This depends a lot on your project and team, which I address a bit in my response you your next question. Of course, if you do good work, make progress, and don't repeatedly make the same mistakes in code you submit for review, your host and team are likely to be happy and give you a good review. Google also has this amorphous (and admittedly subjective) notion of "impact" that sometimes creeps into Googlers' assessment of work (especially if they weren't directly involved in it); for instance, launching a new project is often perceived as having higher impact than adding a feature to an existing one, and therefore a bit more likely to attract interest outside your team. Of course, although you'll have some input into choosing your project, it's not something you control.
In my experience, communication is key. You will definitely not be familiar with the tools that you will use on your project at first, but you definitely have the knowledge and skill to figure it out. However, as is the same with learning anything new, you will have questions, and the fastest way to progress is to ask as many as possible.
I would say the biggest challenge or skill necessary for me is adaptability. The metrics and program change quite a bit, Google's a fast moving digital company. They are always analyzing and finding new ways to do work. There is a lot of analysis and interpersonal skills, but I think you need to be comfortable with change.
Don’t be afraid to ask questions. Obviously, attempt to solve your problems yourself first, but once you get stuck, there is no reason to waste hours trying to debug something that could be easily solved by asking for help. One of the great things about Google is that many of the tools you use will have been made by another Googler, so take advantage of the resources that are available to you. Nobody minds an inquisitive intern!