A trivial way of finding an element in a sequence is looking at every element of the sequence. If the collection does not imply some specific ordering this could be the only way of searching. However, large collection will definitely cause a performance impact.
Binary search is an example of an algorithm that with some amount of preprocessing is able to solve a hard problem. Sorting a collection of elements allows consecutive searches to be really fast.
The strategy that this algorithm employs is very similar to the one we saw in merge sort and quick sort.
You can go through each step of the algorithm here.
Can we search faster than binary? Yes, if we once again limit the problem and require knowledge about the range of the values in the searched interval. If that's the case Interpolation search does a pretty good job improving the performance of binary search.