You are hereProject: SERP Tracking & SEO Analyzer


Project: SERP Tracking & SEO Analyzer


By hagrin - Posted on 05 September 2006

Finally, I've had a chance to play around with one of the Google APIs. I decided to take my crack at the Google Web Search API since search is the cornerstone of Google's operations. After reading the documentation briefly (and being surprised that they included .Net sample code), I started brainstorming what tools I could build. Eventually, I decided on a SERP tracker/analyzer.

Now, I have quite a long way to go, but I have successfully called the API, extracted the data from a search results and stored the information in the database. For the time being, I am writing the application in VB.Net and utilizing MSFT's SQL Server 2000; however, I have every intention of porting this to other languages and platforms. Finally, I plan on releasing the source freely since the Google Wed Search API has a limit of only 1,000 queries per day and a maximum of 10 results per query. With those limitations, the only way that the "community" can really benefit from this is to run their own specific relevant term searches.

Hopefully, the feature list will include (but not be limited to):

  • Port application to .Net, PHP, Python, MS SQL, MySQL, flat text file
  • Allow users to dynamically update their API Key
  • Allow users to dynamically change their search terms and search depth (without violating API terms and conditions)
  • Store historical SERP data
  • Web interface to display contextual charts and graphs based on filtered criteria
  • Evaluate high ranking pages and upwards moving ranked pages to determine why they are so effective including comparing page rank, page titles, page summary, keyword density, META tags, keywords in headline tags and more commonly used SEO techniques.
  • Track daily application usage (to prevent API ToS violations)

The picture above shows, in simplistic form, how the SERP data could be stored in the database. Eventually, by normalizing the tables outward, I can do further analysis on the pages as I parse the sites returned as the Googlebot would see them. I'm really excited about the potential for an application like this and the SEO implications. I'm sure many tools like this exist, but I hope to deliver the application and the resulting data in an efficient and unique user experience while also making it incredibly easy for someone to setup my application using their own API key and search terms.