The Internet can be slow, so the less data sent through it repeatedly, the better.
Imagine walking to the library every single time you wanted to refer to a favorite magazine article. What you do instead is keep a photocopy in your desk drawer. Whenever you need to refer to it, you pull out your local copy—a lot faster than making a trip to the library!
The same mechanism is used to speed up your web browsing experience. Your web browser keeps local "photocopies" of pages and graphics it has seen recently. When you click the Back button, the browser doesn't request a fresh copy from the web server—it simply pulls up the local copy it already stored. When the drawer fills up or a file gets old, it gets purged automatically. This local desk drawer is your browser cache.
Major ISPs, corporate networks, and online services take this concept a step further. Because their internal networks are extremely fast, they build a massive shared "desk drawer" for all users connecting through their network. This is a caching Proxy server.
When your browser requests a web page or advertisement banner through a major ISP, the Proxy server first checks its shared drawer to see if another user recently requested the exact same item. If found, the Proxy server delivers that local copy straight to your screen without reaching out to the originating server across the slower Internet. If not, it retrieves an original, saves a copy in the shared drawer, and forwards it to you.
Because web servers only record direct requests for originals, traditional web server log files miss a massive portion of repeat traffic. As industry experts have noted, server logs fail to measure actual human eyes because the origin server has no way of knowing how many cached copies are circulating across proxy servers and browser caches. Real traffic numbers are often 2 to 3 times higher than what server log files report.