There are some critical decisions that needs to be made that can effectively make or break a rogue campaign before it even starts.

Are you going to be transparent to your user that you're advertising, or not? They both have there pros and cons that break down very simply. 

• If you tell people that this campaign (albeit whatever the shape, model and delivery) IS indeed advertising for your brand you might get the standard resistance that any consumer now has when confronting a good idea but allowing themselves to be jaded because they feel as though they're being sold on something.
• If you decide NOT to tell people it's marketing you stand the risk that if you're caught, before you choose to be, that you will alienate the consumer and create a mountain of distrust. Most agencies that goes this route do so with an approach that is not as risky (from a public perception standpoint) should they get caught the effect can be easily mitigated.

From the professor:

 As David Colburn esteemed Professor at St. Lawrence College stated to me discussing this topic, he says "For me, the way to mitigate risk in a campaign is to answer the most basic questions of marketing:

1. What is the brand message?
2. Who needs to hear it?
3. What do you want them to do?
4. Where is your target audience? 

Go talk to them where they live. If that means finding a new channel -- great! Risk is diminished if you're clear on your answers to the above questions."

I agree with him totally.

After reading my first post it almost sounds too cavalier as if you SHOULD do a rogue campaign, make it super kick ass not matter if it provides any ROI or valid messaging or not. Nothing, of course, could be further from the truth.

Often your 'safe' or traditional campaigns are more informative. "This is new" and "buy this because it's now neutralizes odor" whereas the rogue campaign can offer, more often than not, a platform for entertainment. You're not bound by the need to make sure the audience understands that the Scrub-master 9000 now come with carbon push-rod technology.

Cliff's Notes:

If you trust your agency, and you should otherwise you're paying to much, you should run two (even three) creative executions of your campaign. Vary not only the riskiness of the messaging but the media delivery vehicles as well. Then test and optimize before pulling the plug when you get nervous.

Have you thought of running with an edgy campaign? Do you run more than one campaign idea at a time? If yes, how do you test your campaigns? I'd love to hear more about your clients and campaigns.