Dave's Adventures in Business Intelligence

This post might not seem like it’s related to Business Intelligence at first, but stay with me to the end and I promise I will tie it all together. 🙂

Have you ever purchased a product that came with a coupon that you had to mail in? The idea is that rather than drop the price on a product a company will offer a rebate or other incentive to encourage you to spend your money. There is cost (overhead) involved in processing the rebate, so why don’t they just discount the product to begin with? A few months ago I got a new Blackberry phone (love love LOVE it, by the way). It was $200 but came with a $100 rebate. Nice deal, right? Not so fast. It seems that the rebate wasn’t quite cash in my pocket. It was a $100 “credit” card that I could only use at certain stores. Best Buy couldn’t figure out how to use it, but Walmart could. I eventually got to spend the entire $100 but it took extra effort. And that’s the issue here, the fact that consumers need to make extra effort on rebates.

Several years ago I worked for a company (as a client) that processes rebate requests. (This is one of the reasons I enjoy my job; I get to see all sorts of companies in different industries and learn how they work.) What I learned at this job was that all of those different companies that offer rebates for purchases don’t process the money themselves. Instead they hire one of the few companies that specializes in this area. Here is how that process works, at a very high level.

First, the product company determines the product(s) involved. Then they will talk to the rebate processor to get input on how much the rebate should be. Based on that meeting they will set up the rebate program, which the rebate company then gets paid to manage. Why would they do that? Why would company “A” ask company “B” how to price their product? That’s where the BI comes in.

The dirty secret of rebates is that a large percentage of them go unclaimed. Depending on who you read (I have included some links at the end of this post) the percentage of rebates that get processed ranges from 35-40% down to as low as 10%. Product companies don’t know this, but the rebate processing companies do know. Afterall, they have all of the data. 🙂 Here’s how they might use it.

Initial Research
Acme Widgets has a new super widget coming out but they have lots of old stock of the existing widgets. They want to accellerate sales of the old stock before the release of the new widget. One of the marketing team suggests offering a rebate. The question is, how is the rebate priced? They hire Rebates, Incorporated (RI) to perform this analysis.

Kathy – an analyst for RI – gets right to work. First she will do some basic investigation about Acme Widgets, their products, and their competitors. She finds out that the suggested retail price for the widget is $150, but they’re often on sale for $110. The average retail price paid over the past twelve months is $130. Break-even pricing on the widget is $85. Pricing for competitor products is roughly similar and therefore she decides she doesn’t need to incorporate that into her analysis.

Next Kathy goes into the archive of promotions manged by RI over the past 18 months. During that time span they have run rebates for other companies with similar pricing (but not necessarily similar products). They have also run promotions for similar products but at different price points. Kathy’s goal is to crunch the numbers and determine what the average redemption rate is for anything in the $110-$150 price range, as well as the repemption rate for products in the same or similar categories. Both factors come into play, as consumers are more likely to fill out a $50 rebate than a $5 rebate. Comsumers of certain types of products are also statistically more likely to redeem the rebate. See… I told you that BI would come into play eventually. 🙂

There is a fixed cost of $25,000 to configure the promotion, and RI charges a per-redeption amount to process each claim that is filed. The variable charge is based on the amount of the rebate and the number of components that have to be filed by the consumer. Please note that other than statistics mentioned in the early paragraphs, all numbers in this blog post are completely made up by yours truly.

Ultimately Kathy’s job is to create one or more proposals for Acme to consider. Her proposal might be presented in the form of an interactive Powerpoint presentation with embedded Xcelsius components. By using data from similar past promotions she can come up with formulas or at least enough raw data to support her proposals. The cost to Acme includes the fixed cost to set up the promotion, the variable costs to process each claim, and ultimately the rebate cost paid out to the consumers that bother to send in their claims.

RI Proposal
In this fictional exercise, Kathy designs several proposals based on the results of her analysis. If Acme offers a $20 rebate (on a $150 product) she predicts Acme can expect sales to go up 10% and a redemption rate of 10%. If Acme offers a $30 rebate her prediction calls for sales to go up 28% and a redeption rate of 18%. A rebate of $50 (one third of the suggested retail price) could cause sales to jump 50% but at that level the redemption rate also jumps to 30%. So what should Acme do?

I will continue this in Part II, coming in a few days. 😎

Related Links

• {dead link removed}
• Billions in rebates go unclaimed… (Marketwatch)
• PC World article on rebate process