Introduction
In 2006 we developed a spreadsheet that helped us to increase our Adsense profit/earnings by approximately 45%. We have now made the spreadsheet available for use, at no cost, online. This page describes how we used it to increase our earnings.
(If you are just after a few tips, then see our page on how to increase Adsense profit. Alternatively, if you have a fast internet connection you can watch our video tutorial on how to use the Adsense Simulator)
On Site Testing
The main way for publishers to work out how to improve their Adsense profit has, so far, been on-site testing. This can involve waiting days or even weeks to see the impact of different ad placements. Also, such testing affects your real Adsense profit, so if the new ad placements work poorly then testing can be a costly process.
We therefore developed an Adsense Simulator, based largely on Google's published algorithms, so that we could circumvent much of the testing process. It enabled us to experiment with different ad placements, to see the likely impact on profit, but without affecting our current income.
We started using it to assess the impact of different ad placements on our existing pages, within the limitations of our overall site design. After a lot of offline trial and error, we tested out the best configurations on the live website. We noticed an immediate improvement. When we assessed comparable two-month periods before and after using the simulator, we found our earnings had increased by 43.5%. Our earnings are also more consistent, so we seem to have moved out of the 'smart-pricing trap' (where income can sometimes fall inexplicably despite having a healthy number of clicks).
Some of the key lessons we learned were that some of the less popular channels can make a big difference. We had many pages where we had removed ads from the bottom of the page as they did not have a high CTR. However, the results from the adsense simulator suggested that these adverts had a critical role, because of the smartprice effect.
There is an assumption we have used in the simulator that ads that are outside the immediate hotzones are more likely to convert. What the simulator suggested was that having those low CTR, high converting ads on a page can significantly affect the earnings of the whole page.
We've designed the public version so that publishers can tailor it to fit their own circumstances. It is in a spreadsheet format where the numbers can be adjusted so that you can start off with a near-replica of the reports you see in your Google Account. Also, it doesn't impose any particular page format but uses a 'Position Weight' variable so that you can stipulate the degree of prominence your ads have in your current page design.
The simulator won't replace the final stage of testing, but it will allow a lot of the earlier experimentation to be done in a few hours that otherwise would take weeks or months to complete. There is a lot that Google keep secret so we know the simulator cannot be fully accurate. Also, we have had to make assumptions about how smartpricing works. However, the impact on earnings can be seen with and without smartpricing, so you can make a judgement on whether to use our calculation or not.
Along with the Adsense Simulator, we have also published an online course. This is also free-to-view. It uses spreadsheets to illustrate some of the basics of the Google system, such as Ad Rank, Quality Score, the relationship between bid and price paid, and smartpricing.
The Adsense Simulator is available online: Adsense Simulator
You may prefer to do the online course first as this explains hw to use the Simulator.