Marketing Research Data Sources and Analytics

Flashcards covering primary vs. secondary data, qualitative vs. quantitative research, synthetic respondent bias, POS scanner data, brand-switching matrices, promotion dynamics, and marketing research data vendors.

What is the core definition of primary data?

Original data collected by the user themselves for the specific question at hand.

What is the core definition of secondary data?

Data that was already collected for a different purpose than the specific question at hand.

What key performance penalty was identified in the study of late-night tweeting among $112$ NBA players?

Players who tweeted late at night (between $11:00\\text{ PM}$ and $7:00\\text{ AM}$) experienced a $1.7$ percentage point decrease in shooting accuracy the next day.

How do primary data and secondary data compare across cost, time, depth of insights, and actionability?

Primary data is high in cost, slow in time, high in depth of insights, and high in actionability. Secondary data is low in cost, fast in time, low in depth of insights, and lower in actionability.

What are the main distinguishing features of qualitative versus quantitative research?

Qualitative research uses unstructured, non-numeric data analyzed via interpretation and theme finding on small samples for rich understanding. Quantitative research uses structured, numeric data analyzed via statistics on larger samples to recommend a final course of action.

What are three main limitations of using focus groups in marketing research?

Focus groups rely on small, non-random samples (cannot estimate market share), are subject to group dynamics (dominant voices, conformity, social desirability), and are influenced by moderator effects (wording and probing shape answers).

What five ways do synthetic respondents systematically distort human behavior?

Synthetic respondents distort behavior through: $1.$ Insufficient individuation, $2.$ Stereotyping, $3.$ Representation bias, $4.$ Ideological bias, and $5.$ Hyper-rationality.

In gap analysis, what is the difference between an Attitudes vs. Intention gap and an Intentions vs. Behavior gap?

An Attitudes vs. Intention gap occurs when a consumer likes a product but does not intend to buy it (e.g., too expensive). An Intentions vs. Behavior gap occurs when a consumer intends to buy a product but fails to do so (e.g., out of stock or impulse buying.

What are the three primary types of marketing research based on problem ambiguity?

Exploratory research (used for ambiguous problems), Descriptive research (used when aware of the problem), and Causal research (used when the problem is clearly defined).

What is the difference between active and passive data collection in descriptive research?

Active data collection requires participants to actively answer questions or complete self-reports/surveys. Passive data collection automatically records behavior during daily life without participant effort (e.g., Point of Sale scanner data, search data, location data).

What four main data dimensions compose Point-of-Sale (POS) scanner data?

Geography $\\times$ Product $\\times$ Time $\\times$ Variables (such as price, promotion, and end-of-aisle displays).

What four layers make up the Point-of-Sale (POS) data chain?

$80\\text{--}100$ CPG manufacturers, many major warehouse and distribution centers, $45,000$ supermarkets, and $135,000,000$ households.

Why can aggregate POS sales data be misleading compared to household-level brand-switching matrix data?

Aggregate sales can hide severe customer churn; for example, the brand Silky maintained $150$ sales in both period $t\\_1$ and period $t\\_2$, but household data revealed it retained only $55$ original customers and lost $95$ to competitors while gaining new switchers.

In a brand-switching matrix, what do the diagonal values and off-diagonal values represent?

Diagonal values represent brand-loyal customers who bought the same brand in both time periods ($t\\_1$ and $t\\_2$). Off-diagonal values represent customers who switched brands between period $t\\_1$ and period $t\\_2$.

In the class brand-switching matrix example, what happened to the $300$ customers who bought Sunshine in period $t\\_1$?

In period $t\\_2$, $225$ remained loyal to Sunshine, $50$ switched to Sparkle, and $25$ switched to Silky.

What specific dips occur in sales volume directly before and after a promotional discount peak?

An anticipation dip occurs right before the promotion (consumers wait for the sale), and a post-promotion dip occurs right after (consumers stock up during the sale).

What are four major limitations of Point-of-Sale (POS) scanner data?

POS data may not cover every retailer or channel, observational POS data cannot establish causality, POS data lacks consumer psychographics, and POS data does not reveal the exact consideration set faced by the consumer at the decision time.

What three main reasons justify why companies pay significant amounts for syndicated POS data?

Breadth and comprehensiveness (linking marketing mix variables to aggregate and individual behavior), timeliness for managerial action, and data accuracy.

Which vendor firms represent Group 1 (Internet Communities) and Group 2 (Syndicated Retail Data)?

Group 1 (Internet Communities) includes Alida, C space, Ipsos, and itracks. Group 2 (Syndicated Retail Data) includes NielsenIQ (NIQ), Circana (IRI), and SPINS.

Which vendor firms represent Group 3 (Transaction/Purchase Intelligence) and Group 9 (Location/Geospatial Intelligence)?

Group 3 (Transaction/Purchase Intelligence) includes Bloomberg Second Measure and Cardlytics. Group 9 (Location/Geospatial Intelligence) includes Mogean and Foursquare.