Give them creative freedom to solve your idea in their own way. You may understand the technical specifics and big picture of what you're trying to do, but acknowledging that they know their craft better than you do (that's why you're talking to them) and that you want them to design it out in the way they think is best - just saying that will go a long way. Giving them a mockup helps - but let them know you're ok with their interpretation veering from that.
Obviously if the person you're talking to is an awful designer, nothing will fix it; but if you have a good one - the ideas that will flow will make your product awesome (even if they aren't implemented in the end).
This allows you to describe how you want the system to behave without getting into the nitty gritty on how to make it look good: that is why these mocking systems look as pencil like as possible. Iterate with a designer a few times with these mockups.
I would also recommend showing the mockups to your stakeholders and/or customers and ask them what they think. Do they like the flow, etc.
Another advantage using mockups is that once you have your mockups finalized, you can start implementing the behavior of each mockup while the designer designs.
For this phase, I would recommend using Feature/Behavior Driven Development to drive the development of your code based on the mockups. Use Test Driven Development to fill out the units of your program.
Just a suggestion and I hope this helps.
+1 for the idea by @ericHosick. We use balsamiq to wireframe the details. We take inputs from team to review the details of fields & overall menu placements etc.,
And then we let our designer take it up from there with freedom to come up with various options on his own within the framework. It really helps the designer to work with freedom knowing that the overall content is "almost" set and he is free to implement thoughts based on usability and design aspects.
Edited to add: this is where you'll find out if UX is part of their process. If they begin with things like mood boards, font selections, color palettes, look & feel, no, no and no. It needs to begin with some form of audience/customer, business goals, content strategy, user flows, functionality. Those other things come after, not first.
The reason I like to work like this is because until I work out how the software use actually flows, it's pointless to make it look shiny. And working out that flow itself is a rapid iterative process, which changes the UI many times over. I hate to make my friend design stuff which I will then throw away the next day, although some of that is inevitable.
Over time as the design gets clearer, we work out a visual grammar of sorts, which I can then just draw from as I build more features.
See http://giniji.com for an example of a work-in-progress.