Research Decisions and the Value of Marketing Information The meaning of “limited by budget and time constraints.”

Research Decisions and the Value of Marketing Information

The meaning of “limited by budget and time constraints.”

Value of Marketing Research

• Estimating gains from the “right” marketing decisions, developing market forecasts.

• Framing marketing decisions and the value of information.

Overlapping Information Needs

• Planning: Market Opportunity Analysis

• Implementation: Refining Marketing Actions

• Control: Monitoring

Forecasts vs. Potential

• Market Potential: Sales of all competitors, if all gave maximal marketing efforts.

• Market Forecast: Sales of all competitors if competitors gave historic effort.

• Sales Potential: Sales of your firm, with maximal effort, maximal market share ever.

• Sales Forecast: Sales of a firm given projected efforts.

Forecasts vs. Potential

• “Market Potential” is a frequently used expression, but is an “upper limit” of consumer expenditure with maximum marketing efforts.

• “Market Forecast” is a more realistic projection of consumer expenditure with current competitive efforts.

Women in the U.S.

35 to 39 years 11,387,968

40 to 44 years 11,312,761

45 to 49 years 10,202,898

50 to 54 years 8,977,824

55 to 59 years 6,960,508

Conversion Factors

• Market forecasts convert population and household counts for geographic areas into estimates of:– Number of buyers

– Sales for a particular product category, or consumer expenditures

• Sales forecasts convert market forecasts into share of the market or sales for a particular brand

Placing a Value on a Sales Forecast

• Contribution Margin

• Discounted Time Stream

Market research should be conducted only when the expected value of information to be obtained exceeds the costs of obtaining it.

What then, is the value of information?

Value of perfect information:

Expected payoffs under uncertainty

• Choose between 40% probability of scoring two points, versus

• 98% probability of scoring one point.

• Expected payoffs are .80 versus .98, respectively.

.98 “kick”

.40 “go for two”

1 point

2 points

“kick”

“go for two”

.98 point

.80 point

Decision Tree

98 “kick”

.80 “go for two”

1 point

2 points

“kick”

“go for two”

.98 point

1.6 points

Decision Tree with research

Framing research costs:Expected value of perfect information equals:

• The value of information under certainty (gained from market research) ...

• minus the value of information under uncertainty (operating without market research, trial and error,“learn by doing.”)

1.0 Success

.

$4 millionIntroduce “A”

Do notintroduce

Case A1

$0

Value of information under uncertainty—how much would you pay for marketing research if this were your situation?

.60 Success

.40 “Failure”

$4 million

$1 million

Introduce “A”

Do notintroduce

Case A2

$0

Value of perfect information: How much would you pay for certainty that the product will be successful?

Success

Failure

$2.4 million(.6 x $4m)

$.4 million(.4 x $1m)

Introduce A

Do notintroduce

Case A2

$0

Value of perfect information: Nothing, we would introduce the product regardless of marketing research. On average, the firm would make $2.8 million in gross margins.

+

.60 Success

.40 Failure

$4 million

-$2.5 million

Introduce B

Do notintroduce

Case B

$0

Value of perfect information: How much would you pay for certainty that the product will be successful?

Success

Failure

.6*4=2.4m

.4*(-2.5)= (-$1m)

Introduce B

Do notintroduce

Case B

$0

One year’s perspective: Company would still choose to introduce because on average, on average net would be +1.4m, but information could prevent an average loss of $1m.

Expected value of perfect information:

• The value of information under certainty, (discovered through research), prevents a $2.5 million loss 40% of the time ($1m),

• minus the value of information under uncertainty, worth $0, 60% of the time.

• Expected value of “perfect” information is $1m, or an upper limit for expenditures on research.

Success

Failure

.9*4=3.6m

.1*(-2.5)= (-$.25m)

Introduce B

Do notintroduce

Case B

$0

Across years: Research reduced probability of failure from .4 to .1, providing an average return of $3.35m, a gain of 1.95m over $1.40m.

• As probabilities and costs of failures increase, the expected value of information increases.

• Costs come from accounting records.

• Probabilities come from past experience.

.20 Success

.80 Failure

$0.8 million(.2*4.0 =$.8 m)

-$2 million(.8*-2.5 = -$2m)

Introduce C

Do notintroduce

Case C

$0

The product would not be introduced (net -$1.2.m) without research, prevents a $2.5m loss 80% of the time, or $2m on average.

More realistic scenario

More simply put…

• The dollar value of any given research project depends on the amount of money riding on the decision

• If a decision has already been made and research will not affect it, research has zero value (“window-dressing”).

• The value of research depends on its ability to provide clear direction—formulating the research problem correctly.

“The Company's expenditures for research and development were approximately $121.9 million in 1998, $106.1 million in 1997, and $84.3 million in 1996.” (From Kellogg’s 10-K. Sales and operating profits were $6,762.1 and $895.6 million, respectively.)

Kellogg’s

• Research and development expense (millions): $110.2, $118.4, $104.1 (2001, 2000, 1999, respectively)

• Sales $8,853.3 $6,954.7 $6,984• % 1.5% 1.7% 1.5%• Operating

profit $1,167.9 $989.8 $828.8 • % 9.4% 11.9% 12.5%