s version.
In this series I will show that each of these assertions is false and argue that Facebook are nowhere near being a competitor to Google. Before anyone gets upset, I’m not arguing that Facebook is doomed. This year they will grow their user base enormously, incrementally add new features and continue increasing their profitability further into the green, perhaps even to $3 billion in advertising revenue. They’ll do fine. But they ain’t got nothing on Google.
Myth 1: Facebook have more traffic than Google and this continues to increase
This statement is both wrong and irrelevant – yet has been propogated throughout the media throughout the past week – including Newsweek and The Times.
To understand why it’s wrong you need to first learn a little about how domain names work.
Avoid if you don’t like set theory (or better yet learn set theory then read).
A website’s domain name (e.g. www.example.com) should be thought of as three sets, each of which is a subset of the set to it’s right. Taking the domain “www.example.com” and reading from left to right the first set (www.) is known as the subdomain. A domain (“example” in this case) can have many subdomains (e.g. www.example.com, mail.example.com or app.example.com), all of which are contained within the overall domain (“example”). Now the example domain (which contains all the subdomains) is itself just a subset of the top level domain, in this case (.com). Thus “www.example” is subdomain of “.com” just as “app” or “www” are subdomains of “example”. Differing entries at lower domain levels are all included with the superset to the right, but this relationship does not work in reverse. Thus app.google.com and mail.google.com are both included in google.com, but google.ie not included in google.com since as .com and .ie are different sets, each of which contains every single website or domain within that top level domain.
End avoidance warning.
Now, why does all this matter? One simple and important reason: Statistics for Google.com DO NOT include statistics for Google.co.uk.
Facebook only exists as “Facebook.com”, i.e. it only has one possible top level domain (.com in this case). Unlike Facebook, Google have many different top level domains (e.g. .com, .co.uk, .ie etc.). Thus we have Google.com, Google.co.uk, Google.ie, Google.de and so on. Each of these is a separate site drawing enormous traffic. At time of writing Google.ie is Ireland’s most used website, followed in second place by Google.com, and then third by Facebook.com. As you see the first and second spots in Ireland were held by two versions of Google existing at different TLDs. Whenever Facebook manages to end up in first place for a period this only means it is beating Google.ie or Google.com. It does not, however, mean it is beating the aggregate of the two. The only way to get an accurate picture of Google’s size is to aggregate the statistics for all these entries, something I’ve yet to see a journalist actually do. Indeed in the last week many prestigious publications (including NewsWeek and The Times) made the mistake of thinking Google.com traffic was representative of all their top level domains, thus exhibited an upside down understanding of the way domain system work. One last time for the audience: Google.co.uk is not a subset of Google.com
Even at this point the picture is far too simple.
Google owns a lot of websites besides it’s search core. Some of the biggest are Youtube, Blogger, Gmail, Maps, Docs, Translate, Wave, Picassa, FlickR, Ortuka. For a list of their acquisitions to date (92 at present) see the wiki List of Acquisitions By Google and to see a list of other service they provide see this wiki entry on Google Services. In calculating Google’s size (and quantifying the informational resources available to them), you really need to be taking all of this into account too.
Another dimension to take into account is when Facebook get their traffic. According to analytics firm Hitwise, Facebook was the “most visited site in the United States on Christmas Eve and Christmas Day”, beating even Google. There is a message here if we read between the lines: Facebook is where you go when your bored!
What is left unsaid is that Google is where you go the rest of the time. Surely your own experiences confirm this. If you’re busy at work you close down Facebook. If you’re finals are coming up in college you stop going on Facebook. In both these times your Google usage shoots up. You use Google when it counts.
More fundamentally, the idea of web traffic is misleading and needs further disaggregation. Traffic refers to a lot of different things – hits, unique visitors, time on site. With many of these statistics, less is actually more for Google. Google do not want you spending ages on their search site – they want you to find what you want and leave. They want you to solve your problem. If their results are good enough to give you the correct site in one second then that’s great. More time spend on Google’s site would indicate a failing on their part – you couldn’t find what you wanted so had to keep searching. Now Google ends up paying for not just one search query on your behalf but five or ten. Servers are expensive and the bottom line is that every time you click search you cost Google money.
So, the metric that counts is not simple traffic in Google’s case, but rather is the number of successful queries per day, success being defined as a user having solved their informational problem. As long as Google lead in this respect, it’s irrelevant to them where you spend the rest of your time online. So, in conclusion, even if Facebook oneday did top the Google aggregate in traffic, (something I see as possible in a few years), it still doesn’t even matter.
Next week I’ll deal with Mark Zuckerberg’s confusing quip that Facebook is a killer OS. It ain’t gonna be pretty….
For years revolutionary yay-sayers have piped that the internet will change everything. For precisely as long I’ve responded with scorn, feeling justified by my observation that part of being human is a bias toward thinking your own lot as ever so slightly above average. You can see this phenomenon every day. Just ask a smoker about the risks of getting cancer from smoking and he’ll give you one set of figures. Ask him then whether he honestly thinks those figures apply to him and he’ll coolly explain why he’ll beats those stats. We overestimate our lot; labeling the age we live through as revolutionary is just another symptom of that miracle of “I”. I ignored the cries of revolution and put on another bowl of pasta.
I’ve changed my mind. We are perched on the apex of a revolution – I cannot conclude otherwise. Fearing that the optimistic bias has defeated me too, my goal today is to attempt to prove this hypothesis. I’ll start with my “idea collision” theory, one tenet of my model for the creation of novel knowledge in a society :
“The probability of the formation of new knowledge is directly proportional to the number of collisions between ideas multiplied by the distance between the two colliding ideas.”
What happens when ideas collide
New ideas occur when two previously existent ideas collide for the first time. Combine the old fashioned telephone with battery technology and wireless signaling and you get the mobile phone. Combine the physics of mass, energy and electromagnetism signals and, with a lot of thought, and a stroke of genius, you get relativity. Don’t be fooled into thinking that every collision creates valuable new knowledge. The vast majority of collisions are fruitless and their results must be discarded.
Filtering out the many bad combinations from the few good ones requires testing. Here you have two choices – carry out experiments in the physical world or carry them out in your head using your own internal model of the world to test hypotheses. Carrying out experiments in the world is safer since the real world results provides feedback showing you if you are wrong or right. Carrying out experiments in your head is more efficient and sometimes necessary, given limitations of real world experimentation (such as unavailability of resources, data or time). It was through a vast network of mental experiments that Einstein discovered relativity – he imagined himself as a photon of light observing how things moved around him, eventually arriving at a theory he proved later with mathematics. Be warned, however, to carry out mental experiments you need to have a seriously accurate mental model of your domain of experimentation.
Einstein: the daddy himself
Digressions aside, every Eureka moment eventuating int the formation starts with a collision - followed then by testing. Thus more collisions are better than less for the generation of new ideas.
Moving on to the second part of his hypothesis – the multiplication of theses collisions by the “distance” between the colliding ideas. Lets start by defining distance: if ideas are “close” together this means that they are frequently thought of together by others, meaning the collisions have already occurred and most of the knowledge has already been harvested. Here, we approach a saturation point, whereby collisions no longer results in the generation of new knowledge. Contrast this with when ideas are “distant”. In this case the colliding ideas have seldom, if ever, been thought of together. It during these collisions that the most powerful new ideas are generated. Take the double helix structure of DNA, which was discovered by Crick and Watson. In their research the pair immersed themselves deeply in genetics, biochemistry, chemistry, physical chemistry, and X-ray crystallography. You might say there is little “distance” between the fields, as all are sciences - but remember that distance here is defined as the frequency with which two ideas are thought of together. Very few minds have the drive or ability to simultaneously understand so many complex scientific fields fields and so the intersection of ideas explored by Crick and Watson was at zero saturation, a field ripe for the harvest of fresh knowledge.
Crick and Watson discovered the double helix structure of these jeans
Returning to the title of this post – how is it that internet changes everything? Because it vastly increased the number of collisions occurring between ideas and the distance between the colliding ideas, in exactly the same way that urbanization did (the only comparable revolution in the past, the consequences of which were the industrial revolution).
Before mass urbanization people lived together in small communes, most staying put for the majority of their lives. These rural settlers had low numbers of idea collisions in their lives, given the small number of members in the community (and the time taken to travel to one another) and the small number of non human idea carriers (e.g. books, work practices, etc. – the medium through which an idea collides is irrelevant – what counts is that in someone’s brain the two somehow meet). The infrequent collisions during the agrarian age rarely generated new knowledge since the distance between the colliding ideas was miniscule as most members of the community had spent their entire lives together, preoccupied with the same narrow domains.
Contrast this to the industrial age. Settlers moved from across the countryside to big cities and instead of bumping into five or ten people in a day (and similarly low numbers of other “idea transmitters”), a person might now bump into ten to one hundred times more people or idea transmitters. As these people often came from different villages and different backgrounds the set of ideas they each held differed significantly and so the distance metric increased too.
We still live in cities – and cities today are bigger than before, better connected, and draw people from a broader background that in the industrial age. These factors increase both the number of collisions which occur and the distance metric for these colliding ideas. Distance is also increased too by improvements in the standard of living which enables many in the developed world to devote themselves deeply to the world of ideas. That being said, these improvements do not have the same revolutionary quality as the move from rural to agricultural. I
Now we have the world wide web, where people’s ideas collide with an exponentially greater frequency with ideas of exponentially greater distance. Your Facebook profile puts you in touch with hundreds through it’s mini feed. An individual wiki article may have thirty or forty editors, and by reading through the page you are exposed to the sum of their collisions (remember too that each of their inputs may itself draw on a similarly huge array of past bombardments). There are a lot more collisions – things are heating up.
Idea Collisions on the Wa-Wa-Wee-Wa scale
Not only are there exponentially more collisions – but also these collisions are from disparate fields. With the web an American can communicate effortlessly with other English speakers across the world. The spread of English as a new latin means that other previous boundaries separating distant ideas are now crumbling. Powerful tools like Google Translate provide a Rosetta Stone across language barriers. The rise of links between information on the internet and the Google powered search engine makes delving into distant ideas effortless – the barriers to learning distant fields are dropping harder than a banging London beat.
We are living on the slope of a steeply rising knowledge curve – and it has not been since the rise of urbanization has this slope been so steep. This, I argue, is why we are living in the age of a revolution.
Think otherwise? Agree? Say so in the comments.
Over the weekend a friend asked me why I blog and I thought to myself: what an excellent idea for a blog post. I’d like to make my motivations for blogging transparent so read on and find out why blogging has become as much a part of my life as regular exercise.
1) To Build Trust.
If you are working in the cloud, some of your co-workers, business partners or customers will never meet you. Most of these will still Google you before doing business with you – it’s a natural and understandable instinct. If someone Googles your name and finds nothings, or very little then you are an unknown quantity. If instead they find a blog, with hundreds of articles you have a personality. You have interests and a voice. People can get to know you through your writing, and can dig as deep as they like by looking through your archives. With my blog I hope that my current and future readers will become acquainted with my values, and from that infer information about my companies’ products.
I wear suits and I blog. Therefore I am trustworthy.
2) To Get Feedback on Ideas
The best way to develop an idea is through dialogue. Read through Plato’s Socratic Dialogues before you try to argue otherwise. When I blog, I have an audience of readers who think about what I say, then either comment, email or talk to me in person about the ideas. An added bonsus is that my readers learn about my interests and so bring them up with me when they see me in person, with such exchanges helping me develop my ideas further. My most valued readers will even tell me when I get something wrong – saving me from embarrassment or worse yet, from making a poor decision.
3) To Crystallize Thoughts
In an ideal world you would take half an hour aside everyday to think ideas through. We don’t live in an ideal world, and the truth is that the best way I know to think something through is to write about it. Mistakes in your logic become glaring as you type and the added knowledge that what you write will be read and judged by your readers ensures you keep strict quality controls . When I write, I don’t really have a clue what my conclusion will be until I’m nearly finished. I don’t write in chronological order either. I jot down ideas across the page, throw in various TO-DOs, then play around with the structure and content until it makes enough sense for me to hit “Publish.” The point is to get a reasonably firm conclusion about some topic out of my time spent writing – a new idea I can hopefully apply in the coming weeks.
4) Ego
Right know there are nearly 200 people who have read every single one of my posts. Some of my posts have 2000 readers. Not only that, but they spend a significant amount of time reading these posts. According to Google Analytics, my average reader spends 2:30 on each page. That’s huge. I feel flattered that in the age of the 10 second attention span so many people are willing to wait around to hear what I have to say, then come back for more the next day. It feels good.
5) To Demonstrate Knowledge
I studied law at University, yet I work with computers for a living. Someone might ask, rightly, how I can prove I have any expertise with the Internet since I have no qualifications in the area. My blog provides an answer to this – it lets people know I have thought a lot about what I am doing, and that I know the lingo and the ins and outs of my domain.
6) To Sharpen my Writing Skills
The secret to getting good at writing is to do a lot of it. Surprise surprise. That’s the secret to being good at anything – anyone who tells you otherwise is a liar. By writing regularly I improve my style and, since I hope to write a book in the distant future, I’ll take all the practice I can get.
Hopefully you can now see why I blog and perhaps get inspired to start blogging too.
This is nerdy…. but worthwhile. Never underestimate how important backups are. Most people never really realize this until it’s too late. I lost six years worth of notebooks and diaries when my little brother accidentally dropped my laptop six years ago. Gone. Forever. I felt like I’d lost part of my soul. Ever since I’ve backed up my important files. At first I backed up onto an external hard-drive. This works, but it’s a bit of a pain in the ass, needing to plug the drive in and manually transfer files.
About two years ago I discovered Mozy, a free downloadable program which automatically backs up your PC/Mac. Now if your Hard Drive dies, your Macbook gets stolen or you get drunk and stupidly delete all your files you don’t need to worry. Using Mozy you can easily restore all your documents and pick up where you left off.
Mozy offers 2GB of free space to store your files online. Use this wisely: there is no point in backing up downloaded movies and music. You can download this stuff again easily and you are better off putting these space intensive files on an external hard drive where 500GB only costs £50.
Instead you should use your free Mozy account to backup mission critical files. I use Mozy to backup my computer code, my documents not yet on Google Docs, my finals notes (back when I was in college) and my Mac settings and passwords. Once you install Mozy you select what you want it to backup and that’s it – you’re done. From that point on Mozy will automatically run every day (without distracting you in any way) and backup all your files.
The beauty of Mozy is that it stores your files online. That way you can access the files anywhere – and should there be a fire in your house destroying your external harddrive, you will still have a backup of your important files available online. Now that’s conscientiousness!
1. With everyone asking you to “become a fan” of their crappy business, groups have become the exclusive domain of small projects set up amongst friends. By using a group you positively associate yourself with these close knit communities and people are more likely to give you their attention.
2. Messages sent by the admin of a Page appear only appear in an “updates” panel. Problem: no one ever checks their “updates” panel and so these messages are left unread. Facebook groups send their messages directly to their member’s Inbox and so actually get read. I mean it’s right under your nose appearing alongside personal mail so it makes sense right?
3. You get “invited” to a group, whereas a page only gets “suggested” to you. Semantically, there is a huge difference between an invitation and a suggestion. An invitation feels exclusive, fun and special. A suggestion feels like criticism – it connotes that whatever is being suggested is in some way tainted as it is only worthy of being weakly “suggested” rather than more overtly offered. Invitations switch the power around – an invitation says “hey, we’ve got something great… want to take part? it’s cool if ya don’t, we’re having fun anyway”.
Pic unrelated.
