So, I would think about putting a couple of constraints on this pricing problem. In the paper I go back and try to motivate these constraints from economic theory, but in this case, essentially what I'm going to is I'm going to say that the average store price has to remain unchanged. And that the category sales have to remain unchanged. So if you think about this from a consumer prospective, or a competitor perspective, if my average prices aren't changing the way a competitor would look at it, isn't look at the prices of each product but let's look at the overall prices that this competitor is charging. So what they would do, is they would go out and see, look, their average prices haven't changed so I don't really have an incentive to try to make a difference in my pricing strategy. From a consumer's standpoint, essentially what you're doing is, look, the average price is $1.54 before I do the micro-marketing strategy, after I do it, I'm still going to leave that price at about $1.50. In terms of category sales, I'm going to say that the category sales are not going to change. So if I've been selling 10 million dollars in orange juice before I implemented this micro-marketing strategy, I still want to set 10 million dollars after I implement this micro-marketing strategy. So, essentially what I'm doing, is saying that there's something of my model going on. I'm still trying to come up with these posterior profit functions, what I'm going to do now is I'm going to go back and lets think about what's the posterior profit and what's the posterior distribution of these constrained profit functions?