Web Services are generally used for inter-application communication.
For example, consider a web application that provides aggregated data about trading activity from a number of exchanges.
Each exchange would offer a URL that provided the required information using, let us say, SOAP. Our application would make HTTP requests to the several exchanges and combine that data. It would then make the aggregated data available in whatever representations it chose - perhaps as a web page, an RSS feed, and a SOAP endpoint that could itself be queried as a web service by other applications.
(The question of whether the application would gather all the data every time somebody requested one of these representations (meaning a lot of stuff flying over the net), or refresh a cache of the data independently of client requests, or any other possible approach is something to be considered in the context of your specific requirements. One other approach might be that the application also exposes URLs designed to have updates POSTed to them; the exchanges then push the changes to the application as and when they update their data.)
It's probably worth looking around in the W3C's
Web Services area. Their recommendations (which is the term they use for "standards specifications") define the relevant protocols and so forth.
Also note that, while enterprises tend to go for massively over-engineered solutions using SOAP and WSDL and so forth, there is actually no difference whatsoever between what they are doing and what an amateur web developer achieves when "mashing up" data from (say) flickr, Upcoming, and Twitter for display on their blog, using more lightweight mechanisms such as JSON via REST APIs
