Saturday, September 29, 2012

Seeking "ivansml"

I'm finishing off my forthcoming book Forecast: What Extreme Weather Can Teach Us About Economics and have today been going over page proofs (extreme tedium...fixing typos etc). An important matter: I've referred in the book to some work on learning in macroeconomics (Evans and someone else) that was suggested to me by "ivansml", a graduate student at some European institution. Ivan: can you email me (buchanan.mark@gmail.com)? I would like to point out in a footnote that you directed me to this work. I can do this either using "ivansml" and referring to this site, or by using your real name (which I would rather do). THANKS!

Thursday, September 27, 2012

Bubbles

I just stumbled on this post from a few months back by Noah Smith. Like all his stuff it is a fun and informative read. Essentially, he looks back to the famous experimental paper of Vernon Smith and colleagues which found clear evidence for strong and sustained bubbles in an artificial market with students trading a fictitious asset with real value. The novelty of the experiment was that this asset had a clear and perfectly well-known fundamental value (unlike real financial instruments), and so it was easy to see that the market value at first soared way above the fundamental value, and then crashed down again.

Noah's post looks at why this result, for financial economists, didn't nail the proof that asset bubbles can exist and ought to be expected in real markets. Most of the arguments seem to be centered on the idea that the people acting in real markets are far more sophisticated than those students, and so would never pay more than the true fundamental value for anything. Suffice it to say this argument doesn't hold together at all well in the face of empirical evidence on real trading behavior, some of which Noah reviews.

One thing caught my eye, however, and is worth a short mention. As Noah writes...
If bubbles represent the best available estimate of fundamental values, then they aren't something we should try to stop. But many other people think that bubbles are something more sinister - large-scale departures of prices from the best available estimate of fundamentals. If bubbles really represent market inefficiencies on a vast scale, then there's a chance we could prevent or halt them, either through better design of financial markets, or by direct government intervention.
 I am certainly someone of the latter camp -- convinced that markets often depart from fundamentals (even such values even exist) for long periods of time. But I think the third sentence on what we might do about bubbles needs to be refined a little from a logical point of view.

Bubbles aren't necessarily totally bad things. Perhaps we may find that they are a useful and necessary part of the collective learning process. The foraging of a flock of birds is highly irregular; it moves this way and that, following the lead of different birds at different times, sometimes moving on large excursions in a single direction. A market might be somewhat similar as a collective social process for searching and exploring. We shouldn't expect that what it has found at any one moment is optimal; it may often make huge mistakes. But the process of exploration may be useful anyway.

In that case, we may find that we don't want to stamp out bubbles, unless they get really big. Or if the bubble is one driven by a systematic increase of leverage by investors which sets the stage for a certain explosive episode of de-leveraging, with subsequent long term consequences.

What we do want to stamp out, however, is the dangerous idea (still supported by many economists) that bubbles don't exist. That's the one idea that can make our markets really prone to disasters.

Wednesday, September 26, 2012

The final installment of the Oswald-Green debate on home-owning in The Economist

It is here.

Buttercup



{image source: flickr user law_keven at www.creativecommons.org }
 
I remember the buttercups that used to dot my elementary school playground. 
 I went there for nine years so I knew the field well.  The buttercups would shoot up faster than the blades of grass between mowings.  
 
{image from wikipedia}
 
When we were younger, we'd hold them up under eachother's chins & it was, " Do you like butter?" 
  
A little older and it was, "he loves me, he loves me not"
 
{He loves me!}
 
 You never really knew if he was going to love you or not with buttercups because some types have 5 petals and others have 6.
 
 
 
I loved making flower chains & never really grew out of it.  I remember doing them in the Spring as I waited for my turn at long jump in high school track.  One day I remember making necklaces for the other sprinters & somehow convincing even the guys to wear them as crowns...  (My coach must have loooved that ;)
ahahahaha   (It really still makes me laugh out loud.) 
 
More than anything, I just remember always picking them up & playing with them when I see them.  Long talks on the field with my best friend, scooping down to pick a buttercup, subconsciously.  Crumbling the old flower centers into little pieces between my fingers & rolling them around. 
 
 



 

 
I love these pics of the plants from NC State...  (above & below) 
 





 


 

 
It's for pretty much all of these reasons that I've known for a while I wanted to do a buttercup design for my textile line.  When the power went out for days this Summer & we needed to get out of town, we made an impromput visit to see my dad in Illinois & have him meet the baby...  The car ride is 13 hours and it was the first time after Luke was born (in May) that I'd had some time to sketch.  I  worked on "buttercup" in my lap...


{The beginnings of it...  as you can see, drawing doesn't come easily to me & there are lots of erase marks & mess ups}
 
The design (like all of them) is just an interpretation... not a botanical study.  I add my own little things here & there and I'm sure I leave out pretty important parts, but it's the feeling that I'm after...



{One of my first strike-offs}

I'm working on the final colorways now but I'm really happy with the light earthiness of this first one...

 
A few more options are on their way & I hope you enjoyed a peek into the process.
 
...
 
Do you have any buttercup memories?
 



If you'd like help creating a home you absolutely love, contact me about our design services.

Tuesday, September 25, 2012

SKETCHBOOK SELECTIONS VOL2: AUTOMOBILES


The sketchbooks have been filling up at a fast rate this year, so I've now put together a 2nd volume featuring 24 different automobiles from bubble cars through to Bentleys. 
Limited to 100 signed and numbered issues. 
Click on the following link to buy your's:

Beaulieu International Autojumble 2012 report

I was invited to the Beaulieu International Autojumble by Vintage and Veteran to run a display and sketch on their stand for the weekend. This was my first time at Beaulieu and it's  true what they say about it, it is huge and people are selling everything imaginable, along side a huge auction by Bonhams on the Saturday and many other dealers displaying their vehicles. 
I spent most of the weekend on the Vintage and Veteran stand sketching the incredible array of bikes, with the occasional walk round the jumble resisting a purchase. Even though I had spent the previous Sunday with Vintage and Veteran (see previous report, Founders Day) the bikes on display were different showing what an extensive  and ever changing stock they have. Here are the sketches from the weekend along with some photos of my favourite vehicles from the weekend.

1903 Royal Enfield


(ink sketch)
I was very happy that Vintage and Veteran bought this bike along to Beaulieu, as I got to spend a fair amount of time with it over the weekend. I decided not to only sketch a full view but add some vignettes of the front brake, Brooks saddle, the engine and the Royal Enfield badge.

1917 Harley Davidson 1000cc V-Twin

 
(ink sketch)


1950 BSA Gold Star 500cc

(ink sketch)


1920 FN 4 Cylinder 

This was the star of the Vintage and Veteran stand over the weekend. A fantastic example of a FN 4 cylinder shaft driven machine, considering manufacturers were only just working out how to build motorcycles the amount of ingenuity in this machine is pretty mind blowing.
The 4 cylinder lump
 On the other side there are 4 priming cups which are used when starting up are the carburettor is lower than the pistons it needs a top up to get the fuel flowing.


1925 Henley 550cc (Blackburne engine)

A rare bike indeed these were only made for a short period and apparently are very nice bikes to ride.


1910 Douglas Model D

1936 Vellocette KSS




 Just a few of the grteat bike for sale on the Old Bicycle Company's stall

 1928 Sunbeam model 20.9 pickup
1919 - 20 L'Bleriot 500cc
This incredible Veteran French Bike was on the Yesterdays stand, unfortunately it was sold already. 

 Citroen DS Safari
(I will one day own one of these as a family car)

 Bedford Van
 Chevrolet




Monday, September 24, 2012

Founders Day 2012 report

I was invited by the ever supportive Vintage and veteran to run a display and sketch on their stand at Founders Day at Stanford Hall on the 2nd September. Founders day has a very extensive bike jumble and is well worth a visit. I had a great day sketching the ever changing roster of bikes Vintage and Veteran have, as well as chatting to people and drinking a lot of tea. Here are the from the day:

1913 Rover TT

(ink sketch)
These Rovers are few and far between these days and this is a fantastic example. I particularly liked the  way the oil pump is incorporated into the fuel tank.
 For more photos of this bike please go to the Vintage and Veteran website:

1912 FN Single

(ink sketch)

All original condition and in running and road legal condition. I fell in love with this bike and would have snapped it up if I had the cash, as it has so much history and character there is no need to touch it.
 For more photos of this bike please go to the Vintage and Veteran website:


Here are a couple more of the bikes that were on display:

 1927 Triumph Model N


1903 Royal Enfield

 Billed as the oldest running Royal Enfield this is certainly a sight to behold.

The following weekend Vintage and Veteran invited me to Beaulieu to do more of the same, report to follow.....

Demo & Rebuild

I'm back to the construction zone from a whilrwind trip to NYC and, like I've been mentioning for the past few weeks, we're now down to the studs at our place. (!!)  We're reworking the floorplan- closing doorways, adding new ones, moving walls, removing them... plumbing is being moved, windows are getting enlarged, we're adding a fireplace, redoing the bathrooms & kitchen... it's madness!! (but the good kind of madness ;)  We have the air off because there's so much dust flying around & we want to keep it out of our upstairs "apartment" as it's been dubbed... But as I type this morning, I'm hoping it stays warm a little longer because there is definitely a chill in the air & there's no heat for us until we get the bulk of the contruction completed. 

One of the biggest changes we're making is opening up the kitchen to the living room...  Here it is before:


And here it is now:
 
 
They've finished sanding down the beams.  The red-brown stain is now GONE and we've exposed the bare pine.  (How I wish that pine had been exposed over the past 40 years, aging gracefully but oh well... It's very raw now & we had it hand scraped & wire brushed.  )  We've also had insulation added between the beams because before there was only a layer of plywood before the roof & we're making it a bit more energy efficient. 
 
And of course as  you can see... the wall is DOWN!!!  The kitchen is now completely open to the living room.  We also expanded the kitchen two feet to the right (taking space from the dining room) so that it spans the whole width of the living room now. 
 
Floors are going down this week & next (more on that as soon as I have pics!) and I CANNOT WAIT.  Once we have floors, we can move our things inside from the garage & pod.  Dave went diving in the pod for clothes last week & climbed over all of the front stuff to get to the Fall stuff in the back.... He dug himself a little hole and climbed down into it....  And I  honestly thought he was buried by the stuff at one point.  (thanks to him of course.) 
 
 
{it's since been rearranged}
 
Me, after realizing I hadn't heard him talk in a while:  "Dav, are you okay in there??"
him, very quiet/ far away: "Hellllllllp"
At that point I was about to launch the jogger strollerfrom the pod to the woods behind me as I screamed "DAVE?!!!!!!"
Then I heard "No no no no no!!  I'm fine!!" in a completely normal volume. 
Then laughter.
Oh wow.
 
He's very lucky I couldn't reach him at that point.
 
We've had the discussion before of "fright or flight" in emergency situations.  I'm a "fight" kind of girl.  There's tyipcally no hesitation in my reactions and I find myself doing things before I've even thought about them.  Thanks to him, (yes am still bitter ;)  I almost launched our Christmas breakables into the woods. 
 
Anyway, things are happening around here & I'm in the midst of planning the kithen (and pretty much everything else around here) so I'll be back this week to share some plans!!  Off to some clients for the day & keep your fingers crossed we find our nanny soon!!  My head is busting with things I need to remember & people I need to speak to... send some brain space my way. 
I've never been so busy in my life. 
I'm not at the breaking point yet but feel like I should be. 
Maybe once you actually get here, you're too busy to break. ??? ;)
 
 
{at least I've got what matters most}
 


If you'd like help creating a home you absolutely love, contact me about our design services.

Sunday, September 23, 2012

The Oswald-Green Debate in the Economist: Rebuttals

Andrew says:

If high levels of home-ownership impair the efficiency of the labour market, then the costs are potentially huge. It is the job of economists to point that out.

I say:

Does home-ownership gum up labour markets? A number of papers sought to test Mr Oswald's conjecture and my takeaway is that the impact of tenure on labour markets is marginal.

[Note: as an American, I never myself would write "labour," so The Economist changed spellings to conform to their style].

Thursday, September 20, 2012

Ed Glaeser weighs in

He advocates neutrality:


The government should neither encourage nor discourage home-ownership. Significant public interventions require evidence of significant market failure, and confidence that the costs of state action will be less than the costs of those market failures. These conditions are not met in the housing sector, whether we are contemplating pro or anti home-ownership policies. 
The case for home-ownership often begins with the view that home-owners are better citizens, who create social benefits by investing more in their communities and their governments. My work with Denise DiPasquale does find that home-owners are more likely to work to solve local problems, to vote and to know the names of local leaders. These effects reflect both home-owners' stake in their community and their tendency to live in one place longer. 


If Apple is so cool...

...why doesn't their new map ap give public transit directions?


Wednesday, September 19, 2012

Expose: Google's evil toys Panda and Penguin

It has been 18 moths since Google started experimenting with its new, radical search algorithms, all with good intentions to make the web “free(er) of spam” and to improve user satisfaction by delivering search results that they “should want to see”. Well, whatever Google is doing is not quite working for me…

[image courtesy of  SEO Geezer]

# Google Search getting thin on answers...

For a start, as a Google Search service user, I noticed a gradual degradation in the quality of search results rather than improvement. Big organisations and primary sources of information tend to dominate results page. What’s the benefit of a whole page of search results pointing to a single website? It is all good if you are searching for routine facts as eg. weather – you really want the most respectable provider of the information which in Australia is the Bureau of Meteorology.  Fair enough. Sometimes I just want to visit the same site (bookmarking gets out of hand after a certain number of sites on your “favourites” list) so, search is a good way to get quickly to the right page on a popular site that has the answer to my query (eg. quick lookup of JavaScript syntax). But far more frequently I want to discover new things... and this is where I am getting more and more disappointed with Google.

Google used to be a great tool for exploring the vast resources of knowledge in web universe. I used to be able to search 10 pages deep and almost all results were relevant. Now I am running out of relevant content just after a few pages... Either terms I am looking for became more complex or there is something wrong with Google algorithms.

Does Google’s preoccupation with “personalisation of results” have anything to do with it? It appears that Google’s approach is that it “knows best” what is good for me and therefore what results should be served to me.  So, Google tries to show me what I may like, and not the whole spectrum of possible answers, as it used to... The consequence is that after a few iterations of search terms you start running into the same results, over and over again.

Due to my growing frustration with Google Search I tried Bing for the first time to do the latest search exploration. To my greatest surprise Bing actually pointed a few more relevant results than Google, these results turned out to be much more useful and actually steered me towards the answer I was looking for. Without a doubt I will be incorporating Bing in my future explorations. 

And it looks that I am not alone with such views - even Microsoft picked up the trend and started running a campaign to highlight this fact: BingItOn.


# It is all getting a bit too personal for me...

On the subject of personalisation, it is becoming an issue of great concern for me. Google is not only limiting search results to what I “may like” but also crawling my private emails and snoops the sites I visit to create a profile and serve me the ads that I “should see”. I didn’t really pay much attention to it until a colleague asked me “how did I know” to serve equestrian related ads when she browsed though my blog (I have Google Adsense installed there). Mind you, my blog has nothing to do with horses, and not a single word refers to horses (until now)! It got me thinking... what really Google knows about each of us and how it decides what is "good for us". Ah, the price of “free”... Where is the limit? A few old books come to mind.


# Panda paws cut deep...


This leads me to another point I would like to make about recent experiences with Google – the effects of algorithm upgrades on individual websites. The Panda upgrade which was rolled out in February 2011 was designed to downgrade sites with “thin, non-original or duplicate content including those sites with a high ad-to-content ratio”. On the face of it, eliminating irrelevant, “built for Adsense” sites should be considered a worthy cause but think about it for a moment… For one, not all “content” on the web it text (Google would prefer it to be but it isn’t). And this really discriminates against the sites that work as a single page applications and/or have dynamically generated content where words are not the dominant component. Google crawler is too dumb to make any sense of the code on such pages. Measuring “ad-to-content” for those sites doesn’t really make sense either.

Then there is a whole question of “originality”. Does reproducing eg. earthquakes info from USGS or weather from BOM constitute duplicate content? Taking the extreme view, if answer is yes, then according to Google myriad of  apps with weather and similar recycled info have no right of existence on the web (or at least in their search directory) because it is all non-original and duplicate content! It also means that Google questions the whole idea of opening data, removing licensing constraints and repurposing for economic and human advancement.

Surprisingly Panda didn’t do much damage to my site although it definitely falls into “thin, non-original, high ad-to-content” category, as judged by the number of words per page and considering the content is almost exclusively sourced from third parties via web services. So, one can only question the effectiveness of this algorithm after all. Or its real intent...

The flip side of the implementation is that it creates unforseen problems for both the users of Google Search and “authoritative sites” alike. An example is a recent, let’s leave it as unnamed, event that caused a spike in traffic to a particular site. The site experienced “some problems” due to the volume of incoming enquiries, so the users could not get the answer they were looking for. Google could be seen as the main culprit contributing to the problem due to its monopoly on web search service AND since almost the entire first page of search results was directly pointing to that particular web site. Other information alternatives, which remained fully operational, were far down on “Google suggested list of results” and could only be discovered by those who bothered to look beyond the first page.


# Penguin bites indiscriminately...

The Penguin algorithm, released on 24th April 2012, was aimed at punishing the sites which apparently engage in not accepted by Google Search Engine Optimisation (SEO) practices. As a consequence, a number of sites have dropped down the search rankings or even been eradicated from Google’s search engine “due to a number of web spam factors including keyword stuffing, cloaking, unnatural link building and content duplication”. Er, if it was Google’s main objective it certainly did not work as intended, catching many quality content sites in the net. My site is one such example. Well, being labelled a spammer and being accused of engaging in dubious, according to Google, SEO practices begs a question – on what grounds? What gives Google the right to be the judge and the executor deciding what is ok on the net and what is not? What gives Google the moral right to decide what is good for you and what is not? Only because they have a monopoly on web search service?  Personally, I prefer to decide myself whether the information presented on a particular site is of any value to me or not...

It could be argued that what Google considers “unacceptable practice” cannot be clearly distinguished from reasonable and quite appropriate behaviour. For example, according to Google, if you post a comment on some site and reference a page on your site it is wrong, if you have only one paragraph as “content” of your page (ie. what Google can read) it is wrong, if that paragraph has a word that is repeated 3 times it is wrong, everything is wrong unless… you scramble a meaningless text with non-repeating words (which Google boot can recognise) and you do it at least once a week (ie. fresh scramble) and you don’t include any ads (especially those from non-Adsense network) and you have no outgoing links on the page but you link to that page with a keyword that is unique for each page even if you have 1,000 pages... you may be right (until the next update of the algorithm... perhaps).

Of course I am being sarcastic. Not because Google “punished” me taking away search traffic (by the way, more on this below) but rather to highlight some, let’s call it,  inconsistencies with Google’s still current “Do no evil” motto. After all Google itself is a giant link farm (yes, however you want to spin it, it is still a link farm!) so it should come as no surprise it will do anything to eliminate any form of link competition (certainly Google is not interested in people navigating the web independently of its search engine!). It wants monopoly on links and due to its size it has the power to banish to oblivion anybody “non-compliant with Google rules”.


# Is Google still what we think it is?

I have no doubt that Google has plenty of good, well meaning people however, due to the size of the company its “personality” is no longer shaped by individuals working there. As is the case with any corporate identity when it gets big enough, it becomes its own master, gets its own ego and assumes its own set of values. It can easily become a hard to control monster that can turn against its creators and supporters... a Frankenstein story is quite a good analogy here.

The above examples expose some serious deficiencies in Google algorithms of late that are either the sign of Google’s inability to deal with the complexity of content on the web or are the symptoms of more sinister actions perhaps, but which the company tries to justify in the name of “creating a spam free Internet”. Time will tell whether the broader community will buy into this argument or whether it will gradually turn against Google as dissatisfaction with search service grows. If the company loses you - the searcher, they will struggle as a viable business...



# Where it all started...

And this is where the plot thickens because there is more to search traffic than meets the eye - a very dark secret which is only a hypothesis but not without some merit... Google started as a free service allowing people to find relevant content on the very tangled web. Unsurprisingly, it gained popularity very quickly due to comparatively high quality of search results and simplicity of presentation. These days the main game for Google, above all, is revenue from advertising – generated by clicks on the links that appear above and next to search results – not the search service itself or other fringe revenue streams. Introduction of Adsense and Adwords was the game changer.

By making it so easy to advertise and at the same time to publish the ads, Google was the main culprit in creating the monster it now tries to eradicate, ie. “spammy content on the web”.  In particular, it didn’t take long for people to figure out how to make easy money with Google. The “business model” was very simple: get to the top of Google Search for a popular keyword to generate good flow of traffic to your website, include Adsense ads on all the pages (never mind the quality of the content) and count the money paid by Google every month. Some made big money using this simple formula and when the competition in creating content intensified they moved to teaching others how to make money exploiting quirks in Google algorithms. It gave rise to the whole range of new industries and propelled the profits of Search Engine Optimisation specialists, who continuously devised newer and newer tricks and strategies, some ethical, some less so, to stay ahead in this game.

Google fuelled the new gold rush by providing significant monetary incentives to create content and to carry Adsense code. It was all about aggressive cyber-land grab at the time when Internet content was rapidly expanding. That huge volume of content inventory, which Google was controlling via its search engine and Adsense, became very attractive for advertisers trying to reach mass online audiences. Others could not recognise the opportunity or were rather slow to respond. It made Google, with its Adsense and Adwords services, what it is today: a dominant market player in online advertising market and one of the largest companies in the world.


# Is “spammy web” really the issue?

Initially nobody questioned the ethics and usefulness of all that newly created content. Eventually a derogatory term “made for Adsense” was coined to describe the practices of the most aggressive players. But in all honesty, it is difficult to blame hundreds of thousands of individuals and companies trying to make “a few bucks on the net” by exploiting, to various degrees, the rules of the game they did not control. True, guys and galls flashing big cheques from Google could get your blood boil if you tried to make your content visible on the web with nothing but hard work and good intentions but hey, this is a free world where those who are able to explore opportunities better than you will inevitably be more successful. You have to remember that the wast majority of “made for Adsense” cases were not, and still are not illegal if they stopped short of fraudulent clicking on ads. All those sites are now just inconvenient for Google (but not in the sense you may think...). In fact, the basic premise of providing what is needed (ie. satisfying the demand) is at the core of any business activity. In this case it was collecting information on a particular topic for which was a big demand, publishing it on the web and promoting it with whatever means were available. But when Google reached the dominant position in both online search and online advertising it didn’t like it anymore. Old allies became a burden...


# The truth about search traffic...

You may wonder, why some of those “made for Adsense” sites were so successful? Why people kept coming to those websites for answers if the “quality of content” was so poor?  Our cyberworld is big and as long as Google kept directing new visitors the sites were doing great business. You see, all those “made for Adsense” sites were exploring a very simple truth – people who are in an “active searching mode”, and cannot find the answer on the page they have just opened, will be most inclined to click on the ad if it relates to the subject of their search than those regular users frequently visiting that same site. It could be even argued that the lower the quality of content, or less relevant, the more likely it is for the user to “click away” on the ad (I have some proof of this theory). Don’t forget, this made big money for Google as well...


It is quite an intriguing discovery that the key to making big money on the Internet for sites monetised exclusively with clicks-on-ads are new visitors and not the regular users! It applies equally to dodgy sites as well as respectable media portals. Just to prove the point, when Penguin run its course I lost about 75% of search traffic from Google and 50% of Adsense revenue (by the way, this is “the difference” of being listed on the second page instead of top 3 result in Google Search). But search traffic is only a fraction of all the visitors to my website (ie. about 4%) so, it means that the majority of advertising revenue came from a tiny proportion of users: that is, ad-hoc visitors in an “active search mode”! I have no doubt that the same pattern of user behaviour can be observed for any site, including those in the News or Fairfax portfolios. It leads me to a pretty bold conclusion: you can’t make good money on action oriented ads relying on regular users! If users keep coming back to your site it means they have already found what they were looking for and are not in an “active search mode” anymore. By the way, that should also explain why Facebook is struggling with its advertising program because their main audience is engaged in social activities and not “exploration and discovery”, so action oriented advertising is not appropriate. Touchy-feely stuff could work better but revenue-per-click ads are unlikely to deliver on social platforms. This has all sorts of implications for social media portals as well as advertisers.

It all worked very well for a while for all involved. Google used to focus its efforts on driving traffic to smaller, content specific websites which had keywords that people were searching for and which published contextual Adsense ads (where context for the ad was derived from the search term and/ or dominant keywords on the site). It was a good symbiosis until you - the public, started spoiling the game by clicking less and less frequently on ads.




Again, if my site is representative of what is happening in the bigger online world, it would indicate that in the last 5-6 years click-through rate on contextual ads declined by a factor of 10. Google’s initial response was to lift advertising rates, which increased by the same factor in the corresponding timeframe. The only way for Google to keep growing its revenue was to either increase the inventory (ie. content for serving ads) or get more money from each click. None of these options is easy - these days the growth of content which is adequate for Adsense has slowed down and costs for advertisers reached the point that they started to question the return on their money (global financial crisis was also a contributing factor). Another game spoiler for Google was that Google’s dominant position in online advertising came under attack from other aspiring online companies that were starting to explore (and promised better outcomes with) social engagement and networks as the key determinant for ad content (Facebook in particular). Since there is only so much the advertisers could bear the only move for Google was to cut aggressively into the revenue shared with Adsense affiliates. And this is exactly why Google unleashed Panda and Penguin...



# What is really important for Google...

It is rather easy to determine what Google is really trying to achieve here: to get people to click more on ads on Google Search pages rather than further down the stream (ie. on third party sites that may or may not carry Adsense ads).

Google official position is that by eradicating spammy content from their search results, and therefore increasing “quality” of search results, they will achieve more clicks on ads and hence bigger revenue (this is the story line in their quarterly reports)... Er, that’s not what I discovered and I have evidence that it actually works the other way round. So, it puts in doubt, at least in my mind, Google’s official motives for releasing Panda and Penguin algorithms. Just think about it. If you have found what you are looking for would you be interested in clicking on yet another ad on this topic? I wouldn’t. And consider this - if you can’t find what you are looking for you keep clicking the links, the ads, going back to Google to do more searches... more visited pages means more served ads and more opportunities that you will click those ads. If Google was really “successful” in eliminating spammy content and get you exactly what you are looking for in a single step, and every time, it would reduce the volume of traffic and Google would have killed its golden goose... Unless it would be possible to substitute all genuine search result with “genuinely paid links” (ie. advertising!) but it is a topic for a different post. I am sure Google knows this, its analysts working with the mountains of Adsense, Analytics and Search data are not stupid.


# And what Google is doing about it...

So, in order to accomplish its objective Google had to rebalance its directory of links: relegate “less sticky” sites (ie. the ones with high search traffic proportions, high bounce rates and large click-through rates - conveniently labelled as “spammy”) to the bottom of the list and replace them with “more authoritative” companies and web content providers who are less likely to rely on click through ads (either Adsense or else). What is the benefit to Google? Often those sites don’t carry advertising so, reduced competition to Google own ads. Very often big companies are also big advertisers on Google so more “topic related” traffic to their sites means better return on their money and ultimately better satisfaction with Google and its advertising platform (so they don’t look for social alternatives!). A by-product of Google action is actual reduction in what users perceive as “quality results” which should drive them back to Google Search in larger proportions (because there is a reduced chance of links/ads on target pages to “click away” and after all, there is only one dominant search service provider!).  But the clincher is that if people don’t see what they are looking for on the first page of Google Search results they will, you guessed it, click on the most relevant ad! (By the way, did you noticed that paid ads are starting to blend with search result more and more? Isn’t it hypocritical of Google since this practice is explicitly forbidden for Adsense affiliates?).

So, as you can see, with this counterintuitive but brilliant strategy Google is attempting to reduce clicks on third party sites and increase those on their own pages. Less third party stock for ads (as traffic to Adsense affiliates falls) means Google will be able to reduce the cost of those ads (not to advertisers but rather Google’s share paid to Affiliates as lower traffic means “lower grade ads” means Affiliates will accept lower cost per click).  At the same time this should increase competition for ads on Google own pages, pushing the prices up. A nice by-product of the whole exercise is that those panicking webmaster, whose sites are no longer on top of Google search results, will try to restore the volume of traffic to their sites by buying more Google ads via Adwords. Of course Google could not publicise their true intentions, spinning instead a well rehearsed PR campaign featuring “fighting spam” and “improving quality of content”...


# Clues are all over the web...

Well, you may say, this is all a speculation, Google only have our best interest in mind. My response is that we all tend to forget Google is a giant corporation and not a benevolent organisation like Wikipedia or Open Street Map Foundation. Why would it put our interest before its own? There is a plenty of anecdotal evidence that things are not as clean and transparent with Google as we are all let to believe. For a start, Google’s search engine guru Matt Cutts publicly admitted that sometimes ads are more relevant than organic search results. If it doesn’t make you think, then there is a history of very little transparency about how Google goes about its business.

Consider this - as an Adsense publisher you have almost no control over what ads will be displayed on your site. Keywords? Remember what I said earlier about experience of my friend! Same goes for advertisers, unless they nominate specific sites their ads can be placed anywhere (by the way, there are plenty of embarrassed companies whose ads appeared in a wrong context, eg. coffee ad next to an article about coffee contributing to a hart attack!). Most importantly, Google has the first option on placing the ad – which may be their own site more often than not! So Google in effect “competes” for the most valuable ads with Affiliates. A separate issue is that there is only so much stock of ads available at any particular time so, it is not difficult to imagine that at the time of a severe economic downturn Google could hold to all the stock to protects its profits. Proportions of revenue from Affiliate network is a good indication where Google’s priorities really are.

Yet we all trust Google to “do the right thing” (whatever it means...). So, as a publisher you expect that you will be given a “fair share” for providing advertising space on your site, yet this share is exclusively at Google discretion to determine. Same goes for advertisers - Google promises you that you are “competing on fair grounds” with other advertisers for certain keywords so whatever you pay for your ad is determined by demand and supply plus “Google’s ethics”. But who says Google cannot manipulate those algorithms to their advantage? They will lose some business? This is what monopoly is good for! Your options are really limited… Everyone else is working with Google so it surly must be fair dealing...

The fact is that it would be hard to find many similar examples in terms of business arrangements between suppliers and clients. How many other cases can you name that clients pay “what they feel like” or that you buy supplies without knowing the price upfront? This means that 10c you are being paid for an ad can be 50% share, or only 10% or maybe just 1% of what the advertiser is paying.  It is entirely at Google’s discretion. As an advertiser you bid for a particular keyword but you don’t really know how Google arrives at the price you will ultimately pay. If you nominate $100 daily budget and 50c max cost there is nothing stopping Google taking $50 for a single click from their search page then allowing 200 clicks at an Affiliate site for 15c each. So the advertiser gets what he/she wanted (ie. 200 leads with average 50 cents per click), Affiliate pockets $30 and Google… w whopping 70% of the cash! As long as advertisers and Affiliates think it is all “reasonable” nobody objects. Google managed to condition us to trust that it will “do no evil”.  That is the real power of Google brand!

It is not a surprise that more and more are starting to question Google’s “clean image”. Which leads me to the final observation. I am struggling to find a logical explanation for a wide variation in cost of published Adsense ads on my site, which range from as little as 1c to $8-10 per click on the same page. If Google’s algorithm for placing and costing ads is so smart why such a variation? Wide swings in demand and supply or is something else at play here? If I was getting only 1c clicks I would not bother with Adsense. And probably not many others would. On the other hand, at $5 or more per click Google’s cut would be “too small” and possibly not many would be interested in advertising if Google added their huge margin on top of that $5. So, I can only speculate that variations in cost per click are the result of Google “moderating” the overall advertising revenue for the site.

Firstly, have a look at the graph of post-Penguin search traffic to my site. What strikes me is a perfect symmetry. Search traffic volume used to fluctuate with various events and seasons of the year but the last 4 months has the characteristics of a perfectly moderated result. As if someone had deliberately predetermined how much traffic should be directed to this site – no less and no more, just exactly how much it should get. It is not as crazy as it sounds if you combine it with another interesting phenomenon I observed regarding advertising revenue. That is, if one month is unusually high (ie. “something” went wrong with moderating) the following one or two months have lower revenue – so the monthly average keeps relatively constant. Also, if a given month starts slow, there are always a few days when revenue jumps significantly so it averages out at the end. It is easy to hide “the patterns of intervention” for sites that rapidly grow or decline in search ranking across many keywords (ie. have wide swings of popularity), but they became quite obvious for a site that is referenced with only a few keywords which have relatively stable use pattern over time, as is the case with those pointing to many pages on my site.


[Googe Search traffic pattern: Pre and Post-Penguin]


All in all, it looks that the latest Google algorithms are designed to deliver the right amount of search traffic (ie. by adjusting position on Google Search pages) to get the predetermined volume of clicks that totals a predetermined amount of money for a particular Adsense affiliate. Why? To keep things “reasonable and balanced over time” so nobody objects. After all, Google still needs Affiliates to deliver the traffic for ads at cheap price or dispose of excess inventory that, for whatever reason, it does not want to serve on its own pages. The big reward for all that effort is that Google gets to keep as much money as it deems appropriate.




# Final thoughts...

So, to summarise this whole theory, Google released Panda and Penguin to identify less sticky sites and move them to the end of the search results (ie. to reduce traffic to those sites). At the same time it ramped up advertising efforts on Google Search pages – exploring a well kept secret that poor search results will make people to click more on the served ads. The bonus is happy big customers as traffic to their sites is increasing. This exercise also increased revenue from advertisements for displaced sites. On the micro level, Google is moderating traffic and share of advertising revenue to a bare minimum Affiliate partners will bear without too much fuss. In order to prevent the big backlash from the public Google’s PR machine is spinning “fixing the web” message as the reason for changes.

I would like to believe that Google is doing the right thing and has my best interest in mind. After all, this is what it was telling me and millions of other users from day one. I would like to believe that Google keeps separate its Search, Adsense/Adword and Analytics data and respective future development plans so there is not a chance of any conflict of interest. I would like to believe that what I described above is just a series of unrelated coincidences that are only specific to me personally and my website but sometimes I feel like screaming in desperation... As with any relationship, once you lose trust in the other party (whether it is a spouse, friend or business partner) the relationships gets onto a shaky ground. My trust in Google is somewhat shaken right now. I am not walking out of this relationship just yet but Google, I put you on notice...