Secondary Data In Marketing Research
Chunhua Wu
The University of British Columbia 2015-09-21 Mon
Outline
1 Last Class Review
2 A Closer Look at Secondary Data
3 Research Using Secondary Data: Online Dating Services
Research Approaches and Data Sources
Research Approach Purpose Research Design
Exploratory Provide insights - Focus groups
Research - In-depth interviews
Descriptive Describe market - Surveys
Research characteristics - Observations
Causal Determine cause-and-effect - Experiment
Secondary Data Advantages (p.63-65)
■ Advantages
■ Provide background information, clarify research problem ■ Provide alternative methods for primary data research ■ Build up research credibility
■ Sampling framework
■ Provide a solution to the problem
Secondary Data Limitations (p.65-67)
■ Limitations ■ Availability ■ Relevance ■ Accuracy ■ SufficiencyAlways check the existence of relevant, accurate secondary data that are sufficient to address the question before gathering primary data!
Secondary Data Sources
Secondary Data Sources Government Sources Syndicated SourcesPublications Internal Data Sources
Social Media Data Sources
Other Data Sources
Publications
■ Your secondary data quiz: use UBC library ■ IBIS World
■ Passport GMID
■ Print Measurement Bureau (PMB) ■ Statistics Canada
Internal Data: Customer Database
■ Data structure
Internal Data: Customer Database
■ Typical applications ■ Promotion ■ Targeting ■ Pricing decision ■ Product assortmentInternal Data: Customer Database
■ CRM / Direct Marketing
■ RFM framework: Recency, Frequency, Monetary ■ Ultimate goal: predict profitability
■ http://www.youtube.com/embed/OYohJxp2l9k
■ Use SPSS Statistics direct marketing analysis to gain insight
Internal Data: Web Analytics
■ Raw web log data vs. aggregate services ■ Google Analytics
Internal Data: Web Analytics
■ Typical applications
■ Search Engine Optimization (SEO) ■ Behavioral targeting
■ Customer segmentation ■ …
Government Data
■ A census is the procedure of systematically acquiring and recording information about the members of a given population
■ Application: market entry, etc.
Standardized/Syndicated Data
■ IRI scanner panel data
■ Provided by Information Resources Inc.
IRI Panel Applications
■ Typical applications
■ Customer segmentation
■ Market share and brand management ■ Promotional effectiveness
■ Pricing decision
■ Exercise
■ Q: sales,p: price ■ c: unit cost, 40 cents ■ Q=170−100×p
■ What is the optimal price?
Social Media Data: Customer Reviews/Feedbacks
Customer Reviews/Feedbacks
■ Applications
■ New product development ■ Service improvement
■ UNIQLO
Social Networks Data: User Activities
■ Application Programming Interface (API) ■ Facebook Analytics
■ http:
A Little Bit on Privacy Issues
■ Privacy issues
■ Behavioral tracking: browser wars ■ Identity theft
■ Information in exchange for rewards: Airmiles card ■ Personal information being sold
■ PRISM
Research Using Secondary Data: Online Dating
Services
A Dating Platform
Product Positioning
■ Positioning map ■ Free vs paid ■ Young vs old ■ Mass vs niche ■ Machine vs human ■ Traditional vs controversial ■ Casual vs seriousPricing
■ Revenue models of dating websites ■ “Free”: OkCupid
■ Ads
■ Premium: Match.com, Chemistry.com, etc
■ Membership fee
■ Communication fee
■ Viewing fee ■ Freemium: POF
■ A combination
The Dating Process
Registration Browsing Messaging
Reply Dating Leave Out A B C D E ■ User experience?
The Big Problem…
■ Websites’ incentives?
■ Consistent with customer needs?
■ “Dead” Profiles
■ What is the implication? ■ Chance people can reply?
■ Number of messages you should send?
Should You Pay for Online Dating?
■ Never
■ Why you should never pay for online dating? ■ 2010-04-07 by Christian Rudder
■ Paying for dates on match.com and eHarmony is fundamentally broken ■ http://web.archive.org/web/20101006104124/http:
//blog.okcupid.com/index.php/
Reasonable Arguments?
■ eHarmony
■ 20M users, $250M revenue (2009)
■ Membership fee: $19.95/month(one year), $29.95/month (6-month), $59.95/month
■ Subscriber turnover: 6.5 months
■ What is the real number of subscribers?
■ What is the chance that you are messaging to someone who could reply?
■ Match.com
■ Claimed profiles: 20,000,000; confirmed subscribers: 1,377,000 (6.9%)
Reasonable Arguments?
■ Desperation feedback loop
■ Message to more people, less reply rate
■ Less reply rate, need to message to more people
■ Cost of online dating
■ Match.com 4,380 people got married a year ■ $342M revenue
Anything Missing? – Why Still in Business
■ Member searching is non-random ■ Local market
■ Distinguish subscribers vs. non-subscribers
■ “Matching function” provided by the website ■ Recommendations
■ Algorithms, i.e. chemistry.com
■ Self selection
■ If one could easily get married in the offline channel, he is less likely to move to the online dating services
■ Try and learning
Would “Free” Solve the Problem?
■ “Free” means reduced user experience ■ Dis-utility from ads
■ Website less motivation in enhancing user experience
■ “Free” does not necessarily solve the “dead profile” problem ■ People try and leave
■ Observation
Bottom Line
■ Online dating is a growing, billion dollar industry
■ User experience and platform revenue interests are in conflicts
Online dating service provides would leverage user experience concerns and profitability for long-term sustainability.
One More Thing
Ashley Madison
Source: Almost None of the Women in the Ashley Madison Database Ever Used the Site
Critical Thinking Question
No critical thinking question today. :) Focus on your research project!
Punch Line
■ Secondary data sources
■ Applications for marketing research
■ Online dating services industry ■ Logic in the analysis
■ A few numbers may help deeply understand the industry ■ Criticize the arguments
Next Class
■ Content ■ Qualitative research ■ Focus groups ■ Projective techniques ■ Readings ■ Chapter 4 ■ What’s due?■ Critical thinking question 3
■ Bring some used magzines to class
Problem Formulation Research Design Data Collection Data Analysis Results & Recommendation Introduction Research Process Research Approaches Secondary Data Qualitative Research Observation Measurement & Attitude Scales Sampling TBD Midterm Data Analysis Overview Hypothesis Test: Association Hypothesis Test: Means Statistical Lab Correlation and Regression Advanced Topic TBD Presentation I Presentation II