Pricing Tests and Price Elasticity for one product

You can customize this template by filling in a simple form, without editing a spreadsheet.
This is a small working sample of the template.

A customized template is a flexible model that you can adapt to your situation by filling in a simple form,
without editing a spreadsheet or its formulas. For example, you can specify time range and time grain;
number and names of items in a dimension (such as your products and product families); and include or
exclude major features. The resulting spreadsheet matches your needs better than any standard template.
Get a customized version of this template on our website.

ModelSheet provides you with customized templates in three ways.

1. Order a customized version of this template.

Click "+" for more information.
Explore our customized templates.

2. If you want more customizations, retain ModelSheet Software to build them for you.
Click "+" for more information.
Learn more about consulting services.

3. Use the ModelSheet Authoring Environment to build and customize your spreadsheet models.

The ModelSheet Authoring Environment is a SaaS application for developing and maintaining business
models and delivering them in conventional spreadsheets.
Click "+" to learn more about ModelSheet technology that makes customized template possible.

We have more to tell you about ModelSheet and we'd like to hear about your needs for templates and
models.
Please visit our website at www.modelsheetsoft.com
or contact us at info@modelsheetsoft.com.

Description of the Price Elasticity model

Challenge #1: How do we set the price for a single product to maximize its revenue?

This application uses results of pricing tests to estimate impact of prices on unit sales and revenue. The
results can help you optimize revenues with limited pricing experiments.

The key to estimating sales from pricing tests is to estimate the price elasticity from test results. The
mathematics of price elasticity is described in the appendix below.

The model computes revenue as price * sales units.

Challenge #2: How do we set the price for a single product to maximize profits?

The model computes profit margin as Revenue – Costs. Costs can be cost of goods or total costs or
whatever you choose, in order to compute gross margin, operating margin or any other profit margin you
choose.

ModelSheet is a trademark of ModelSheet Software, LLC page 1 of 15

 Pricing Tests and Price Elasticity for one product

The key step in computing profit margins is to estimate costs as a function of sales levels, as described in
the Technical Notes below.

The model includes Excel charts that provide graphical views of key variables. These charts are part of the
model, and they are included by default in exported Excel workbooks. You can add more charts, import
them, and the new charts will be included in exported Excel workbooks.

As you explore the model, we suggest that you

• Read some of the Excel comments that are attached to Analysis Variables throughout the workbook.
These comments also appear in ModelSheet in convenient places.
• View worksheet "Formulas" which shows the named variables and symbolic formulas of the model
in a compact and readable form. The symbolic formulas are not active in this Excel workbook, but
they
give you some idea how the model works, and how it looks in ModelSheet.

Technical Notes

Challenge #1: How do we set the price for a single product to maximize its revenue?

The model assumes that, at a reference price (variable Reference_Price), sales units are known (variable
Reference_Sales_Units).

Assume we have a number of test markets, each of which sells a known number of units (variable
Test_Ref_Sales_Units) at the reference price. The characteristics of test markets that affect the validity of
pricing tests are described in the appendix below.

The model uses the test results (variable Test_Sales_Units) at prices (variable Test_Prices) that you
specify for each test market to estimate the price elasticity (Test_Elasticity) using the regression formula
Test_Sales_Units = Test_Ref_Sales_Units * (Test_Price/Price0) ^ Test_Elasticity

Where does this equation come from?

This equation is the exponentiation of the linear regression equation
Ln(Test_Sales_Units) = Ln(Test_Ref_Sales_Units) + Ln(Test_Price/Price0) * Test_Elasticity
This equation is the exponentiated equivalent of a linear approximation, or a first-order Taylor series.
Both of these equations come from accepting the approximation that Test_Ref_Sales_Units and
Test_Elasticity are constant over a range of prices. Their greatest weakness is that, in reality, the
elasticity may change as price changes.

Using the results of the regression, the application predicts sales units (Predicted_Sales_Units) for a range
of prices (Prediction_Prices) acording to the formula
Predicted_Sales_Units = Reference_Sales_Units * (Prediction_Price / Reference_Price) ^ Test_Elasticity

Challenge #2: How do we set the price for a single product to maximize profits?

• You provide a value for costs at the reference level of sales units, and an estimate of elasticity of
costs
with respect to sales units (variable Cost_Elasticity).

ModelSheet is a trademark of ModelSheet Software, LLC page 2 of 15

 Pricing Tests and Price Elasticity for one product
This cost model enables you to provide accurate cost information at one sales level, and to model
non-linearities in costs due to economies of scale and capacity limitations.
• The model estimates costs at each predicted level of unit sales using the formula
Predicted_Cost = Reference_Cost * (Predicted_Sales_Units / Reference_Sales_Units) ^ Cost_Elasticity

Appendix: Introduction to Price Elasticity

Price elasticity is important because it tells how price affects sales levels.

Definition of price elasticity (denoted epsilon)

Let's denote the sales units of a product by Q (for quantity) and the price by P.
Price elasticity relates changes in price to changes in unit sales, according to the formula
ΔQ/Q = epsilon ΔP/P .
Price elasticity, denoted by 'epsilon', is defined as
epsilon = ΔQ / ΔP * P / Q
This can be stated more compactly as
epsilon = Δ ln(Q) / Δ ln(P)

Use of price elasticity to predict sale units at a given price

For price P, the unit sales Q will be, for some constant K,
Q = K * P ^ epsilon

If we have a reference price P0 at which we know that unit sales are Q0, then the relationship can be
stated as
Q = Q0 * (P/P0) ^ epsilon

If prices rise by 1%, then unit sales change by epsilon * 1%, and revenues change by (1+epsilon) * 1%.

Different ranges of values for price elasticity indicate different types of reaction to price changes.
epsilon > 0: increasing prices increases unit sales. This occurs for some luxury goods with status appeal.
epsilon = 0: changing price will not affect unit sales, so revenue grows directly with price.
-1 < epsilon < 0: increasing price decreases unit sales, but increases revenues.
epsilon = -1: increasing price decreases unit sales but does not affect revenues at all.
epsilon < -1: increasing prices decreases unit sales so much that revenue declines.

Goods with price elasticities in these ranges often have the following characteristics.
epsilon > 0: luxury goods with status appeal
epsilon = 0: price-inelastic goods
epsilon ≤ 0: normal commodities
epsilon < -1: commodities that are overpriced

Using pricing tests to estimate price elasticity

If you can offer different prices in different markets, you can measure the impact of price changes on sales.

ModelSheet is a trademark of ModelSheet Software, LLC page 3 of 15

 Pricing Tests and Price Elasticity for one product
Many factors can distort the results of pricing tests.
• The different markets may have different distributions of customer behavior, so that combining them
in one
test will distort results.
• Extraneous factors may distort results unevenly across test markets, such as seasonal effects,
amount
of advertising and other promotion, strength of distribution channels.
• Price sensitivity may differ over the range of prices tested. To get accurate estimates of price
sensitivity, you
must perform separate tests over smaller price ranges.
• Customers in one market may find out about lower prices in other markets. This may cause people to
buy
in markets where you did not expect them to buy, which is likely to distort test results.
• Customers may learn that the prices are temporary. If this happens, lower prices will generate
artificially
high sales, and higher prices will generate artificially low sales, as people wait for the test to end.

If your test design minimizes the impact of these problems, you can probably perform a pricing test that
yields a good estimate of price elasticity.

Cost elasticity

We can use price elasticity and cost elasticity to estimate the price that maximizes profits.

Assume we know that, when sales are Q0 units, then costs are Cost0. (This can be cost of good or Cogs
pluas opertating expense, depending what kind of margin you want.) Assume costs as a function of units
sold are:
Cost = Cost0 * ((Sales Units) / Q0) ^ Cost_Elasticity

Then profit margin is

Margin = Revenue - Cost = Price * (Sales Units) - Costs

Challenge #1: How do we set the price for a product line with several products to maximize its revenue?

This application uses results of pricing tests to estimate impact of prices on unit sales and revenue for
several related products. The results can help you optimize revenues with limited pricing experiments.

The key to estimating sales from pricing tests is to estimate the price elasticity for each product, and price
cross-elasticities for each pair of products from test results. The mathematics of price cross-elasticity is
described in the appendix below.

The model computes revenue as sum(price[product] * sales_units[product], product in product line)

Challenge #2: How do we set the price for a product line of several products to maximize profits?

The Advanced version of the model computes profit margin as Revenue – Costs. Costs can be cost of
goods or total costs or whatever you choose, in order to compute gross margin, operating margin or any
other profit margin you choose. This model assigns costs to each product, but costs can be joint costs of
several products.

ModelSheet is a trademark of ModelSheet Software, LLC page 4 of 15

 Pricing Tests and Price Elasticity for one product
The key step in computing profit margins is to estimate costs as a function of unit sales levels for the
products, as described in the Technical Notes below.

As you explore the model, we suggest that you

• Read some of the Excel comments that are attached to Analysis Variables throughout the workbook.
These comments also appear in ModelSheet in convenient places.
• View worksheet "Formulas" which shows the named variables and symbolic formulas of the model
in a compact and readable form. The symbolic formulas are not active in this Excel workbook, but
they
give you some idea how the model works, and how it looks in ModelSheet.

This Excel workbook was generated by ModelSheet on April 22, 2010, except for this worksheet of
comments.

Copyright © 2009, 2010 ModelSheet Software, LLC

ModelSheet and the ModelSheet logo are registered trademarks of ModelSheet Software, LLC.

ModelSheet is a trademark of ModelSheet Software, LLC page 5 of 15

 Sales Units versus Price in Test Markets Normalized S
in
Sales Units in each Test Market

160 1.40

Normalized Sales Units

150 1.30
140
1.20
130
1.10
120
1.00
110
0.90
100
90 0.80

80 0.70
$80 $90 $100 $110 $120 $130 $140 0.80 0.90 1.00 1.10
Price in each Test Market Normalized Price in ea

Predicted Price Sensitivity Predicted R

100 $1
Predicted Total Sales Units

95 $1
90 $1
85 $1
80 $1
75 $
70 $
65 $
60 $
55 $
50 $
$80 $100 $120 $140 $160 $180 $200 $220 105 128 151 17
Price Price

ModelSheet is a trademark of ModelSheet Software, LLC page 6 of 15

 Normalized Sales Units versus Price
in Test Markets

1.40
Normalized Sales Units

1.30

1.20

1.10

1.00

0.90

0.80

0.70
0.80 0.90 1.00 1.10 1.20 1.30 1.40
Normalized Price in each Test Market

Predicted Revenue and Profit

$1
$1
$1
$1
Predicted Revenue
$1 Predicted Margin
$
$
$
$
$
$
105 128 151 174 197
Price

ModelSheet is a trademark of ModelSheet Software, LLC page 7 of 15

 ABC, Inc.
Widgets
Price Test with Standard Product
Price Test

Shaded cells are input cells. You can enter data in them.
Excel formulas in shaded cells are starting suggestions. You can overwrite them.
Reference Price $100.00 Test Prices Test Prices Normalized
Test 1 $105.00 Test 1 1.050
Test 2 $116.00 Test 2 1.160
Test 3 $128.00 Test 3 1.280
Total $116.33 Total 1.163

Test Ref Sales Units Test Sales Units Test Sales Units Normalized
Test 1 333.33 Test 1 Test 1 0.000
Test 2 333.33 Test 2 Test 2 0.000
Test 3 333.33 Test 3 Test 3 0.000
Total 1,000.00 Total 0.00 Total 0.000

Regression Statistics

Test Elasticity Err:502 Test StdError Elasticity #VALUE! Test R squared #VALUE!
Regression equation:
Test_Sales_Units/Test_Ref_Sales_Units = (Test_Price/Ref_Price)^Test_Elasticity

ModelSheet is a trademark of ModelSheet Software, LLC page 8 of 15

ModelSheet is a trademark of ModelSheet Software, LLC page 9 of 15
 ABC, Inc.
Widgets
Price Test with Standard Product
Sales Predictions

Shaded cells are input cells. You can enter data in them.
Excel formulas in shaded cells are starting suggestions. You can overwrite them.
Prediction Prices Predicted Sales Units Predicted Revenue
Price 1 $105.00 Price 1 Err:502 Price 1 Err:502
Price 2 $128.00 Price 2 Err:502 Price 2 Err:502
Price 3 $151.00 Price 3 Err:502 Price 3 Err:502
Price 4 $174.00 Price 4 Err:502 Price 4 Err:502
Price 5 $197.00 Price 5 Err:502 Price 5 Err:502
Sales Units = Reference_Sales_Units * (Prediction_Price/Reference_Price)^Units_Elasticity

Predicted Costs Predicted Margin Predicted Margin %

Price 1 Err:502 Price 1 Err:502 Price 1 Err:502
Price 2 Err:502 Price 2 Err:502 Price 2 Err:502
Price 3 Err:502 Price 3 Err:502 Price 3 Err:502
Price 4 Err:502 Price 4 Err:502 Price 4 Err:502
Price 5 Err:502 Price 5 Err:502 Price 5 Err:502
Cost = Reference_Cost *(Predicted_Sales_Units/Reference_Sales_Units)^Cost_Elasticity

Cost Parameters

Reference Sales Units 1,000.00 Reference Cost $56,000.00 Cost Elasticity 0.45

ModelSheet is a trademark of ModelSheet Software, LLC page 10 of 15

ModelSheet is a trademark of ModelSheet Software, LLC page 11 of 15
 ABC, Inc.
Widgets
Price Test with Standard Product
Formulas

Variable Display As Dimension Index

Company_Name Company Name

Cost_Elasticity Cost Elasticity

Predicted_Costs Predicted Costs Predictions Data:

Predicted_Margin Predicted Margin Predictions Data:

Predicted_Margin_pct Predicted Margin % Predictions Data:

Predicted_Revenue Predicted Revenue Predictions Data:

Predicted_Sales_Units Predicted Sales Units Predictions Data:

Prediction_Prices Prediction Prices Predictions Data:

Product_Name Product

Reference_Cost Reference Cost

Reference_Price Reference Price

Reference_Sales_Units Reference Sales Units

Test_Elasticity Test Elasticity Global Data:

Test_Ln_Prices Test Ln Prices Tests Data:

Test_Ln_Sales_Units Test Ln Sales Units Tests Data:

Test_Name Name of Test

Test_Prices Test Prices Tests Data:

Test_Prices_Normed Test Prices Normalized Tests Data:

Test_R_squared Test R squared Global Data:

Test_Ref_Sales_Units Test Ref Sales Units

Test_Sales_Units Test Sales Units

Test_Sales_Units_Normed Test Sales Units Normalized Tests Data:

Test_StdError_Elasticity Test StdError Elasticity Global Data:

ModelSheet is a trademark of ModelSheet Software, LLC page 12 of 15

 Formula / Data

Reference_Cost*(Predicted_Sales_Units/Reference_Sales_Units)^Cost_Elasticity

Predicted_Sales_Units*Prediction_Prices-Predicted_Costs

Predicted_Margin/Predicted_Revenue

Prediction_Prices*Predicted_Sales_Units

Reference_Sales_Units*(Prediction_Prices/Reference_Price)^Test_Elasticity

minr(ranged("Tests", Test_Prices))+(dimitemnum("Predictions")-1)*(maxr(ranged("Tests", Test_Prices))-minr(ranged("Tests", Test_Prices)))

linest(1, 1, false, ranged("Tests", Test_Ln_Sales_Units), ranged("Tests", Test_Ln_Prices))

ln(Test_Prices_Normed)

Ln(Test_Sales_Units_Normed)

round(1.1*prevde(1.1^(-0.5*length(ranged("Tests", Test_PricesX)))*Reference_Price, "Tests"), 0)

Test_Prices/Reference_Price

linest(3, 0, false, ranged("Tests", Test_Ln_Sales_Units), ranged("Tests", Test_Ln_Prices))

Test_Sales_Units/Test_Ref_Sales_Units

linest(2, 1, false, ranged("Tests", Test_Ln_Sales_Units), ranged("Tests", Test_Ln_Prices))

ModelSheet is a trademark of ModelSheet Software, LLC page 13 of 15

 ABC, Inc.
Widgets
Price Test with Standard Product
Labels

Variable Display Label

Company_Name Company Name
Cost_Elasticity Cost Elasticity

Predicted_Costs Predicted Costs

Predicted_Margin Predicted Margin

Predicted_Margin_pct Predicted Margin %

Predicted_Revenue Predicted Revenue

Predicted_Sales_Units Predicted Sales Units
Prediction_Prices Prediction Prices
Product_Name Product
Reference_Cost Reference Cost

Reference_Price Reference Price

Reference_Sales_Units Reference Sales Units
Test_Elasticity Test Elasticity

Test_Ln_Prices Test Ln Prices

Test_Ln_Sales_Units Test Ln Sales Units
Test_Name Name of Test
Test_Prices Test Prices
Test_Prices_Normed Test Prices Normalized

Test_R_squared Test R squared

Test_Ref_Sales_Units Test Ref Sales Units

Test_Sales_Units Test Sales Units

Test_Sales_Units_Normed Test Sales Units Normalized

Test_StdError_Elasticity Test StdError Elasticity

Dimension (item) Display Item As Total As Level As

Predictions Predictions Total Predictions
Price_1 Price 1 Predictions
Price_2 Price 2
Price_3 Price 3
Price_4 Price 4
Price_5 Price 5

Tests Tests Total Tests

Test_1 Test 1 Tests
Test_2 Test 2
Test_3 Test 3

ModelSheet is a trademark of ModelSheet Software, LLC page 14 of 15

 Comment
Name of the company or organization doing the pricing test
The elasticity of cost with respect to changes in units sold; that is, the price
sensitivity of costs to unit sales. The cost used in the model can be cost of goods
or operating cost, depending on what kind of profit margin you want to measure.
The predicted costs for each level of unit sales predicted by the price elasticity
model and the cost elasticity model
The predicted profit margin (e.g. contribution margin or operating margin) at each
price level using the predictions for sales units and costs
The predicted profit margin percentage, defined as (profit margin) / revenue, for
each prediction price level
The predicted revenue at each price level using the predictions for sales units
The predicted sales units at each price level
The prices at which sales units of the product are predicted

The cost incurred when sales units equal the reference amount
(Reference_Sales_Units)
The reference price at which we know that sales units equal
Reference_Sales_Units
The number of units sold when the price is the reference price (Reference_Price)
The elasticity of sales units with respect to changes in price, estimated from the
pricing test
Ln(Test_Prices / Reference_Price), for use in the regression equation
Ln(Test_Sales_Units / Reference_Sales_Units), for use in the regression
equation
A name that uniquely identifies this pricing test
The prices used in the various pricing tests
The ratio (test price)/(reference price) for each test market. The log of this variable
is used in the regression equation to compute elasticity.
The coefficient of determination (commonly known as r-squared) for the
regression of log units versus log prices in test markets
The reference sales units that can be sold in each test market when the price is
the reference price (Reference_Price)
The sales units measured at the test price in each test market in the pricing test
The ratio (test sales units)/(test reference sales units) for each test market. The
log of this variable is used in the regression equation to compute elasticity.
The standard error of the estimate of the elasticity from the market test

Comment
A list of the price levels at which sales units are predicted from the market test
data

A list of the test markets where the pricing test is performed

ModelSheet is a trademark of ModelSheet Software, LLC page 15 of 15