One aspect of good web programming is returning easy-to-use errors. If your application detects no coding errors, the webserver returns the error details in the browser. In some cases, the error details contain user names, passwords, and database-specific data that you do not want visitors to see. Most users bounce off a serving website, but not Google. Googlebot is just a crawler to scan the web for content, index it, and then present the results to search engines. If your web server returns coding error errors, Google indexes the content of these errors, which could mean that confidential login information is indexed. The result is that your web server provides search content that can be used to hack your website.

 Google Dorking refers to the process of using common error phrases that relate to a specific response code generated by a programming language. For example, your PHP application might have the wrong database username and password stored in your backend code. PHP returns a database-specific login error that should be handled when encoding an application. "Handling" an error refers to catching it and sending a user-friendly response to the browser. If you don't handle the error, the database login credentials are returned to the browser. When Googlebot visits your pages, it sees the PHP errors and indexes the content. This content is available to anyone who knows what PHP phrases to search for.