Optimizing Street Price with Demand Curves
by Jonathan Bein, Ph.D.
In this issue of Price Matters, we look at a case study of price optimization for a $100M software vendor. This vendor processes several hundred thousand transactions annually with consumer and business customers. Z2M4 used pocket price analysis and price elasticity analysis to enable this client to gain incremental revenue of more than 2%. Pocket price1 reflects the true price, the net price after all on-invoice and off-invoice discounts and promotions have been applied to a transaction. The gap between list price and pocket price is frequently large and surprising. Price elasticity analysis identifiesthe price at which revenue will be maximized for a given demand curve.
Situation
This vendor has a multi-tier product line with a wide price range sold through direct and indirect channels globally. The vendor's initial question was whether the list price was appropriate. Z2M4 analyzed 16 different scenarios based on 4 products, 2 channels, and 2 geographic regions. This inspection revealed a broad and arbitrary pattern of discounts. Approach
- The approach used for this client is as follows:
- Perform pocket price analysis for each scenario.
- Perform elasticity analysis for each scenario.
- Determine target pocket price for each scenario.
- Recommend a set of actions to achieve target pocket price.
The elasticity analysis quantifies a theoretical optimum whereas the average pocket price quantifies what is really happening in the field. The primary insight in our approach is to iteratively evolve the pocket price toward the theoretical optimum whether the pocket price is above or below the optimum price. Initially, the vendor believed that pocket price analysis would not be insightful.
Pocket Price Analysis
The pocket price analysis for one of the 16 scenarios is shown at the right. It plots the percentage of list price paid in a transaction versus the total amount from the account in a year. The red line depicts how the price should vary with size, namely that the discount on a transaction should be proportional to the size of the account. However, as the graph shows, there is a stark pattern of small clients receiving significant discounts and larger clients receiving smaller discounts.
1 Pocket Price was defined by Michael Marn at McKinsey in 1992.
One of the channels showed a similar trend in all of its eight scenarios. The other channel exhibited much better discipline in its discounting practices. Upon further inspection, it turned out that the undisciplined channel was routinely competing on price to win deals. Because software products have effectively 0% cost-of-goods-sold, there was a perception on the part of the sales personnel that the company will still make money even with a large discount. Of course, this practice trains customers to expect (large) discounts and ultimately erodes the perceived value of the product.
Elasticity Analysis and Target Price Determination
The elasticity analysis for each scenario was based on actual transactions in a one year period. Because the discounting practices were fairly arbitrary the data was noisy. As a result the linear regression was supplemented with expert judgment to derive the demand curve for each scenario. In the graph shown below, there is a demand curve (the straight line in red) and a total revenue curve (the curved line in green.)
In this scenario, the price that optimizes total revenue is $629. However, the pocket price is $525, significantly less than the list price and the optimal price. As mentioned before, the approach is to set a pocket price target that is closer to the optimal price, in this case $553, a 4% increase. The net effect for this scenario was to increase revenue 2.6% from $8.485M to $8.713M, while decreasing units sold from 16,000 to 15,600.
Naturally, it is tempting to set the target price equal to the optimal price. But remember, the optimal price involves the subjective element of expert judgment, so an incremental and iterative approach is the safest move.
In another scenario, the pocket price, $181, was above the optimal price, $165. Once again, the incremental approach is to move toward but not all the way to the optimal price, so the pocket price target was set at $169. The net effect for this scenario was to increase revenue 1.1% from $4.351M to $4.396M, while increasing units sold from 24,000 to 26,000.
Recommendations
Once the target pocket price is set, there is still work to realize the price optimization. We spent a good deal of time determining the moves that will allow the vendor to increase the pocket price without losing revenue. Besides a stronger enforcement of discount policy, here are some of the recommended actions:
- Perform buyer power analysis (described in Price Matters 6-17-2009) to quantify account price sensitivity and bargaining leverage. The buyer power analysis allows sales people to give more tailored discounts.
- Calculate commission on pocket price attained as well as gross revenue. The traditional revenue based incentives encourage the wrong behavior on the part of sales people. Including pocket price in the incentive aligns sales people with corporate objectives.
- Create an economic value model/TCO model for SMB and Enterprise products. The economic value model is critical in justifying prices for corporate customers. Otherwise, value is completely subjective.
Results
Some of the recommendations have been implemented and others are underway. The benefit realized so far is a 2% increase in average selling price. As the other recommendations are implemented, the total benefit will likely increase further. For the vendor it was an interesting realization that street price/pocket price matters even more than list price in a B2B environment. An adjustment to list price is a mean to attain a higher street price.
You can also download this article as a PDF.