Just a quick theory about Panda. Google essentially does an economic analysis on all sites just like with its Adwords system. If you have Adsense on your pages, your site is pushed to the top of the queue. You also have big brand sites which cannot be affected by panda because users recognize them and wouldn't use Google if they didn't exist. Google knows exactly the amount of money they can make if they get a click on your site (via your ads) versus simply sending them to the ads within their search results. Therefore, your site gets hit by Panda, if more money can be made by just sending them to an ad on their search results page. So, search results are purposely degraded to a certain extent (there is some sort of user satisfaction algorithm in place so they are not degraded too much) to send more clicks to their ads on their search results.
You can also look at it from an advertiser’s point of view. Google essentially lowers the amount of available inventory for ads by demoting publishers. Therefore raising the CPC amounts that advertisers pay, creating a win, win scenario.
This can also be applied if your site competes in a niche with a Google product or a potential Google product. The fact of the matter is why share advertising revenue with you, when they can keep 100% if it’s on their own pages. So, each iteration of Panda increases Google's profitability by eliminating the percentage paid to publishers hit by panda (not completely because they do have to send traffic to some publishers so they could potentially keep 15 or so percent of the difference between your site, the new publisher result, and the profit from their ads on their own search results (insert the big data analysis here)). So, each panda judgment is ultimately done with Google's bottom line in mind.
Does Panda actually make some judgments about the quality of the website? Yes, this is true, but I argue it plays a much smaller role than they say it does. Because with this latest Panda Update you have people saying it happened on the 15th, 22nd, 23rd of February etc., etc... Rather, this essentially was when the economic analysis of these sites took place by machine or human with help of machine (Far fetched, but I can literally imagine a red or green light (true or false) based upon a very extensive algorithmic economic analysis that is shown or not shown to a human).
Why could this theory be true? Because Google hasn't come up with any good ideas lately to increase profitability, and android ad revenue only accounts for 2 billion in Revenue (correct me if I'm wrong on that). They have a share price and investors to worry about.
What can small to medium publishers do? Unfortunately, not much. I only see this Panda problem getting progressively worse as more and more traffic is squeezed from publishers and sent back to Google. One area of hope for the internet is the democratization of it through other forms of traffic such as Twitter and Facebook (both hate Google) and indicative that competition is good.
Some final notes. I think it would be interesting to do an analysis of the dates that Google does Panda and the resulting increase in traffic back to Google Properties. A site by site analysis would be difficult but something a data warehouse like Alexa could do. It would also be interesting to use the corporate quarterly statements and compare the resulting increase or decrease to Google Properties and Panda sites based on Panda dates.
Finally, it would be cool to start a search engine dedicated to websites hit by Panda. Quality and if a site contains malware (toolbar’s and unwanted browsers) are incredibly subjective measurements especially when there is a big profit motive. I bet the quality of the results to the “user” might be surprisingly high.
Disclaimer: This is speculation and is an opinion
[edited by: tedster at 10:23 pm (utc) on Feb 29, 2012]
