This document describes a process used to profile mortgage loan data and compare it to national profiles in order to maximize TV exposure for funded mortgages. Key steps included appending a commercial segmentation scheme (PRIZM) to loan records, identifying clusters with over-penetration, and profiling these clusters based on demographics. Metrics were then created to rate TV channels and potentially optimize ad purchasing.
Step by step guide for a 2 hour introduction to PRIZM and some baseline to understand DEMOGRAPHIC options better. It is PDF export from PREZI
Altough Claritas Prizm will be replaced soon by "Nielsen Segmentation & Market Solutions" but IMHO, PRIZM is the best way to understand & feel demographic segmentation better.
The document discusses how market research can help businesses in three ways: 1) better targeting of customers geographically, 2) knowledge of customers' media habits, and 3) more accurate messaging. It then provides an example cluster profile from the PRIZM market research system called "Kids & Cul-de-Sacs", describing them as affluent suburban families with children. Finally, it argues that targeting customers based on their demographic and lifestyle clusters can help businesses more efficiently reach similar potential customers.
This presentation was the culmination of a 3 month long team effort. It is the result of extensive research, discovery and creativity. It was presented in March of 2008, in Marketing Behaviors class.
The document is a retail market analysis for the city of Lockhart, Texas. It summarizes the population demographics and psychographic profiles of Lockhart and surrounding trade areas. There is $149 million in potential retail sales in Lockhart's primary trade area and $79 million in potential sales in its central business district, with $68 million leaking out. The top lifestyle segments in the trade area include families with children, middle-aged singles, and retired individuals. The analysis identifies retail sectors that are underserved and provides recommendations to strengthen local businesses.
Experience Mazda Zoom Zoom Lifestyle and Culture by Visiting and joining the Official Mazda Community at http://www.MazdaCommunity.org for additional insight into the Zoom Zoom Lifestyle and special offers for Mazda Community Members. If you live in Arizona, check out CardinaleWay Mazda's eCommerce website at http://www.Cardinale-Way-Mazda.com
This document provides a segmentation and targeting analysis for marketing efforts. It identifies three key customer segments - Single Fabulous, Energetic Family, and Old Empty Nesters. For each segment, it describes the demographics, lifestyle and interests. It then recommends marketing channels and geographic focus areas for each segment. The insights suggest prioritizing efforts on the Single Fabulous segment due to their higher spending, while also targeting the larger Energetic Family segment. Partnerships with dealers and discounts are recommended to boost sales.
Marketing Research and Consumer Behavior Report for Sandals Resorts
Created a professional brand management market segmentation report for Sandals Resorts including a market situation analysis, psychographic, demographic, geodemographics, and product-related behavioral characteristics, consumer market segmentation analysis, identified Sandals Resorts target market, VALS2, PRIZM, consumer benefit segmentation, market size and sales potential, consumer decision process model, identified consumer problems and suggested marketing mix solutions to all problems, and identified opportunities for growth.
This document provides information about marketing segmentation for the Dallas Mavericks organization. It analyzes different fan segments including overall Mavericks fans, attendees, and listeners. It examines demographics like age, gender, income, and education. It also discusses business goals around ticket sales, TV ratings, merchandise sales, and growing the Hispanic fan base. Finally, it performs a zip code cluster analysis to identify key target areas for marketing based on demographics and lifestyles.
Step by step guide for a 2 hour introduction to PRIZM and some baseline to understand DEMOGRAPHIC options better. It is PDF export from PREZI
Altough Claritas Prizm will be replaced soon by "Nielsen Segmentation & Market Solutions" but IMHO, PRIZM is the best way to understand & feel demographic segmentation better.
The document discusses how market research can help businesses in three ways: 1) better targeting of customers geographically, 2) knowledge of customers' media habits, and 3) more accurate messaging. It then provides an example cluster profile from the PRIZM market research system called "Kids & Cul-de-Sacs", describing them as affluent suburban families with children. Finally, it argues that targeting customers based on their demographic and lifestyle clusters can help businesses more efficiently reach similar potential customers.
This presentation was the culmination of a 3 month long team effort. It is the result of extensive research, discovery and creativity. It was presented in March of 2008, in Marketing Behaviors class.
The document is a retail market analysis for the city of Lockhart, Texas. It summarizes the population demographics and psychographic profiles of Lockhart and surrounding trade areas. There is $149 million in potential retail sales in Lockhart's primary trade area and $79 million in potential sales in its central business district, with $68 million leaking out. The top lifestyle segments in the trade area include families with children, middle-aged singles, and retired individuals. The analysis identifies retail sectors that are underserved and provides recommendations to strengthen local businesses.
Experience Mazda Zoom Zoom Lifestyle and Culture by Visiting and joining the Official Mazda Community at http://www.MazdaCommunity.org for additional insight into the Zoom Zoom Lifestyle and special offers for Mazda Community Members. If you live in Arizona, check out CardinaleWay Mazda's eCommerce website at http://www.Cardinale-Way-Mazda.com
This document provides a segmentation and targeting analysis for marketing efforts. It identifies three key customer segments - Single Fabulous, Energetic Family, and Old Empty Nesters. For each segment, it describes the demographics, lifestyle and interests. It then recommends marketing channels and geographic focus areas for each segment. The insights suggest prioritizing efforts on the Single Fabulous segment due to their higher spending, while also targeting the larger Energetic Family segment. Partnerships with dealers and discounts are recommended to boost sales.
Marketing Research and Consumer Behavior Report for Sandals Resorts
Created a professional brand management market segmentation report for Sandals Resorts including a market situation analysis, psychographic, demographic, geodemographics, and product-related behavioral characteristics, consumer market segmentation analysis, identified Sandals Resorts target market, VALS2, PRIZM, consumer benefit segmentation, market size and sales potential, consumer decision process model, identified consumer problems and suggested marketing mix solutions to all problems, and identified opportunities for growth.
This document provides information about marketing segmentation for the Dallas Mavericks organization. It analyzes different fan segments including overall Mavericks fans, attendees, and listeners. It examines demographics like age, gender, income, and education. It also discusses business goals around ticket sales, TV ratings, merchandise sales, and growing the Hispanic fan base. Finally, it performs a zip code cluster analysis to identify key target areas for marketing based on demographics and lifestyles.
The document discusses social class and its relationship to consumer behavior. It defines social class as the hierarchical division of society based on status, and examines how social class is measured through subjective, reputational, and objective means. Some of the key social class categories discussed are the upper-upper class, lower-upper class, upper-middle class, lower-middle class, upper-lower class, lower-lower class, and the affluent consumer. The document also explores indexes used to measure social class composites like the Index of Status Characteristics and Socioeconomic Status Score.
2010 Chatham Square Case Study In Healthy Neighborhood Approaches To Communit...Lee Cruz
This document summarizes a case study of the Chatham Square neighborhood in New Haven, Connecticut that utilized a "healthy neighborhood approach" to community development. Key aspects of this approach included encouraging neighbor participation and cooperation, measuring success by quality of life outcomes rather than outputs, and focusing on building social capital and relationships within the community. Through this approach, the neighborhood saw improvements in image, physical conditions, management of issues, and real estate market activity.
To scale and grow your digital marketing agency, you need to offer more to your clients. Easy right? Not so fast.
Do you spend more money and time hiring and training staff to support new services, or do you risk quality of work by bringing in independent contractors?
Check out these webinar slides to learn clever, time-saving ways to expand your agency offerings and increase profits. You’ll also learn how to:
• Fill gaps that are hurting your revenue.
• Gain and retain quality clients.
• Use a secret swiss army knife of targeting tools not seen anywhere else.
Using Analytics to Understand Consumers, Neighborhoods and AdvertisingPrecisely
Customer analytics vary by industry – what’s important for retail clothing stores isn’t always important for a telecom client, etc. Join Precisely Demographics product managers Andy Peloe and Dylan Conrad as they describe how Precisely has done the work of aggregating diverse types of data so that you can more easily answer important questions for your business as well as your customers.
This webinar will help you:
- Partner with your clients and choose the right data variables and granularity
- More accurately identify target customer markets based on dynamic demographics
- Understand retail performance and site selection based on neighborhood profiles
- Use measured analytics to tell a story about a neighborhood
- Identify trends to support investment strategies
-Accurately assign risk
Castro Village Bowl is a bowling alley located in Castro Valley, California that has been in business since 1961. While it has a loyal customer base and offers affordable prices and frequent promotions, it is missing out on attracting younger customers. Research methods including surveys, interviews, and analyzing social media and review sites found that the bowling alley lacks a strong social media presence and many of its events are geared toward older patrons. A proposed strategic plan includes partnering with the local high school on youth-focused events, updating the alley's appearance, and growing its social media presence in order to attract more young customers and their families.
The document describes a system called LOANTRAX that was developed to help banks and regulators assess lending performance and potential. It did this by analyzing credit bureau and demographic data to model potential demand for various financial products across geographic areas. This provided banks and regulators with metrics to quantify lending performance versus potential, helping banks address areas of high potential but low presence and assisting regulators in ensuring compliance with laws like CRA and HMDA.
Urban Roots is expanding its nonprofit operations from Austin, Texas to Atlanta, Georgia. The new location in Old Fourth Ward offers opportunities to engage potential supporters in the trendy neighborhood. Urban Roots' target market is wealthy, educated young mothers who value healthy living and the environment. While Atlanta has many similar organizations focused on urban farming and food access, Urban Roots stands out by donating a higher percentage of its produce and offering paid youth internships. The new location provides a way for Urban Roots to make a positive impact in Atlanta.
This document discusses how knowledge management can help build strategic partnerships between communications and campaign teams. It outlines how analyzing constituent data through tools like surveys, predictive modeling, and market segmentation can help organizations understand their audiences. This understanding then allows organizations to test assumptions about constituents and craft targeted communications strategies. Effective knowledge management requires integrating information from various systems and partnering between teams to acquire, analyze and act on constituent insights.
Advertising Awards given during the SDNA Convention in Aberdeen, SD, April 28...David Bordewyk
This document lists the winners of the 2016 Better Newspaper Advertising Awards for South Dakota newspapers. It provides the winners in different newspaper circulation categories for "Best 2x4 Ad" and "Best Advertising Sales Tool". Winners are listed for weeklies under 1,150 circulation, between 1,151-2,000 circulation, and over 2,000 circulation. Winners are also listed for all dailies. First, second, and third place winners are provided for each category and circulation level.
This document discusses strategies for identifying and cultivating major and planned gift donors, referred to as "top of the pyramid" or "principal" donors. It provides an overview of research on donor profiles and giving behaviors. Donors who make major gifts or planned gifts tend to share demographic characteristics like age, household size, wealth indicators, and philanthropic histories. The document outlines approaches like wealth screening, predictive modeling, and cluster analysis that nonprofits can use to identify and understand top prospective donors based on their giving behaviors and characteristics.
Regional Open House Presentation-April 2014Heartland2050
John Fregonese, Principal of Fregonese Associates and lead consultant on the project, provided the public with an update and review of the four scenarios for growth over the next 40 years.
This presentation was also presented at the April 17th Steering Committee meeting.
Lo-Cal International is launching a line of low-calorie frozen meals targeting dieters, especially women aged 25-54. Their $8 million media budget will be spent on national advertising. Research shows their target audience watches morning and primetime TV, reads magazines about health, home, and lifestyle. Lo-Cal will run ads in SELF, Shape and Weight Watchers magazines monthly or quarterly to reach readers interested in health, fitness and dieting. TV and magazines will allow visualization of products and reach consumers nationwide during peak sales seasons.
The marketing proposal summarizes CBS Radio Houston's strategy to help drive women ages 35-54 to call ABC Home & Commercial Services for home services. The strategy includes radio spots on MIX 96.5, SportsRadio 610, and The Bull stations, which heavily target homeowners in this demographic. It also proposes email marketing, video pre-roll ads, and sponsorships of various community events to raise awareness of ABC and increase attendance at A Child's Hope Gala to help more people in need. The total proposed investment is $313,595 for the year.
◆ Cooperation with High-Level Executives
◆ Uniforming Visual Aesthetic
◆ Fully Editable Format
Quarterly report - Yates on United States Demographics. Analysis of Life Modes, Socioeconomic Traits, etc.
If you search for a presentation designer, please email me at design@natakostenko.com.
A Place for Us Presentation (June 22, 2007)Andy Carswell
This slide show describes some of the challenges facing Athens, GA and some of the surrounding areas in terms of housing. It was presented in mid-2007.
UPSTART Live Spring Summit - Rocking The AgesWorkforceNEXT
Rocking The Ages: Retention from a Generational Perspective.
Presented By Mayerland Harris, HEB; Laura Ramey, Crestwood Midstream Partners; Edda Tinis, Air Liquide
Real Estate Residential Market Research Study in North Carolina CabaData
Moving to a different state is stressful. However, if you have hashtag#Industry hashtag#Professionals that can help you find a home, your stress levels will reduce dramatically like this ->
This was the situation with my friends moving from Miami, FL to Charlotte, North Carolina. I wrote a hashtag#linkedin hashtag#article about it found at the link -> https://lnkd.in/eDS7ah5
Also, I placed the entire hashtag#research completed on hashtag#slideshare. I will post the link in the comment area below as well.
#Comment #like and give me your opinion
#northcarolina #northcarolinalivingfornow #rentalmarket #northcarolinaapartments #northcarolinalifestyle #northcarolinajobs
This document provides a socio-demographic and health overview of Rugby Borough in the United Kingdom. It summarizes data on the borough's demographics, deprivation levels, economy, education, community safety, population subgroups as defined by a Mosaic classification system, and key health issues like obesity, alcohol, smoking, and mental health. The purpose is to identify current and future health and wellbeing needs to inform local priorities and commissioning.
This document discusses the development of a model to predict patient mortality using continuous electrocardiogram (ECG) and laboratory data collected from patients in the pediatric intensive care unit (PICU). Data from multiple sources with different time units were converted to a common 2-minute time frame. Predictive models were developed using time and frequency domain features extracted from the ECG data as well as physiological lab tests. The models provide risk scores and interactive reports in real-time to allow clinicians to monitor patients and update diagnoses continuously. Dimensionality reduction and hierarchical clustering were used to organize and visualize the relationships between high-dimensional patient feature vectors over time.
This document analyzes changes to TV media purchases between Q4 2005 and Q1 2006 and Q3 2006 by looking at the top 20% channels purchased, day of week, and day part. It identifies monthly, daily, and time period changes in ad purchases across different visualizations to understand how TV spending has changed and how this relates to other marketing initiatives.
The document discusses social class and its relationship to consumer behavior. It defines social class as the hierarchical division of society based on status, and examines how social class is measured through subjective, reputational, and objective means. Some of the key social class categories discussed are the upper-upper class, lower-upper class, upper-middle class, lower-middle class, upper-lower class, lower-lower class, and the affluent consumer. The document also explores indexes used to measure social class composites like the Index of Status Characteristics and Socioeconomic Status Score.
2010 Chatham Square Case Study In Healthy Neighborhood Approaches To Communit...Lee Cruz
This document summarizes a case study of the Chatham Square neighborhood in New Haven, Connecticut that utilized a "healthy neighborhood approach" to community development. Key aspects of this approach included encouraging neighbor participation and cooperation, measuring success by quality of life outcomes rather than outputs, and focusing on building social capital and relationships within the community. Through this approach, the neighborhood saw improvements in image, physical conditions, management of issues, and real estate market activity.
To scale and grow your digital marketing agency, you need to offer more to your clients. Easy right? Not so fast.
Do you spend more money and time hiring and training staff to support new services, or do you risk quality of work by bringing in independent contractors?
Check out these webinar slides to learn clever, time-saving ways to expand your agency offerings and increase profits. You’ll also learn how to:
• Fill gaps that are hurting your revenue.
• Gain and retain quality clients.
• Use a secret swiss army knife of targeting tools not seen anywhere else.
Using Analytics to Understand Consumers, Neighborhoods and AdvertisingPrecisely
Customer analytics vary by industry – what’s important for retail clothing stores isn’t always important for a telecom client, etc. Join Precisely Demographics product managers Andy Peloe and Dylan Conrad as they describe how Precisely has done the work of aggregating diverse types of data so that you can more easily answer important questions for your business as well as your customers.
This webinar will help you:
- Partner with your clients and choose the right data variables and granularity
- More accurately identify target customer markets based on dynamic demographics
- Understand retail performance and site selection based on neighborhood profiles
- Use measured analytics to tell a story about a neighborhood
- Identify trends to support investment strategies
-Accurately assign risk
Castro Village Bowl is a bowling alley located in Castro Valley, California that has been in business since 1961. While it has a loyal customer base and offers affordable prices and frequent promotions, it is missing out on attracting younger customers. Research methods including surveys, interviews, and analyzing social media and review sites found that the bowling alley lacks a strong social media presence and many of its events are geared toward older patrons. A proposed strategic plan includes partnering with the local high school on youth-focused events, updating the alley's appearance, and growing its social media presence in order to attract more young customers and their families.
The document describes a system called LOANTRAX that was developed to help banks and regulators assess lending performance and potential. It did this by analyzing credit bureau and demographic data to model potential demand for various financial products across geographic areas. This provided banks and regulators with metrics to quantify lending performance versus potential, helping banks address areas of high potential but low presence and assisting regulators in ensuring compliance with laws like CRA and HMDA.
Urban Roots is expanding its nonprofit operations from Austin, Texas to Atlanta, Georgia. The new location in Old Fourth Ward offers opportunities to engage potential supporters in the trendy neighborhood. Urban Roots' target market is wealthy, educated young mothers who value healthy living and the environment. While Atlanta has many similar organizations focused on urban farming and food access, Urban Roots stands out by donating a higher percentage of its produce and offering paid youth internships. The new location provides a way for Urban Roots to make a positive impact in Atlanta.
This document discusses how knowledge management can help build strategic partnerships between communications and campaign teams. It outlines how analyzing constituent data through tools like surveys, predictive modeling, and market segmentation can help organizations understand their audiences. This understanding then allows organizations to test assumptions about constituents and craft targeted communications strategies. Effective knowledge management requires integrating information from various systems and partnering between teams to acquire, analyze and act on constituent insights.
Advertising Awards given during the SDNA Convention in Aberdeen, SD, April 28...David Bordewyk
This document lists the winners of the 2016 Better Newspaper Advertising Awards for South Dakota newspapers. It provides the winners in different newspaper circulation categories for "Best 2x4 Ad" and "Best Advertising Sales Tool". Winners are listed for weeklies under 1,150 circulation, between 1,151-2,000 circulation, and over 2,000 circulation. Winners are also listed for all dailies. First, second, and third place winners are provided for each category and circulation level.
This document discusses strategies for identifying and cultivating major and planned gift donors, referred to as "top of the pyramid" or "principal" donors. It provides an overview of research on donor profiles and giving behaviors. Donors who make major gifts or planned gifts tend to share demographic characteristics like age, household size, wealth indicators, and philanthropic histories. The document outlines approaches like wealth screening, predictive modeling, and cluster analysis that nonprofits can use to identify and understand top prospective donors based on their giving behaviors and characteristics.
Regional Open House Presentation-April 2014Heartland2050
John Fregonese, Principal of Fregonese Associates and lead consultant on the project, provided the public with an update and review of the four scenarios for growth over the next 40 years.
This presentation was also presented at the April 17th Steering Committee meeting.
Lo-Cal International is launching a line of low-calorie frozen meals targeting dieters, especially women aged 25-54. Their $8 million media budget will be spent on national advertising. Research shows their target audience watches morning and primetime TV, reads magazines about health, home, and lifestyle. Lo-Cal will run ads in SELF, Shape and Weight Watchers magazines monthly or quarterly to reach readers interested in health, fitness and dieting. TV and magazines will allow visualization of products and reach consumers nationwide during peak sales seasons.
The marketing proposal summarizes CBS Radio Houston's strategy to help drive women ages 35-54 to call ABC Home & Commercial Services for home services. The strategy includes radio spots on MIX 96.5, SportsRadio 610, and The Bull stations, which heavily target homeowners in this demographic. It also proposes email marketing, video pre-roll ads, and sponsorships of various community events to raise awareness of ABC and increase attendance at A Child's Hope Gala to help more people in need. The total proposed investment is $313,595 for the year.
◆ Cooperation with High-Level Executives
◆ Uniforming Visual Aesthetic
◆ Fully Editable Format
Quarterly report - Yates on United States Demographics. Analysis of Life Modes, Socioeconomic Traits, etc.
If you search for a presentation designer, please email me at design@natakostenko.com.
A Place for Us Presentation (June 22, 2007)Andy Carswell
This slide show describes some of the challenges facing Athens, GA and some of the surrounding areas in terms of housing. It was presented in mid-2007.
UPSTART Live Spring Summit - Rocking The AgesWorkforceNEXT
Rocking The Ages: Retention from a Generational Perspective.
Presented By Mayerland Harris, HEB; Laura Ramey, Crestwood Midstream Partners; Edda Tinis, Air Liquide
Real Estate Residential Market Research Study in North Carolina CabaData
Moving to a different state is stressful. However, if you have hashtag#Industry hashtag#Professionals that can help you find a home, your stress levels will reduce dramatically like this ->
This was the situation with my friends moving from Miami, FL to Charlotte, North Carolina. I wrote a hashtag#linkedin hashtag#article about it found at the link -> https://lnkd.in/eDS7ah5
Also, I placed the entire hashtag#research completed on hashtag#slideshare. I will post the link in the comment area below as well.
#Comment #like and give me your opinion
#northcarolina #northcarolinalivingfornow #rentalmarket #northcarolinaapartments #northcarolinalifestyle #northcarolinajobs
This document provides a socio-demographic and health overview of Rugby Borough in the United Kingdom. It summarizes data on the borough's demographics, deprivation levels, economy, education, community safety, population subgroups as defined by a Mosaic classification system, and key health issues like obesity, alcohol, smoking, and mental health. The purpose is to identify current and future health and wellbeing needs to inform local priorities and commissioning.
This document discusses the development of a model to predict patient mortality using continuous electrocardiogram (ECG) and laboratory data collected from patients in the pediatric intensive care unit (PICU). Data from multiple sources with different time units were converted to a common 2-minute time frame. Predictive models were developed using time and frequency domain features extracted from the ECG data as well as physiological lab tests. The models provide risk scores and interactive reports in real-time to allow clinicians to monitor patients and update diagnoses continuously. Dimensionality reduction and hierarchical clustering were used to organize and visualize the relationships between high-dimensional patient feature vectors over time.
This document analyzes changes to TV media purchases between Q4 2005 and Q1 2006 and Q3 2006 by looking at the top 20% channels purchased, day of week, and day part. It identifies monthly, daily, and time period changes in ad purchases across different visualizations to understand how TV spending has changed and how this relates to other marketing initiatives.
This document section describes extracting relevant data from streaming ECG data, clinical lab results, and demographic patient charts to develop multivariate models for predicting patient mortality risk. It outlines processing steps like filtering ECG data to identify heartbeats, calculating heart rate and frequency domain variables, and developing derogatory biomarkers from lab tests and charts by identifying intervals that correlate with higher mortality risk. The goal is to handle variable data lengths and correlations to include meaningful information from these diverse clinical sources in predictive models.
Regulatory credit risk reporting was optimized by developing a centralized CASE-Condition database defining key metrics like balances and interest rates across all consumer credit products. This helped establish consistent definitions enterprise-wide. The database was used to audit and optimize legacy SQL code retrieving the metrics from origination and portfolio tables for each product monthly. The optimized SQL reduced runtime and storage usage by up to 80%. Common metrics were also collected in a new reporting system with specific CASE conditions and aliases.
The document summarizes an analysis comparing actual media buys for early morning spots in Q1 2007 to optimized solutions developed using linear programming. The linear programming model aimed to maximize a multivariate target score while adhering to budget and channel constraints. Results showed the optimized solution selected more spots within the same budget, including stations not previously bought, though further testing is needed to confirm the expected benefits on qualifying applicants.
This document discusses the build history of a neural network model that identifies an optimal point between a hold out sample and model building database. It also examines time series data broken into trend and cyclical components, noting that peaks are generally underestimated. Additionally, it determines the impacts of different factors on data, how a target is affected by changing one factor, and what happens when two factors are changed simultaneously.
Major shifts occurred during July-August 2006 as leads went flat, the average 30-year mortgage rate went down, and the market composite and refinance index went up, while the refinance percentage of business went up and purchase percentage went down. Additional major shifts happened after the first quarter of 2007 as leads went down, spots and spot costs increased, the average 30-year rate rose, and the market composite, refinance index, and refinance percentage of business fell, while the purchase percentage of business rose.
This document analyzes sales data from 2007 to identify peak sales periods and best-selling brands and items within different product categories like coffee, beer, candy, tobacco, and water. Key findings include that beer sales peak at 5PM and Budweiser and 12-packs of Bud Light sell best. Coffee sales are highest at 7AM. Candy sales increase in the afternoon. Tobacco sales peak at 5PM and Marlboro lights are the top-selling brand and product type. Water sells best in Virginia, particularly the city of Fredericksburg, and popular brands include Glaceau, Aquafina, and Gatorade.
Daniel Kocis is the president of Applied Multivariate Algorithms Inc. He has extensive experience developing SAS models and reporting systems to support regulatory risk reporting and credit risk management at a large bank. Some of his work includes:
1) Developing a model risk management tool for consumer credit cards that automated model building, validation, and tracking.
2) Creating risk reporting and data governance processes across multiple lines of business.
3) Developing models and reports to track credit performance, delinquency rates, and risk exposures across all of the bank's consumer credit products.
4) Using credit bureau data to profile auto and specialty loan portfolios and track their credit risk characteristics.
This document describes services from Multivariate Algorithms Inc. to analyze web log data and visitor behavior in three main points:
1. They convert web logs into sessionized visitor records and analyze page requests to determine what content each visitor accessed and their discovery paths through the website.
2. As an example, they performed an analysis for Abbott Laboratories that showed how TV advertisements drove increases in website traffic and how most visitors engaged with specific product pages.
3. Their sessionization and analysis provides more accurate traffic counts and visitor behavior than other tools by accounting for all log file records and identifying non-human traffic that can inflate numbers.
The document discusses a statistical tool called the Multivariate Media Model that can identify the characteristics of high performing media placements based on factors like creative, placement, cost, and call volume. It finds over 1,000 combinations that are classified as "Keepers" that generate more calls at a lower cost per call compared to "Losers". Adopting the insights from this model could potentially save over $3 million by shifting placements to higher performing time periods, day parts, and formats. The tool uses a proprietary technique to analyze interactions between multiple predictor variables and identify significant rules that describe high response segments.
2. 2
Overview
This report summarizes a process that was used to create a quantitative selection
process that focuses on characteristics of current business practices and
can be used to maximize our TV exposure for funded 1st mortgages by:
* profiling data and comparing it to national profiles
* providing profiles on specific outcomes of our marketing efforts
* describing how to integrate Nielsen TV ratings to these key demographics
* creating key metrics to rate each Channel by daypart
* combining these metrics into a single composite score
* using these scores to pick BEST channels by daypart
* compare this outcome with current AD purchasing
3. 3
Background, rational and setup
We used a data base profiling approach that looked at different outcomes of all 2006
business and included
83,996 fundings (26,437 1st Mortgages and 55,229 2nd Mortgages),
422,398 cancellations (237,108 1st Mortgages and 162,567 2nd Mortgages), and
100,000 dead leads (single time callers).
The first step was to append a commercial segmentation scheme called PRIZM to each
record. This segmentation classifies each record into 1 of 66 different clusters based upon
age, income, marital status, home owner status, household composition, ethnicity, and
geographic location. We calculated an index against the National profile and identified
13 clusters where we over-penetrate (index 135+) for 1st Mortgages
19 clusters where we over-penetrate (index 135+) for 2nd Mortgages
13 clusters where we over-penetrate (index 135+) for 1st Mortgages Cancellation
18 clusters where we over-penetrate (index 135+) for 2nd Mortgages Cancellation
4. 4
The clusters where is dominant for 1st Mortgages
Label COUNT PERCENT HH USPercent index Urbanicity HH Income
White Picket Fences 652 2.47 1,403,531 1.25 1.97 Second City Midscale
Kids and Cul-de-Sacs 832 3.15 1,828,699 1.63 1.93 Suburban Upper-Mid
Upward Bound 785 2.97 1,793,920 1.6 1.86 Second City Upscale
American Dreams 1,004 3.8 2,447,099 2.18 1.74 Urban Midscale
New Homesteaders 919 3.48 2,254,616 2.01 1.73 Town Upper-Mid
Beltway Boomers 416 1.57 1,079,269 0.96 1.64 Suburban Upper-Mid
Fast-Track Families 713 2.7 1,950,575 1.74 1.55 Town/Rural Upscale
Blue-Chip Blues 489 1.85 1,400,592 1.25 1.48 Suburban Midscale
Winner's Circle 412 1.56 1,239,200 1.1 1.42 Suburban Wealthy
Suburban Sprawl 490 1.85 1,473,003 1.31 1.41 Suburban Midscale
Kid Country, USA 489 1.85 1,500,755 1.34 1.38 Town Lower-Mid
Home Sweet Home 669 2.53 2,062,147 1.84 1.38 Suburban Upper-Mid
The Cosmopolitans 422 1.6 1,317,884 1.17 1.36 Urban Midscale
Label HH Age Range HH Comp HH Tenure HH Education HH Employment HH Race & Ethnicity HH IPA
White Picket Fences Age 25-44 HH w/ Kids Mix, Owners Some College BC, Srv, Mix W, B, A, H, Mix Moderate
Kids and Cul-de-Sacs Age 25-44 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix Above Avg.
Upward Bound Age 35-54 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix Above Avg.
American Dreams Age 35-54 Mostly w/ Kids Homeowners Some College WC, Srv, Mix W, B, A, H, Mix Above Avg.
New Homesteaders Age 25-44 HH w/ Kids Mostly Owners College Grad BC, Srv, Mix W Above Avg.
Beltway Boomers Age 45-64 HH w/ Kids Mostly Owners College Grad WC, Mix W, B, A, Mix Above Avg.
Fast-Track Families Age 35-54 HH w/ Kids Mostly Owners College Grad Management W Above Avg.
Blue-Chip Blues Age 25-44 HH w/ Kids Mix, Owners Some College BC, Srv, Mix W, B, A, H, Mix Below Avg.
Winner's Circle Age 25-44 HH w/ Kids Mostly Owners Grad Plus Management W, A, Mix High
Suburban Sprawl Age 35-54 HH w/o Kids Homeowners College Grad Professional W, B, A, Mix Moderate
Kid Country, USA Age 25-44 HH w/ Kids Mix, Owners H.S. Grad BC, Srv, Mix W, B, H, Mix Below Avg.
Home Sweet Home Age <55 HH w/o Kids Mostly Owners College Grad Professional W, B, A, Mix Above Avg.
The Cosmopolitans Age 55+ Mostly w/o Kids Homeowners Some College WC, Mix W, B, A, H, Mix High
5. 5
The clusters where is dominant for 2nd Mortgages
Label COUNT PERCENT HH USPercent index Urbanicity HH Income
Kids and Cul-de-Sacs 2,076 3.76 1,828,699 1.63 2.31 Suburban Upper-Mid
Upward Bound 1,934 3.50 1,793,920 1.60 2.19 Second City Upscale
New Homesteaders 2,207 4.00 2,254,616 2.01 1.99 Town Upper-Mid
Winner's Circle 1,163 2.11 1,239,200 1.10 1.91 Suburban Wealthy
Fast-Track Families 1,724 3.12 1,950,575 1.74 1.79 Town/Rural Upscale
White Picket Fences 1,236 2.24 1,403,531 1.25 1.79 Second City Midscale
Country Squires 1,866 3.38 2,152,742 1.92 1.76 Town/Rural Upscale
Beltway Boomers 900 1.63 1,079,269 0.96 1.70 Suburban Upper-Mid
Greenbelt Sports 1,294 2.34 1,612,141 1.44 1.63 Town/Rural Upper-Mid
God's Country 1,347 2.44 1,735,899 1.55 1.57 Town/Rural Upscale
Brite Lites, Li'l City 1,282 2.32 1,684,994 1.50 1.55 Second City Upscale
Home Sweet Home 1,552 2.81 2,062,147 1.84 1.53 Suburban Upper-Mid
Movers and Shakers 1,343 2.43 1,807,572 1.61 1.51 Suburban Wealthy
American Dreams 1,795 3.25 2,447,099 2.18 1.49 Urban Midscale
Big Sky Families 1,472 2.67 2,014,484 1.79 1.49 Rural Upper-Mid
Country Casuals 1,274 2.31 1,807,787 1.61 1.43 Town/Rural Upscale
Blue-Chip Blues 970 1.76 1,400,592 1.25 1.41 Suburban Midscale
Pools and Patios 1,009 1.83 1,470,884 1.31 1.39 Suburban Upper-Mid
Blue Blood Estates 731 1.32 1,094,703 0.98 1.35 Suburban Wealthy
Label HH Age Range HH Comp HH Tenure HH Education HH Employment HH Race & Ethnicity HH IPA
Kids and Cul-de-Sacs Age 25-44 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix Above Avg.
Upward Bound Age 35-54 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix Above Avg.
New Homesteaders Age 25-44 HH w/ Kids Mostly Owners College Grad BC, Srv, Mix W Above Avg.
Winner's Circle Age 25-44 HH w/ Kids Mostly Owners Grad Plus Management W, A, Mix High
Fast-Track Families Age 35-54 HH w/ Kids Mostly Owners College Grad Management W Above Avg.
White Picket Fences Age 25-44 HH w/ Kids Mix, Owners Some College BC, Srv, Mix W, B, A, H, Mix Moderate
Country Squires Age 35-54 HH w/ Kids Mostly Owners Grad Plus Management W High
Beltway Boomers Age 45-64 HH w/ Kids Mostly Owners College Grad WC, Mix W, B, A, Mix Above Avg.
Greenbelt Sports Age 35-54 HH w/o Kids Mostly Owners College Grad WC, Mix W Above Avg.
God's Country Age 35-54 HH w/o Kids Mostly Owners College Grad Management W High
Brite Lites, Li'l City Age 35-54 HH w/o Kids Mostly Owners College Grad Professional W, A, Mix Above Avg.
Home Sweet Home Age <55 HH w/o Kids Mostly Owners College Grad Professional W, B, A, Mix Above Avg.
Movers and Shakers Age 35-54 HH w/o Kids Mostly Owners Grad Plus Management W, A, Mix High
American Dreams Age 35-54 Mostly w/ Kids Homeowners Some College WC, Srv, Mix W, B, A, H, Mix Above Avg.
Big Sky Families Age 25-44 HH w/ Kids Mostly Owners Some College BC, Srv, Mix W Moderate
Country Casuals Age 35-54 HH w/o Kids Mostly Owners College Grad Management W High
Blue-Chip Blues Age 25-44 HH w/ Kids Mix, Owners Some College BC, Srv, Mix W, B, A, H, Mix Below Avg.
Pools and Patios Age 45-64 HH w/o Kids Mostly Owners College Grad Professional W, A, Mix High
Blue Blood Estates Age 45-64 HH w/ Kids Mostly Owners Grad Plus Management W, A, Mix High
6. 6
Where are cancels and single time callers coming from ?
The same process was repeated for cancellations, but we are reserving identifying these
clusters until we can differentiate applications that were denied versus true cancels.
Single time callers that never pursued an application were labeled dead leads (see below)
Label COUNT PERCENT HH USPercent index Urbanicity HH Income
Beltway Boomers 611.00 1.35 1,079,269.00 0.96 1.40 Suburban Upper-Mid
Blue Highways 899.00 1.98 1,644,447.00 1.46 1.36 Rural Lower-Mid
Upward Bound 985.00 2.17 1,793,920.00 1.6 1.36 Second City Upscale
Fast-Track Families 1,071.00 2.36 1,950,575.00 1.74 1.36 Town/Rural Upscale
New Homesteaders 1,236.00 2.73 2,254,616.00 2.01 1.36 Town Upper-Mid
Kids and Cul-de-Sacs 999.00 2.20 1,828,699.00 1.63 1.35 Suburban Upper-Mid
Label HH Age Range HH Comp HH Tenure HH Education HH Employment HH Race & Ethnicity HH IPA
Beltway Boomers Age 45-64 HH w/ Kids Mostly Owners College Grad WC, Mix W, B, A, Mix Above Avg.
Blue Highways Age 35-54 HH w/o Kids Homeowners H.S. Grad BC, Srv, Mix W Moderate
Upward Bound Age 35-54 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix Above Avg.
Fast-Track Families Age 35-54 HH w/ Kids Mostly Owners College Grad Management W Above Avg.
New Homesteaders Age 25-44 HH w/ Kids Mostly Owners College Grad BC, Srv, Mix W Above Avg.
Kids and Cul-de-Sacs Age 25-44 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix Above Avg.
7. 7
How to identify Total US Mortgage Market through
a definitional set of criteria within PRIZM
Segment Nickname Urbanicity HH Income HH Age Range HH Comp HH Tenure HH Education HH Employment HH Race & Ethnicity
Upper Crust Suburban Wealthy Age 45-64 HH w/o Kids Mostly Owners Grad Plus Professional W, A, Mix
Blue Blood Estates Suburban Wealthy Age 45-64 HH w/ Kids Mostly Owners Grad Plus Management W, A, Mix
Movers & Shakers Suburban Wealthy Age 35-54 HH w/o Kids Mostly Owners Grad Plus Management W, A, Mix
Young Digerati Urban Upscale Age 25-44 Family Mix Mix, Owners Grad Plus Professional W, A, H, Mix
Country Squires Town/Rural Upscale Age 35-54 HH w/ Kids Mostly Owners Grad Plus Management W
Winner's Circle Suburban Wealthy Age 25-44 HH w/ Kids Mostly Owners Grad Plus Management W, A, Mix
Money & Brains Urban Upscale Age 45-64 Family Mix Mostly Owners Grad Plus Professional W, A, H, Mix
Executive Suites Suburban Upper-Mid Age 35-54 HH w/o Kids Mostly Owners College Grad Professional W, A, Mix
Big Fish, Small Pond Town/Rural Upscale Age 45-64 HH w/o Kids Mostly Owners Grad Plus Management W
Second City Elite Second City Upscale Age 45-64 HH w/o Kids Mostly Owners Grad Plus WC, Mix W
God's Country Town/Rural Upscale Age 35-54 HH w/o Kids Mostly Owners College Grad Management W
Brite Lites, Li'l City Second City Upscale Age 35-54 HH w/o Kids Mostly Owners College Grad Professional W, A, Mix
Upward Bound Second City Upscale Age 35-54 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix
New Empty Nests Suburban Upper-Mid Age 65+ HH w/o Kids Mostly Owners College Grad Retired W
Pools & Patios Suburban Upper-Mid Age 45-64 HH w/o Kids Mostly Owners College Grad Professional W, A, Mix
Beltway Boomers Suburban Upper-Mid Age 45-64 HH w/ Kids Mostly Owners College Grad WC, Mix W, B, A, Mix
Kids & Cul-de-sacs Suburban Upper-Mid Age 25-44 HH w/ Kids Mostly Owners College Grad WC, Mix W, A, H, Mix
Home Sweet Home Suburban Upper-Mid Age <55 HH w/o Kids Mostly Owners College Grad Professional W, B, A, Mix
Fast-Track Families Town/Rural Upscale Age 35-54 HH w/ Kids Mostly Owners College Grad Management W
Gray Power Suburban Midscale Age 65+ Mostly w/o Kids Mostly Owners College Grad Retired W
Greenbelt Sports Town/Rural Upper-Mid Age 35-54 HH w/o Kids Mostly Owners College Grad WC, Mix W
Country Casuals Town/Rural Upscale Age 35-54 HH w/o Kids Mostly Owners College Grad Management W
The Cosmopolitans Urban Midscale Age 55+ Mostly w/o Kids Homeowners Some College WC, Mix W, B, A, H, Mix
Middleburg Managers Second City Midscale Age 45-64 HH w/o Kids Mostly Owners Some College WC, Mix W
Traditional Times Town/Rural Upper-Mid Age 55+ HH w/o Kids Mostly Owners Some College WC, Mix W
American Dreams Urban Midscale Age 35-54 Mostly w/ Kids Homeowners Some College WC, Srv, Mix W, B, A, H, Mix
Suburban Sprawl Suburban Midscale Age 35-54 HH w/o Kids Homeowners College Grad Professional W, B, A, Mix
New Homesteaders Town Upper-Mid Age 25-44 HH w/ Kids Mostly Owners College Grad BC, Srv, Mix W
Big Sky Families Rural Upper-Mid Age 25-44 HH w/ Kids Mostly Owners Some College BC, Srv, Mix W
White Picket Fences Second City Midscale Age 25-44 HH w/ Kids Mix, Owners Some College BC, Srv, Mix W, B, A, H, Mix
Blue-Chip Blues Suburban Midscale Age 25-44 HH w/ Kids Mix, Owners Some College BC, Srv, Mix W, B, A, H, Mix
Mayberry-ville Town/Rural Upper-Mid Age 35-54 HH w/o Kids Mostly Owners H.S. Grad BC, Srv, Mix W
Domestic Duos Suburban Midscale Age 55+ Mostly w/o Kids Mostly Owners H.S. Grad WC, Mix W, Mix
8. 8
How to get from clusters to TV Nielsen Ratings and Channel dominance
While earlier slides identify where the brand is dominant given their current
advertising, the previous slide used PRIZM to define the total presence of home mortgage
by using the same characteristics that described that current business to estimate a total
Mortgage Market. Results came from a key informant that concluded 33 clusters were needed.
This resulted in 6 profiles that can be described from a segmentation perspective using
dominant clusters (Funded 1st and 2nd, Cancelled 1st and 2nd , Dead Leads, and Target 33 )
These profiles were then used to provide TV viewing behavior across 80 cable channels by
the four major dayparts (early morning, day time, early fringe, and prime time) *. For each
channel Nielsen provides a key audience metric that counts the total eyeball watching at the
midpoint of every 15 minute interval within that daypart. These numbers were summed from
the first quarter 2007.
Eight unique performance indexes were created to rate each channel and then quantitatively
pick out those whose audience is best suited for our 1st mortgage product. Creatively using
these indicator maximizes our pull and dominance while minimizing cancels and dead leads
* ON and LN are currently available for Target 33 and dead lead only
9. 9
Performance Index Definitions by Daypart
Index33 = (Channel Minute 33 / Total Minute33) /
(Total Minute Daypart / Grand Total Minute Daypart)
This used a definitional set of clusters to calculate Total Homeowner Market Potential
FF_Index =(Channel Funded 1st Minutes / Total Funded First Minutes) /
(Total Minute Daypart / Grand Total Minute Daypart)
This used PRIZM matches against 26,437 Funded Firsts in 2006 that index 135+
FS_Index =(Channel Funded 2nd Minutes / Total Funded 2nd Minutes) /
(Total Minute Daypart / Grand Total Minute Daypart)
This used PRIZM matches against 55,229 Funded 2nd in 2006 that index 135+
CF_Index =(Channel Cancel First Minutes / Total Cancel First Minutes) /
(Total Minute Daypart / Grand Total Minute Daypart)
This used PRIZM matches against 237,000 cancelled Firsts in 2006 that index 135+
CS_Index =(Channel Cancel 2nd Minutes / Total Cancel 2nd Minutes) /
(Total Minute Daypart / Grand Total Minute Daypart)
This used PRIZM matches against 162,000 cancelled 2nd’s in 2006 that index 135+
10. 10
Performance Index Definitions by Daypart
Deadlead _Index =(Channel Deadlead Minutes / Total Deadlead Minutes) /
(Total Minute Daypart / Grand Total Minute Daypart)
Using a sample of 100,000 phone numbers of one time callers that never called
back, a reverse address append resulted in 45,000 PRIZM coded records
Pull_Index = (Channel Minute 33 / Channel Total Minute) / Mean Pull
This index measures the percent of total eyeball minutes seen by Target 33’s indexed to
the grand average percent for that daypart. The higher the index then more minutes of
that channel are being seen by our target 33 profile
pull=target_33_minutes/total_minutes;
pull_index=pull / mean pull for that daypart
Dom_index = (Channel Target33 minutes / Grand Total Target33 Minutes) / Mean Dominance
This index measures the dominance in total minutes of each channel indexed to the
average dominance for that daypart
dominance=total target33_minutes by channel / grand total target33 minute
dom_index=dominance / mean dominance for that daypart
11. 11
Comparison to March 2007 Early Morning Spot Picks
Top 10 Quantitative Selections in BLUE
Channel COUNT PERCENT Channel COUNT PERCENT
FIT 171 5.6195 HI 36 1.1830
FCS 162 5.3237 GAMC 34 1.1173
CW+ 134 4.4035 FXSC 33 1.0845
MSNB 131 4.3050 BBCA 32 1.0516
FX 116 3.8120 SLEU 32 1.0516
A&E 102 3.3520 ESPN2 30 0.9859
TOC 98 3.2205 SOAP 28 0.9201
STYL 97 3.1876 DIY 27 0.8873
HGTV 94 3.0891 BETJ 25 0.8216
TMC 93 3.0562 DISM 23 0.7558
CNN 91 2.9905 WGNS 23 0.7558
OXY 91 2.9905 ESPC 20 0.6572
HLN 86 2.8262 GAME 20 0.6572
HIST 85 2.7933 FOOD 19 0.6244
DSCH 82 2.6947 BIO 17 0.5587
FNEW 82 2.6947 TNT 16 0.5258
HOM 79 2.5961 TWC 16 0.5258
SCFI 78 2.5633 FXRE 12 0.3943
TBS 73 2.3989 ANPL 11 0.3615
HELV 72 2.3661 DSKD 8 0.2629
TTC 69 2.2675 TRAV 6 0.1972
CRNT 67 2.2018 HALL 5 0.1643
OUTD 61 2.0046 USA 5 0.1643
TVL 51 1.6760 DSCN 4 0.1314
CNBC 48 1.5774 DISC 2 0.0657
CMT 41 1.3474 DIST 2 0.0657
CSTV 41 1.3474 FINE 2 0.0657
AMC 40 1.3145 FSN 2 0.0657
BRAV 39 1.2816 G4 2 0.0657
NBA 38 1.2488 SPKE 2 0.0657
ESNW 36 1.1830 BFC 1 0.0329
This page shows what the
Past purchase plan was for
March 2007. The top 10
Quantitaive selections are
highlighted in BLUE. Comments
are that this appears to be
Sub-optimal and fragmented.
Note the top select for this
Daypart BBCA is number 35
and was purchased at only 1%.
Additional dayparts were also
done but not included.
12. Using Jan-March 2007, Grand Total Spots per month rounds to 20,000
March07 19,422 spots were distributed over various day parts as follows:
EM= 15.6% DT=37% EF=32.45% PR=7.96% LN=3.9% ON=2.9%
EM= 3,043 DT=7,188 EF=6,303 PR=1,546 LN=773 ON=569
Using the PRIZM coded Nielsen Reports distributions, suggests this pattern
EM= 9.5% DT=30.5% EF=25.7% PR=34.0%
EM= 1,845 DT=5,923 EF=4,991 PR=6,603
The following pages detail each channel’s specific profile using
8 unique performance indexes that will weight us heavier into PRIME Daypart
AND increase our funded 1st portfolio by focusing our AD buy in targeted spots
Why we should shift our spending to PRIME Daypart
12
13. 13
How to create a pick list from the Master file
• There are many ways to pick from this master list (see attached Updated Master
.xls) and some of the indexes are to be included (target33, FF, FS) while others are
to be excluded (deadlead, CF and CS). In addition two other indexes profile both the
pull or reach of a station to our target 33 profile as well as the dominance of that
channel to draw total eyeball minutes within that daypart.
• Cost information is only available for channels that were previously purchased. A
linear program was executed by the team to assign an optimal number of spots per
daypart (see attached LP solution). One issue with this approach is that fact that NO
major shift to Prime Time is delivered
• Using a factoring approach that assigned a statistical distance between channels was
calculated. Two composite scores were created – one to maximize our pull of
mortgage applicants, then second weighted to maximize FUNDED applicants.The
idea is choose channels with HIGH Target 33 AND FF profiles, but LOW cancel first
and LOW deadlead profile while pulling HIGH and being dominant with that daypart