Price Optimization Software is changing the face of retailing - For the big guys

@# Intro = If you raise the price for something, you may make more money if you sell it, but fewer people will chose to buy it. So how do you find the magic price at which the maximum number of people will still chose to buy it?
And the opposite is true: if you want to promote something, what is the least you need to drop the price to make more people chose to buy it?
The technical term for this question is ‘price elasticity’. If the price is elastic (it stretches) people will still buy the item as its price goes up, for instance. And, at this time, the biggest players in the retailing industry appear to be rushing to buy a new breed of software-called price optimization systems - that lets them determine the price elasticity of every item in a supermarket or apparel store. The results are often several percentage points of improvement in gross margin, which is a dramatic difference in the small-margin field of retailing. Its rumored that Wall Street analysts will upgrade their evaluation of a retailer if they see it acquiring price optimization software.

Volume, promo’s, margin
The problem is that the price of the optimization software itself is, so far, not very elastic. A full implementation usually costs more than a million dollars, and the software is currently considered too costly for retailers with yearly gross sales of less than $100 million.
Of course, there was a time when you had to be that big to afford a computer, and today computers are often cheaper than the desks they sit on. So it seems inevitable that price optimization technology will sooner or later become affordable for computer dealers. Basically, there are three kinds of optimization systems:

Category pricing, which involves setting everyday base prices, or setting prices for new items. This is especially interesting for grocery stores.

Promotional pricing, to drive sales of selected items with temporary price reductions or coupons. This is interesting for high-end retailers who do a lot of promotions.

Markdown or clearance pricing, for seasonal items or anything that is otherwise being discontinued. Such software is especially interesting for the fashion industry, since styles are often seasonal.
In all three cases, the software is typically run by the vendor, and the retailer sends the vendor as many as two years of point-of-sale data. That can amount to trillions of bytes. Before anything is done, teams of statisticians may have to ‘clean’ the data. The data is then subjected to sophisticated mathematical modeling.

Drawbacks
The retailers can specify goals for the pricing, such whether they want to achieve maximum profit, or volume, or market share. They can add additional business rules, such as how long an item should be on the shelf before a price change is attempted, or the minimum amount for price changes.
The data is usually updated weekly, and the results are accessed with a Web interface. The software will usually let the retailer play what-if games, to gauge the impact of alternate price changes.
‘The return on investment is clear-usually in 12 to 18 months-and the analytics are strong, so I think that we are going to see more firms offer packages for tier 2 and tier 3 retailers,’ predicted Jerry Sheldon, an expert in the field and vice president of the IHL Consulting Group. ‘There are two drawbacks,’ he added. ‘One is emotional - the existing pricing analysts will partially be replaced by software. The other is that you have to have the labor force available to go into the stores and change the prices. And if the customers ask about the changing prices, you need to educate the staff to tell the customers the truth.
’As for the benefits of using the systems, a conservative estimate is that it generates a one-half to one percent improvement in gross margins,’ said Hung LeHong, research director at research firm GartnerG2. ‘That is the business case that I would propose. On the higher side I have seen better results (than one percent) but I have also seen pilot projects where the gross margin actually went down. Perhaps a major competitor launched a promotion that was not reflected in the historical sales data. You still cannot predict outside circumstances.’
With the costs currently being so high for the software, it is not surprising that the market penetration is still low. ‘We know of 65 retailers who have adopted the technology, and 85 percent of them have sales in excess of a billion dollars. Of the top 500 retailers, I would estimate that less than 15 percent are using it.’ However, one of the vendors said their business has been doubling every year.
As for who the leading optimization vendors are, Sheldon identified the leading ones, currently, as: ProfitLogic, DemandTec, i2 Technologies, KhiMetrics.
And doubtless others will begin appearing. The trick is to have saved your point-of-sale or transaction data. Almost certainly, an affordable tool will someday come along that will let you probe that data.

Lamont Wood