With our current recession, we are seeing online retailers offer various discounts on their merchandise. A question every online marketer ponders is what offer or discount will get the best response. At what price point will they get the attention from their customers and site visitors that will result in a purchase. At the same time online marketers have to watch the gross margin and get the ROI required.
Understanding the sensitivity between price and demand is critical to all marketers. We want to find the smallest change in price (ie, smallest discount) that leads to the biggest change in quantity demanded, resulting in the optimal increase in sales. In addition to doing an A/B split test to see if one discount works better than another, it is important to measure the actual sensitivity. If the price sensitivity is known, then more effective online promotional campaigns can be designed to get the desired sales result.
In math and economics measuring this sensitivity is called the price elasticity of demand, which is characterized by the symbol “η”. And there is a reasonably simple formula for calculating that.
η = price elasticity of demand
η = - (% change in orders) / (% change in price)
η = - [Δ Q / (Q1 + Q2) /2] ÷ [Δ P / (P1 + P2) /2]
If η is > 1, then you have elastic demand, and a change in price can make a big difference in demand, and total sales made. But if η is < 1, than you have inelastic demand, and a change in price is not going to affect demand. You’ll just sell the same quantity but make less revenue because you gave a discount.
Let’s take an example of an online apparel retailer. They sell a wide variety of women’s clothing. They want to promote their spring trench coats and their new line of polo shirts. Demand is lagging and they are contemplating doing a sales promotion to sell their spring inventory. They would like to test different discounts with their email list first to determine how effective the discount would be on their website and in their search campaigns. They run the following A/B split tests in their weekly email campaign and get these results:
| Spring Trench Coats |
| Discount |
Price |
Quantity Sold |
Total Sales |
| Regular Retail |
$ 150.00 |
20 |
$3,000.00 |
| With 17% discount |
$ 125.00 |
36 |
$4,500.00 |
Plugging in the numbers to the formula for price elasticity we get:
η = - [(20-36) / (20+36)/2] ÷ [($150-$125) / ($150 + $125) /2]
η = 3.14
Every 1 % discount represents 3.14% in increased orders.
| Basic Polo Shirt |
| Discount |
Price |
Quantity Sold |
Total Sales |
| Regular Retail |
$ 25.00 |
45 |
$1,125.00 |
| With 20% discount |
$ 20.00 |
48 |
$ 960.00 |
Plugging in the numbers to the formula for price elasticity we get:
η = - [(45-48) / (45+48)/2] ÷ [($25-$20) / ($25 + $20) /2]
η = .29
Every 1 % discount represents .29 % in increased orders.
While they could just look at the sales results and see that the spring trench coat sale of $25 off (or a 17%) discount was effective, by measuring the price elasticity of demand for the trench coats they had better statistical data to support publishing the sale through search and on their website. Trench coats had a very elastic demand and it was worth adding the $25 off promotion to the website and to the search campaigns.
The basic polo shirt sold more orders but was it effective? By measuring the price elasticity it became very clear that the polo shirts had an inelastic demand. They actually made less revenue. It was now clear that it was not going to be effective to run a discount promotion on the website or in the search campaigns for the polo shirts. It actually would hurt total revenue.
Next week we’ll use a bit more math to help you calculate what a statistically valid sample size with an acceptable sampling error should be. This will help you determine how large of a marketing test do you have to conduct before you can have a high degree of confidence that full campaign will get you the results you are forecasting.
Sharing how to use math in marketing, Cindy.
Tags: E-Mail, Math in Marketing, SEO // Add Comment »
