Here's what I would do.
Choose a few applicants you like and pre-interview them on what they're looking for, management experience, what kind of budgets they've been use to working for and such. This is for the "bonus" aspect.
Of those take the responses that are the most aligned with working in tight budgets and timelines and ditch the ones that are expecting big name clients with months to spend on a single campaign.
Now for the actual interview here's my suggestion:
Pick a generic topic, anything. Ahead of time take yourself and current coworkers (or friends/family if its just you right now) and do this exercise:
60 seconds to write down as many things that pop into your head as possible.
Let's say its you and John Doe that do the pre-test with the topic beer bottle:
- You write down: drink, smash, crack, shiv, candle, light, blow into
- John writes: drink, candle, light, blow into, share, ice, miller, recycle
Now when an interview shows up give them the same topic and same 30 seconds. They say:
- Interviewee: drink, candle, throw, share, malotov cocktail
Could score it 1 point for each item and an additional point for anything that doesn't appear on one of the pre-interview tests. So throw and molotov cocktail would each be worth 2 since they're "unique"
This is how I would go about testing someone's ability to come up with creative ideas quickly. The interviewees that score the most points either by coming up with the most ideas, more unusual ideas, or combination of both are the ones to hire. (Pending the rest of the interview of course)
Make sure to do the pre-interview test with a few people. Otherwise you won't have a very good methodology to know which answers are more unique then others in the event two interviewees come up with the same number of ideas.
Edit
I want to point out that others are objecting to the "spirit of the question." I'm not saying I agree with doing this or agree with what @blueberryfields is looking to do. But the question remains - what methods are available to determine if someone has the "ability to consistently quickly come up with good, original, creative ideas "
The methodology I'm proposing is not something I'm just making up here. It's a classical test of creativity developed by J.P. Guilford known as Alternative Use Test or Alternative Use Task. It is commonly found as part of larger Torrance Tests.
References: