The magnitude of the pricing problem becomes apparent when we consider its implementation at the individual product level for each store.
In this case we're going to have about thirty products, or thirty different universal product codes (UPCs) in the orange juice category. Now, just to define my terminology, each individual product is identified by a UPC code, which is the bar code that you read on the bottle. Each individual product is uniquely identified. You may wonder, ``well, how do we get thirty different orange juices?'' Orange juice can differ just in the fact that it has pulp or it doesn't have pulp, it's from concentrate or it's premium orange juice, and it's Minute Maid or Tropicana. Once you start multiplying you get a multitude of products. Now, on top of the product problem, you've got 120 weeks which is two and a half years of data. Dominick's has an even longer historical database that could go back 5 or 6 years. They have 83 stores and the notion is that you're not just operating one store, you're operating a lot of different stores. When you multiply the product, the weeks, and the stores together, it means that Dominick's is going to have to make 360,000 pricing decisions. Obviously, this is going to start becoming a very large problem. If you were to think about this in its entirety, and say ``let's not just focus on an individual category, but let's focus on 30,000 products,'' then you're starting to get the flavor of what a retailer has to do to figure out what pricing strategies they need to follow.
What's happening is that retailers essentially feel they're in a very competitive business. They're thinking ``what is my competitor charging?'' and they're going to determine what to charge by saying ``all right, my competitor is charging this price.'' If you imagine that your competitor is charging a uniform pricing strategy, one possibility is to customize your pricing strategy to the individual store level and take advantage of -- or exploit -- some of the differences that exist. This may allow you to better market your product.
Now, that leaves us with a very rich set of pricing strategies, which is really why retailers are going to be interested in this. First of all, it gives them a device to give better service to their customers. In Target's case it's just a matter of carrying the types of products that people are really interested in. In Dominic's case, it's giving people some price discount to buy certain products, and giving them some kind of inducement to buy more orange juice. The first type of micro-marketing strategy, and what all product pricing strategies are essentially in contrast to, is having the exact same pricing strategy across all stores (as discussed in the following section).
What may happen is that Minute Maid orange juice is $1.50, and Tropicana is $1.60. According to the uniform pricing strategy, it doesn't matter whether you are located in Chicago or you're in the suburbs. Every store is going to have the same pricing strategy. Now the first thing that you can do with a micro-marketing strategy is say, ``we know that in the inner city, people are more sensitive to price changes, so we'll uniformly decrease all our prices.'' Instead of having an average price of $1.50, we'll lower the price to $1.40. In some of the suburban stores we'll think about increasing our price to $1.60. So, this is the first type of micro-marketing pricing strategy that I want to mention. Yes?
Instead of just increasing or decreasing all of the prices, perhaps there's a group of products for which we'd like to increase the prices and another group for which we'd like to decrease prices. One example is that you've got private labels, the store brands. For instance, Dominick's sells their own brand of orange juice. And then you've got national brands like Tropicana and Minute Maid. What Dominick's might want to do with the micro-marketing strategy is that in some stores, they might want to keep the gap between private and national brands close. They still would have the discount, because your brand is perceived as lower quality, and you want to give people some kind of price discount to buy your brand. Dominic's might like to price so that in some stores they have a 10 cent price gap and in other stores they have a 20, 30 or 50 cent price gap. The question is not just a matter of increasing or decreasing all the prices -- it's really thinking about competition that is going on between the brands. The idea is that maybe I can induce people to switch to one brand or another by using price.
Finally, you can think about large blocks of products or you can go to the individual product level. You can decide that for UPC #1, what I want to do is charge $1.45 in Chicago, store #1. In store #2 I'd like to charge $1.10. In store #3 I'd like to charge $1.20. The idea is that you can have a large range of prices. You can ask yourself ``what's going to be the optimum pricing strategy?'' (Optimal in the sense of what's maximizing my profits).
Let me move on and come to what the data looks like. For those who aren't familiar with this marketing data, specifically what's going to drive this problem is the retailer. And what is it from the retailer's standpoint? What information set does the retailer have? Obviously it would be nice to go out and know what every consumer does in the marketplace. That would allow us to think about micro-marketing policies not just at a store-level, but at an individual level. That's a slightly more complicated problem. Let's solve the store problem first.
Go to written version of paper