L11 - Testing and Experimenting Flashcards
Media Planners:
Essentially, a media planner needs answers to questions such as these:
- In which geographic markets should we place extra money to achieve our marketing objective?
- Will extra media weight generate enough additional sales to justify the cost?
- Which advertising timing schedule is best for this brand?
- How much money is optimum to achieve the brands goals?
Why Test or Experiment?
Testing or experimenting is necessary for a number of reasons:
- To help the planner make decisions
- To avoid making costly errors
- To modify the data with their own judgement
How they differ:
- A test is a simple piece of research in which one measures a variable (or treatment) introduced into the market to see what effect it has
- An experiment resemble a test in that simular markets are selected for treatments, but great care is taken to make sure that the markets are equivalent
Test or Experiment, Which is better?
- Experiments provide a higher degree of objectivity
- Experiments are better controlled
- A test is less expensive than an experiment
- Test usually take less time to conduct than experiments
Advertisers use tests almost exclusively to obtain guidance and support for their market decisions
Measure Examples:
- Weekly gross rating points by daypart commercial length
- Product distribution
- Number of grocery store facings
- Price of their own and competitors products
- Weekly GRPS of the primary competitors
- Dates and values of consumer promotions
- Weather where relevant
- Seasonal influences
Test Marketing:
Is the use of controlled tests or experiments in one or more geographic areas to gather certain kinds of information or to gain experience in marketing a brand
Test marketing is most common for new brands in existing product categories
Purpose of Test Marketing:
- To examine the viability and potential profitability of new products
- To test alternative media, creative and marketing strategies for existing brands
Nielsen Company’s BASES:
The company uses three measures:
- Product distribution as a percent of all commodity volume
- The estimated awareness of the new brand, based on the proposed media plan
- A survey of consumers intent to buy after seeing a description of the product’s concept
Research Designs Used in Test Marketing:
Research design refers to a plan of actions to be taken in the testing
Plan is carefully worked out to obtain certain kinds of information
Research designs can be very simple
- Observing a market
- Introducing an experimental variable
Control Markets:
Validity means that the test measures what it is supposed to measure
- E.g. if advertising is introduced into a test market as an experimental treatment, the planner will want to know the extent to which advertising had any effect on the outcome
Control market is one that does not receive any special treatment
- Must be selected carefully so that is shares all the demographic and economic characteristics of the test market
Randomized Block Designs:
Is one in which subjects to be measured are first grouped together into blocks
- Each subject is carefully chosen so that he or she will be much like every other subject in that block
- The selection process is done at random
- Randomized blocks are designed to prevent situations in which the outcome results from differences with the sample, rather than differences in the treatment
Latin Square Design:
A block divided into a number of cells containing coded letters and arranged so that a letter appears only one in each row
More precise than the randomized block
Media Testing Example:
Complete media plan: a national media plan might be developed for a new product that has not been marketed before. The typical objective of the marketing part of a test for such a product is to learn a brand’s sales or the market share that could be developed on a national scale.
- Alternative spending levels
- Alternative schedules
- Alternative media mixes
- Alternative commercial length or ad sizes
Testing a Campaign
Impressions:
‘Impressions’ are the total number of viewing occasions (announcements) that a campaign delivers - duplicate numbers)
The number of impressions depends on GRP and population size
Impressions = GRP/100 x number of people in the population group
For example, a campaign with 200 GRP (ie 200%) delivered to a 1m population (F:18-34) gives 2 million impressions per week
Impressions = 200/100 x 1m = 2m impressions
