or that?” Or, “How do I distribute my trade funds?” We’d like to be able, with a push of a button, to take a national budget, let it allocate across all our accounts, push it down and optimize it. Presto!

DALE HAGEMEYER: (Laughs) No one in the marketplace is so naïve as to think it’s: Just push the button, and out comes the answer.

ALEX RING: Let me back up. Again, it’s a matter of being careful about our expectations. We all face competitive pressures. If we don’t meet competition in certain ways, we lose business and market share. If we spend too much, our profitability drops.

So let’s say for a moment that we have, last week, modeled –- perfectly –- the relationship between our deals and those of a competitor –- pricing, distribution, etc. In other words, let’s pretend I have a perfect model of what to do, depending on what our competitors are doing.

DALE HAGEMEYER: You’re saying competition is consistent.

ALEX RING: I’m saying, Let’s say the model is perfect. It’s not relative, it’s absolute. And yet! We have to be careful about the ROI. Why? Because even though the model itself is perfect, it will work only if I know exactly what my competitor is going to do.

Now, think about this: I have to commit to a promotion nine months from now. Right? How DO I know what my competitor is going to do? Do I even know if I going in to present first? Or has a competitor already offered? If so, suppose I go in and beat their deal. Well, they could come back in and sweeten it.

So you see, even with a perfect model, our ROI expectations have to be tempered, because there are situations where, yes, I see what the model is suggesting, and it looks great. But the fact is, either I really don’t have all the relevant data, or there is simply no way to factor it in.

So we have to set the expectations appropriately — which is no easy thing.

DALE HAGEMEYER: Agreed. The key, in my view, is simulation –- the ability to say: ”Here’s Scenario ‘A’, and here’s Scenario ‘B’. Which do I believe?” If the competition stays constant in its approach, then maybe “A” is the right road. If, on the other hand, I think there may be some deviation from the normal pattern, then maybe “B” is the way to go.

In short, simulation will suggest a couple of options. But ultimately, yes, I’ll have to use my own head to decide between or among them.

The need to use our heads, by the way, is why there are so many companies still struggling with post-event analysis. It’s so much work! Typically, as the joke goes, “post-event analysis” consists of the salesperson pulling out of the store parking lot, looking in the rear-view mirror, seeing the big banner in the store window, and saying: “Yup, we’re good! I got the display!”

In reality, as we all know, post-event analysis has been the bane of CPG’s existence. How DO we pull all that data –– shipment, point of sale, cleared deductions — together? How do we make sense of it all?

Today, we can actually do that in a meaningful way. It’s still not easy. But companies will have to pull those pieces together and automate the process to the degree possible, because that becomes the springboard, their leverage for the future. The capabilities embedded in these modeling systems -– regression analysis and other dynamics — require solid information to leverage: “We did this, we got that. If we do something different in the future, how does our trajectory change?”

So again, it won’t be easy. But it does force discipline, and a game plan that says: “For starters, let’s have a server-based solution that we can use to plan promotions, one that captures and uses the information we need.”

This is a far cry from circulating spreadsheets around, saying: “Here’s Version two-dot-five or Version nine-dot-four. Which one do you have, Alex? … Oh, you have a totally different version because the latest version is buried on my desk. We’re not talking about the same thing at all.”