I'd read a book about UX/ID processes just to know what to expect.
Also, you have to keep in mind: there are two ways of UX: model based and research based.
If you hire a UX designer, but don't give him enough money and/or development resources to do user research, the only thing he could do is to say nice things to you which might or might not work.
User research can be setting up measuring technologies, which range from configuring google analytics or click heatmap tests to have two different versions of your site served for a small - let's say, 1 percent - of your audience to test if that version is better. This costs you development efforts to put the needed codes into your system, or to develop the alternatives.
User research can also be based on people, this goes from the cheap (but not free) solution of asking them online what do they think of the site, to actually bringing carefully selected people from the street into a laboratory, which has devices to watch what do they look at and how frustrated are they.
If you give money only enough for him to eat, he can bring you general ideas, perhaps some wireframes, some change recommendations.
These recommendations might or might not have some scientific merit, and they might or might not have any good effects on your product. There are good and worse UX consultants, I've seen too many of them living off by looking confident enough and tell the customer to copycat facebook/twitter/iOS, whatever was their current fad of the day, and when asked for explanation, they told to the audience, "I'm the expert, if you hired mel trust me"
So, beware or who you hire and be wise with your funds, spend them on the things which will actually do have an impact.