I definitely agree with what you said, marinersk. From the context, I would think that what is slow is collecting the data on the Internet, therefore BrowserUK's solution to multithread the web scraping part is certainly the first solution that came to my mind and I believe most probably the right one. Having said that, sundialsvc4's response also makes an interesting point which is worth taking into consideration (and can very easily be tested with minimal code changes).
The point is that we would really need profiling of the program to figure out what is really slow.
