As you can probably imagine, it's difficult to come up with a topic when the objective is not defined, and there's no specific indicated action based on the findings and conclusions of any research.

Right now many/most marketers are fixated on big data and web analytics, so you probably want to include those as you look for useful/relevant topics. My inclination would be to talk with some of your clients and see what THEY suggest. One of them may be facing a situation where your PhD topic could be particularly useful.

If all else fails, what about looking at the interaction of brand loyalty and retailer loyalty. Which is more dominant/important in selected industries? In which industries is one more important than the other?

There is a real problem with "big data" - how much of it is the result of poor thinking? It is easy to set up a measure, it is not so easy to set up a measure that is actually meaningful. 'Garbage in, Garbage out' as they say. This is more common than you might think, just look at split testing - as the master Glenn Livingstone says, "most people test two things twenty times, instead of testing twenty things twice". The data you get from the first will be largely meaningless, the data you get from the second will be meaningful - simply because you asked clear questions. It is harder than you think, and given the state of PPC, more common than most people acknowledge.

Tying that into what Mr. Goodman has said might give you some interesting avenues to pursue.