Earnings and traffic are definitely intertwined. Google AdSense's tools provide only limited functionality when it comes to telling publishers what works best, in the form of the channels that we talked about in the previous section. Unfortunately, Google limits the number of channels to 200 (up from 100 as of August 2005!), making it impossible to assign a channel to each file for sites that are made up of many HTML files which need to be tracked independently, such as photo albums. As a result, it is important to find alternative ways to determine what generates earnings. However, without Google's help, it is virtually impossible to accurately track what goes on. Tools exist that help publishers track the numbers of impressions and the number of clicks each page generates. AdsenseLogger is one such tool, it is open source and works quite well. Commercial offerings also exist (I have no experience with them). While these tools can go a long way in giving publishers a better idea of what goes on, they do not link page impressions and clicks to actual earnings, simply because Google does not provide that information. Pages with similar CTRs and number of clicks can generate wildly different earnings due to the different amounts of money advertisers are willing to pay to have their ads seen on those pages. That is, of course, due to the fact that these pages most likely target different keywords.
Given this state of affairs, whatever information publishers can gather from their website statistics will most certainly help. All decent hosting plans provide some form of usage statistics, but many are pretty basic and do not include things like visitor paths, browser resolutions and other statistics that may be important to have access to. Commercial offerings exist to fill that need, but they can be quite expensive. An interesting compromise is StatCounter, which only analyzes the last 100 hits, but is free as long as the site does not exceed 250,000 pageloads a month.
An important thing publishers may want to keep an eye on is the type and percentage of HTML errors. By monitoring my stats I found that Yahoo's image search was directing visitors to invalid URLs on my website. For some reason, Yahoo was adding a "~" in front of the top-level directory in the URL, i.e. "http://hulubei.net/~tudor/photography/T/u/Tulip-Flower-1.html" rather than "http://hulubei.net/tudor/photography/T/u/Tulip-Flower-1.html". Mistakes like this can very easily be handled through Apache redirect rules on the server, recuperating traffic that would otherwise be lost:
Redirect /~tudor http://hulubei.net/tudor
«Previous 1 2 3 4 5 6 7 8 9 10 11 Next»