Note that I have only started using Google AdSense in July 2005. So you might want to double the 2005 stats to get a more comparative number, assuming the traffic level in 2005 is constant. Also although there are still ½ month to go for December 2008, the stats shown above should give you a general idea. I have also changed to a new theme which may/may not affect the revenue generation. Finally, many of my most popular pages on this blog (mainly those WordPress plugin pages + the front page of this blog) don’t actually have ads shown on them.

So, what can I conclude from the stats generated from above?

• Lazy bloggers won’t make truck load of money. Actually, even if you are a heavily effective pro-blogger who can consistently produce a few well-written insightful blog posts a day, you probably still won’t make as much money as an investment banker or a seasoned J2EE/.Net engineer. Your blogs probably won’t sell for more than their redundant packages either, in the current economy climate.

  Still, a few hundred US greenbacks are more than enough to pay for the web hosting for this site and many others (which are currently on USD$20/month SliceHost VPS). Still got some spare changes for a treat every now and then…

• People are getting more and more blind to the online ads. While the number of impressions stays pretty constant, the number of clicks (and thus the click through rate) has dropped year after year. Assuming Google has maintained the quality of ads shown on this blog (finger crossed), I can only concluded that people have fed up with the online ads and become blind to them.

  I guess that’s probably the reason why there are more and more aggressive ads on big news sites like SMH, where they take over the whole screen to make sure you see them. Did I say aggressive or annoying?

• Not just CTR, CPC has been dropping too! If you divide the revenue by the number of clicks, you will get the Cost Per Click (CPC), and you’ll see them drop year after year as well. That means you now need to get 150 clicks to make the same amount of money where you used to be able to make it with 100 clicks. That sucks more because CTR is dropping at the same time.

  I guess there might be multiple reasons behind it. Google AdWords bidding is still a mystery that I don’t think I will ever understand. No one wants to bid on my site anyway after seeing all the crappy writing I have here. The financial crisis and economy downturn has also reduced the marketing budget of many advertisers. Finally there are many more websites parked with AdSense to compete for the cake. No wonder my slice gets smaller and smaller.

Well. That’s about my it — my thoughts on Google AdSense on this blog. With imminent recession in 2009, I guess I can pretty much predict the number..

It was fun. It was rewarding — maybe not in financial terms but more of “beating the system”. It is exhausting — yes it does take time to read, write and tinker with small details. Despite my previous dislike of banner ads, I think I now have a better understanding on how the whole thing ties together, and will continue to put those little text ads on this site (if Google permits).

First of all, here’s a chart of impressions verse earnings over the last 12 months. Sorry no actual number but feel free to guess what they are :)

Things I have observed after running AdSense for 12 months on two hobby sites (playground and bargain blog):

• Figures don’t just go up. Yeah, the first few months are pretty much going up and up, but then it stalls, went down a bit, and up again. A bit like a roller coaster ride than highway cruse. Many reasons contribute to this.
• It eats your time. I don’t write very fast (although it did make me write faster). Nor did I think fast. So when you are busy — church, work, family, etc — you have less time working on websites, thus produce less traffic, thus less people click, thus less income. But things ain’t that simple.
• Google is a mysterious blackbox. Most of the time I just have no idea. Why some clicks pay more than the others. Why some pages have better targeted ads. Why some pages rank better thus bringing more visitors. Google is mysterious — and it is not a good thing if your income depends on it.
• It won’t replace a full-time job. At least in my case, it never will. It pays peanuts in comparison with my day time job, and I do not see how I can actually make decent money even if I work full time on it. Probably I just don’t exactly know how it works, yet. Or maybe I just don’t have any quality stuff to write.
• Still a nice pocket money. Especially when you are blogging anyway, and the extra pocket money from the big G is just enough to pay for all my hosting (currently 2x VPS + 2x shared hosting), my ADSL connection and all phone bills. No complain :) Can’t bring Vivian to eat lobster too often though :)
• Content is *not* everything. You still need good original quality content though — diary style blogs don’t really work. However, a good informative piece of article does not always lead people to advertisement, and most of the time you need to work out psychologically “why do people click on ads”.
• Attract visitors in “spending mode”. Thus organic traffic is good — people came in search of “something”. Review a product positively is important, and I suspect I can never be a good marketer as I am always too critical on things.
• Traffic! Traffic! Traffic!. After looking at my statistics over the past year, eCPM goes up and down, but always within a range. Therefore in order to increase the earning, you definitely need more traffic. My traffic is pretty much organic, and 80+% from Google. I guess there’s more to work on.
• Niche! Niche! Niche!. Well, traffic is important, but each niche is different — they have different advertisers, and they attract different visitors with different views on banner ads. This site is a mix bag of things. Bargain blog does well as visitors are usually in “spending mode”. FuCoder.com does very poorly (0.x% CTR where x < 5) — I guess programmers don’t click on ads, thus I just take them all off.
• Ad position is critical. Moving the skyscrapper to the left-hand side on bargain blog helps. 3 fold!! Too bad that I don’t really have time tinkering with the ad format. Nor am I a fan to blend the ads right inside the content (way too ugly). Still, AdSense heatmap works.

Well, I am sure there are lots more that I have acquired though this exercise. Except I cannot all share here as it’s almost 12am and my brain has SpeedStep to 10% of the performance. I’ll add more onto the list, if I remember.

Another interesting use is, using this preview tool to see how Google categorises your website. On my index page, I got:

• Free Bible Study Course (who are United Church of God?)
• Apostolic Sermon Outlines (pentecostal sermons, yik!)
• Martin Luther King Jr. (biography software)
• Voice of the Martyrs (VoM Australia — do they advertise as well?)
• Flying Colors (get rich quick via property scheme)
• Free Bible Software (enter Jesus survey and win $300, what?!)

Seems to be Christianity-biased, although less than a quarter of my posts are about church and Christianity. Still, more than half of those ads are dodgy, and now I feel sorry for those who have actually clicked on them.

Update: I actually wondered the same thing two and half years ago!

At the same time, a new site that makes $2 per day (which won’t make Google referral payout requirement) might be brining $2,000 a day in 12 months time, keeping both the publisher and Google’s shareholders happy. Who knows. Except the original referrer who missed out on the reward.

By the way, I’ve found ThreadWatch.org getting rather negative about Google after change of ownership. Maybe it is just an art of link-baiting?!

Now, shall I frame it (my very first cheque from Google), or shall I cash it? Hmmm…

LcF blogged about this sale, concluding:

Weblogs Inc. is a successful blogging network. Many of its blogs are top blogs of the Wold. The network earns USD$1 milllion from Google Adsense in a year. That’s still exclude the premium advertisements revenue. USD$20 millions is a good deal.

I have never seen USD$20mil in my life, but was it a “good deal” because Weblogs Inc. has potential to generate 1 million dollars a year “revenue” from AdSense? Let’s say, WIN gets 1.5 million dollars from Google, and another 1 million dollars from other advertisers and affliates. However what about its cost? Besides site hosting, domain name registration (which are negligible), other infrastructure cost, they still need to pay 100+ bloggers under their wings, quite a few staffs and Jason himself.

What’s left as “proft”?

Besides, being a subsidiary of AOL — would that damage their relationship with Google? Wouldn’t AOL resale the advertising themselves without profiting the “middle-man”?

I do not think it was the scenario in the AOL board room a few weeks ago:

Adviser 1: Weblogs Inc — They are just n00b in the media network.
Adviser 2: But Jason claimed he can make one million bucks from Google!
Adviser 1: OMG! w00t! Maybe we should buy them.
Adviser 2: Would 20mil be enough?
Jon M: they pwned.

I am sure AOL has its plans for its newly acquired blog network, but I don’t think “ability to generate 1 million dollars a year from AdSense” is the reason why it has been bought.

Well, I guess I am only doing it just to cover my costs. I have got a nice developer job during day time, and thinking working over-over time in a competitive world for almost 365 days a week like these guys are just insane. However, I always felt fascinated by all these techniques on search engine optimisation, and in the case of Google AdSense, the “AdSense-optimisation”. It turns out having a sophisticated reporting tool is important, so you can know exactly how ads have been shown, and which ones have been clicked on. All the details can then be feed back into optimisation, helping the webmasters to build their next golden goose.

However, AdSense’s reporting tool might be nicely characterised as “simple”, or cruelly described as “lacking”. You have your daily total of impressions, number of clicks, and how much have you made — and that’s it. You can set up “channels” to aid tracking, but I personally find it very awkward to use.

Then there are third party “logging” and “reporting” tools that let you track impressions and clicks in real time. Do a Google search on “AdSense tracking“, you’ll find the net is populated with these tools. There are some free (in beer) and open source efforts, and the most popular amongst them is currently the Adsenselogger.

How does it work?

Since Adsenselogger has been GPL’ed, it is not difficult to look at the mechanics. It requires you to add an almost-2kb Javascript snippet at the end of your document, before </body>, and it does two things when a visitor viewed your site with a Javascript capable browser.

1. It loops through all IFRAME’s on the page, looking for ones that point to Google’s ad servers, and then reports the impressions to a server-side PHP script.
2. When a click has been made, the script will then report which ad has been clicked, and try to figure out the target URL of the ad.

It does the reporting by creating a dummy image object and then assign it with the crafted URL. After the server-side script has collected enough impressions and clicks, you can then use some of the pre-designed reports to work out how you should optimise your AdSense site.

Adsenselogger Issues

The only reference I have is Adsenselogger 1.1 posted on their website. I know their developers are working on 2.0, but I can only evaluate according to what I can get my hands on.

There are quite a few issues that make it less and ideal tool to track your AdSense traffics.

1. Click-capture code only works for Internet Explorer. It uses only onfocus event on the IFRAME window instance to track whether an ad was about to be clicked, which does not work with about anything but IE. On Mozilla/Firefox, you really need to attach the onfocus event onto the contentWindow instance, i.e. the window container object of the document inside the IFRAME. However due to Mozilla’s security model a script cannot window/document objects from a different server.

   I did manage to build a prototype to capture the click event by using a combination of onmouseover, onmouseout and onunload events, and that works with both Firefox and Opera. Again it is tricky with Opera, as its onmouseout event is not that reliable on a IFRAME window, so my solution needs to make an educated guess on whether a click has been generated or not.

2. Adsenselogger only can track one ad unit on each page. AdSense TOS allows you to have three ad units + 1 adlinks on each individual page, and its impression/click code does not take multiple ad units into account.

   Again, it can be easily fixed by taking individual ad units into consideration, before packaging the impression logs to be sent back to the logger.

I have since been writing my own AdSense tracking script that work on at least IE, Firefox and Opera — to track my pathetic AdSense earning records. My plan is to write an architecture that is generic enough to adopt to other ad networks as well, just in case 6 months down the track I need to re-code for YPN or MSN ads tracker. Since I have got my inspiration from GPL’ed Adsenselogger, I shall release it as open source when the code is stablised.

Against Google’s TOS?

Because those AdSense trackers relies on binding extra event handlers onto objects created by Google’s own scripts, I was wondering whether it is against Google’s terms of services. From Google AdSense’s Program Policies, it says:

Code Modification

Any AdSense ad code or search box code must be pasted directly into Web pages without modification. AdSense participants are not allowed to alter any portion of the ad code or change the layout, behavior, targeting, or delivery of ads for any reason.

Does attaching extra events to trap clicks count towards changing the behaviour or delivery of the ads? I searched the web on Adsenselogger Google TOS, and most results are saying that there is no issue what so ever. Basically mos people reckon you can continue using those loggers without infringing AdSense’s TOS.

However, I think I shall just send the AdSense team an email just to be sure. I worded it this way:

… I looked at the code (of Adsenselogger) to see how it tracks clicks, and noted that it uses Javascript to add an event handler on AdSense’s IFRAME object. Does that violate AdSense’s Program Policies?

I guess the real question is, can I use AdSenselogger without getting banned from the AdSense program?

Or, is there a clear guide on what I can or cannot do with the DOM objects created by Google AdSense?

Two days later I received a reply from Allen of the Google AdSense Team:

We understand that you would like to have more detailed reporting regarding your AdSense account. However, we do not endorse or encourage the use of any third-party tools for this purpose…

…Please note that AdSense participants are solely responsible for verifying that any tools or software used in conjunction with AdSense do not violate the AdSense Terms and Conditions.

Aarrgh! I don’t like vague answers — I want Yes or No! However, it did sound a bit contradicting with other sources, where nearly everyone claims no issue with Google’s TOS. So who is right? Or is it a canned message? Google might just be too sick verifying every piece of software whether they are okay or not — Google wants to reserve the right to boot someone off at will.

Anyway. If you are using a third party AdSense tracking software, make sure you know how it works and be very careful not violating Google’s TOS.

I read Nicole Simon’s article on ProBlogger this morning, and he looked at the option of separating your daily readers and visitors, when you display your ads. I happen to have quite a few daily readers who track me through services like Bloglines and FOCUSer.net’s recent updates, and I like to hide ads for them. At the same time, I am also getting a few hundred visitors landed here via Google, Yahoo!, MSN Search and others, searching for some particular information – these are the people I especially like to target the ads with.

So Simon’s solution is – only show ads to the blog posts that are a few days old. A great idea! A few minutes of PHP hacking, here’s a list of rules that governs whether AdSense will be shown:

• Post is at least 7 days old & not in some specific categories.

• Referrer is either Google, Yahoo! or MSN Search.

I am now showing ads in two places – at the end of blog entry, and at the top of the index page. I have PHP/WordPress code that looks like this:

  <?php if (show_ads()) { ?>
  <script type="text/javascript"><!--
  google_ad_client = ...
  //--></script>
  <script type="text/javascript"
      src="http://pagead2.googlesyndication.com/pagead/show_ads.js">
  </script>
  <?php } ?>

And my PHP function that checks whether ads needs to be shown:

<?php
function show_ads() {
    global $posts;
    $referer = $_SERVER['HTTP_REFERER'];
    if (isset($referer)) {
        if (preg_match('/^http:\/\/(www)?\.?google.*/i', $referer) ||
            preg_match('/^http:\/\/search\.yahoo.*/i', $referer) ||
            preg_match('/^http:\/\/search\.msn.*/i', $referer))
        {
            return true;
        }
    }
    if (is_single()) {
        $t = $posts[0]->post_date;
        $t = mktime(substr($t, 11, 2), substr($t, 14, 2), substr($t, 17, 2),
                    substr($t, 5, 2), substr($t, 8, 2), substr($t, 0, 4));
        if ((time() - $t) > 604800) {
            if (!in_category(21) && !in_category(3) &&
                !in_category(40) && !in_category(14))
            {
                return true;
            }
        }
    }

    return false;
}
?>

It checked against 4 different categories (mainly christianity-related) and will not show ads if someone browses these blog posts. I know I should have made it into a plugin, but was too lazy to do so and the code is really very site-specific. I am just putting them here hoping it might be useful for some.

What you need to do now, is keep on clicking on the ads below until I am filthy rich. Bwahahahah!!! (click, click, click)Which of course should not be stated anywhere on this website as it is against Google AdSense’s TOS…

I used to hate banner ads, popup/pop-under ads and even Javascript text ads that I will try to avoid if I can. I have ad-blocking regular expression filtering installed on my Squid proxy, plus AdBlock installed on Mozilla to completely eliminate JS/Flash ads. However, is that stealing, as I have taken away the opportunity cost of the advertisers and content providers? It strikes closer to home as I am about to launch a new site in the next few weeks (not related to FOCUS), which will be my first site with Google’s AdSense, and through selling the advertisement I can hopefully recover some of the operational costs on this server.

The reason I say “opportunity cost” is, advertising only brings in revenue if the user makes a click-through, as most ad agencies are paying on a click through rate. With this point of view, I can safely say that I have never and probably will not steal any revenue from an ad-serving content site, as I never click on a banner ad. But I guess each individual varies. Banner ads survive – because they bring in enough revenue to both the advertiser and the content provider – because people do click through them and make purchase.

Do you click-though banner ads and make purchase? Do you block them using a proxy or browser filter? What’s your view on the ethics of blocking the ads? Would you spend that much time on-line, if suddenly all the worth-reading content sites now demand subscription? That’s to be discussed.