Simulated Test Markets Market Intelligence Julie Edell Britton
Simulated Test Markets Market Intelligence Julie Edell Britton Session 10 October 10, 2009
Today’s Agenda • Simulated Test Markets • Contadina Case • Course Wrap-up
Why do an STM? • Get trial rate, repeat rate, & purchase amounts • Fast and Cheap • about $100 K for full STM with projections for 3 alternative marketing plans • Secret from Competitors; no sabotage • Accurate -- when assumptions implemented • Over 1000 validations of BASES forecasting accuracy; average forecast falls within 9% of actual sales; 91% fall within 20% of actual sales. • For high risk, follow with true test market
Two Kinds of Simulated Test Markets • Concept Test Simulated Test Market (BASES). Cheaper. • Trial Rate: Consumers exposed to 2 -paragraph concept, not final ads. Adjusted “Top Box”. • True Simulated Test Market (e. g. , ASSESSOR, LITMUS) • Trial Rate: Consumers come to laboratory store. Exposed to half hour sitcom with finished ads for new product, competition. Then led to mini-store where new product, competitors available for purchase w/ own $. • Concept Tests winning because accuracy not that different and trends have made it more difficult to recruit representative samples at mall-based laboratory stores. BASES now uses Web panels.
Simulated Test Markets • Universe Size (Households) • %Trial Rate (from BASES) • %Adjusted Trial = % Aware * % ACV * % Trial Rate • Trial Units = Trial HHs * Trial Ave. # Units • = (%Adj Trial * Universe) * Trial Av. # Units • • % Repeat (from BASES) • Repeat Units/Yr = Trial HH * %Repeat * • Av. Units/Purchase * # Repeat Purchases/Yr
Contadina Case
Contadina Takeaways • Stages of product development • Moving parts of simulated test market • Simulated test markets good at forecasting demand if inputs right • Competition must be modeled as in Clancy & Schulman • When product usage and risk are key, an actual test market might be necessary
Market Intelligence: 3 Skills Backward Market Research Getting data & judging its quality Analytic methods for classic marketing decision problems
Backward Market Research • Determine how research will be implemented (what will final report look like) • Figure out what info is needed and what analysis will support recommendations • Design the study to gather the data -- if info doesn’t already exist • Crunch data & make recommendation
Getting Data • Process • Identify management decision, alternatives, information needs • Does secondary data exist? If not, do a study. • Exploratory research for hypotheses • Conclusive Research • Descriptive. Do a survey. • Causal. Do an experiment.
Judging Data Quality • • • All data has error –systematic or random Systematic error is bias Direction of the bias? Implication of this direction for decision? Tools for judging data quality • For secondary: Ad Age & MBA value • For surveys: Measure reliability, validity; sampling • For experiments: Threats to internal validity For forecasts: Reproduce current market shares
Statistical Analysis Tools • Crosstabs and Chi Square • Zero order v. partial effects • Experiments & factorial designs using Analysis of Variance • Conjoint Analysis • Multiple regression
Marketing Analysis Tools • Competition: Multiattribute attitude model, perceptual mapping, conjoint, cannibalization analyses • Segmentation: Index numbers, Interaction, a priori v. clustering • New product concept screening: Conjoint analysis, BASES • Pricing: Conjoint analysis and Regression • Promotion effects: Sales Promotion analysis, Multiple regression • New Product Forecasting: perceptual mapping, conjoint, BASES
Day 1: Purpose & Approach • Users of market intelligence in consulting, marketing management, entrepreneurship, finance – Drowning in data – Faced with decisions with imperfect information • Gain practice as evaluator of research – before and after it is undertaken • Act as a provider to become a sophisticated user – soft and hard skills
Day 10: Technical Skills Learned • Research Process: Exploratory to Conclusive (Conjoint, Contadina). • • • Disaggregate forecasting (Conjoint) Pricing (Conjoint) Product Design (Conjoint) Segmentation (index #s, interaction, clustering) Quantitative tools for marketing decision making: • Perceptual maps for positioning, regression & promotion (Doritos), BASES (Contadina)
Thank You and Best Wishes!
- Slides: 16