One of the contributing factors to its speed is the fact that Go is statically typed. This makes the language ideal for building systems at a large scale and being confident in how your program will run in production. Also, since Go programs are built with a compiler instead of being run with an interpreter, it allows you to catch more bugs at compile time and greatly reduces the dreaded runtime errors.
This safety net is also extended to the Go garbage collector. Garbage collection means that you do not need to manually allocate and deallocate memory. This helps prevent memory leaks that might occur from mishandling objects in your code. Some may argue that garbage collection impedes the performance of your application, however, the Go garbage collector adds very little overhead in terms of interfering with your code execution. Many source report that the pauses caused by Go's garbage collector are less than one millisecond. In most cases, it's a very small price to pay to avoid chasing down memory leaks in the future. This certainly holds true for web scrapers.
As web scrapers grow in both size and complexity, it can be difficult to track all of the errors that may occur during processing. Thinking on the scale of processing thousands of web pages per day, one small bug could cause significantly affect the collection of data. At the end of the day, data missed is money lost, so preventing as many known errors as possible before the system is running is critical to your system.