The process of defining a problem or opportunity, then systematically collecting and analyzing information to aid in decision making.

Define, collect data, analyze, make decision

ie. problem price of gas up/opportunity gas alternatives

Reduce uncertainty - gain information

Hard to do well (lots of bad data)

Bad research worse than no research (bad research people are mislead, no research proceed with caution)

ie. told pool is 10 ft deep, dive in/actually 2 ft deep w/o information would be cautious. Led to believe people will like a product, mass produce, loose money.

If the expected benefit exceeds the expected cost.

cost is easier to calculate than benefits.

Define problem

Set research objectives

Develop research plan (constraints of time and money and data collection method-surveys)

Collect data

Analyze data

Make decision

Very hard to do well due to poor questions.

Non-probability samples use people because they are convenient have a known non-zero chance fo being selected.

ie. (mall intercepts) not generalizable

Each sample has a known and equal chance of being selected.

Frequently used in research; can generalize

ie. deck of cards always 1:52 chance of picking a card. 11 red out of population = 55% red, 45% black + - error factor 6 or 7

to generalize from a probability sample to the population it was taken from.

ie. get the views of many people from sampling only a few using statistics to tell us what a sample reveals to us about a population.

Mail survey-takes long time for returns, low return rates, can be annonymous and allows answers at convenience.

Telephone interview-quick, but needs to be short, lot of not at home, or refusals, sensitive issues are a problem.

Personal interviews-expensive can probe, explain questions, but interviewer bias and sensitive issues.

On-line-cheap, can be quick, answers at convenience, automatic tally, not anonymous, low response, self selection bias (only interested answer), sensitive issues about income.

Leading and loaded questions lead you to an answer or use loaded words or phrases.

ie. does senator smith's communist background make you support unlikely? or Isn't the new safety feature a marvelous idea?

Implied alternatives-How long has it been since you stopped caring about doing a good job at work?

Double barreled-The staff is friendly and efficient (two in one)

Ambiguous-How often do you go out?

Categories-Mutually exclusive (overlapping categories 1-5 and 5-10) and exhaustive (no place for all categories 24 and up, 12 and below)

Social Networks-data mining-look what people are saying

Panels data over time, changes in response to environment - Experiments manipulate ads, see responses.

Information Technology and data mining-use technology to look at data for patterns to see what people buy together and then place them together on the shelves.

Projects future needs, used for hiring, buying machines, etc (basis for all plans)

Hard to do-you will probably be wrong (use an optimistic, likely, and a pessimistic to be prepared). Hard to predict future.

Too high, excess hiring, too much equipment, projects are not cost effective, excess inventory, computer chips are quickly outdated.

Too low, not enough people, inventory shortage-bad

ie. Apple lost lifetime customers-switched to competitor, beanie babies-no substitutes, will not switch.

Survey of buyer intentions-ask buyers (good in volatile times, but do they know) most buyers are over-optimistic.

Salesforce survey-ask sales people deal with customers (may be good in volatile times but over-optimistic by nature and bias because sales people think they may be used for quotas.

Ask executives-they are in charge and may have some insight but common bias get news/views from same sources.

Direct forecast-one step

Lost Horse forecast-start with last known value, add or subtract based on positive/negative market and market factors.

Trend extrapolation-past sales predict future sales (good if clear trend, but not in volatile times) ie. linear trend between sales and year and trend line (/)

Regression-what factors/variables that affect sales. Find past relationships between sales and ie. price (use statistics to predict future)-good because use other factors, not just past sales.

Need to gather information to make marketing decision-4P's.

Need to gather right people, ask right questions, or look at right secondary data.