Nielsen PRIZM Segment Narratives - DonorScape

2y ago
205 Views
37 Downloads
1.11 MB
49 Pages
Last View : 2d ago
Last Download : 10d ago
Upload by : Callan Shouse
Transcription

Nielsen PRIZMSegmentNarrativesDecember 2011

PRIZM, P YCLE, and ConneXions are registered trademarks of The Nielsen Company(US), LLC. Nielsen and the Nielsen logo are trademarks or registered trademarks ofCZT/ACN Trademarks, L.L.C. Other company names and product names aretrademarks or registered trademarks of their respective companies and are herebyacknowledged.This documentation contains proprietary information of The Nielsen Company.Publication, disclosure, copying, or distribution of this document or any of its contentsis prohibited, unless consent has been obtained from The Nielsen Company.Some of the data in this document is for illustrative purposes only and may not containor reflect the actual data and/or information provided by Nielsen to its clients.Copyright 2011 The Nielsen Company. All rights reserved.

ContentsIntroduction . 1Overview .1Data Description .2Methodology .3New Statistical Techniques .3New Data Sources.5Nielsen Urbanization Measures .5Demographic Predictors .6Nielsen PRIZM Components .6Contact Information .8Nielsen PRIZM Demographics . 9Interpreting Nielsen PRIZM Demographics .9Predominant Urbanization Class .10Predominant Household Income Class .10Predominant Age Range .11Predominant Age Class .11Predominant Household Composition Class .12Predominant Tenure Class .12Predominant Education Class .12Predominant Employment Class .13Predominant Race & Ethnicity Range .13Predominant Race & Ethnicity Class .14Predominant Income Producing Assets Class .14Nielsen PRIZM Social Groups . 15U1: Urban Uptown .16U2: Midtown Mix .18U3: Urban Cores .19S1: Elite Suburbs .21S2: The Affluentials .22S3: Middleburbs .24S4: Inner Suburbs .26C1: Second City Society .27C2: City Centers .28C3: Micro-City Blues .30T1: Landed Gentry .32T2: Country Comfort .34T3: Middle America .36T4: Rustic Living .38i

Nielsen PRIZM Lifestage Groups . 40Younger Years .41Y1: Midlife Success .41Y2: Young Achievers .41Y3: Striving Singles .42Family Life .42F1: Accumulated Wealth .42F2: Young Accumulators .43F3: Mainstream Families .43F4 - Sustaining Families .43Mature Years .44M1: Affluent Empty Nests .44M2: Conservative Classics.44M3: Cautious Couples .45M4 - Sustaining Seniors.45ii

IntroductionOverviewNielsen PRIZM offers a seamless transition between household-level segmentationand traditional geodemographics by providing the same segments to analyze at allgeographic levels. Having the ability to downshift from geodemographic tohousehold-level targeting makes it possible for marketers to move seamlessly frommarket planning and media strategy to customer acquisition, cross-selling, andretention.PRIZM classifies every U.S. household into one of 66 consumer segments based onthe purchasing preferences of a household. PRIZM offers a complete set of ancillarydatabases and links to partner data, allowing marketers to use data outside of theirown customer files to pinpoint products and services that their best customers aremost likely to use as well as locate their best customers on the ground. PRIZMenables marketers to create a complete portrait of their customers by answering theseimportant questions: Who are my targets? What are they like? Where can I find them? How can I reach them?PRIZM external links allow for company-wide integration of a single customerconcept. Beyond coding customer records for consumer targeting applications,Nielsen provides estimates of markets and trade areas for site analytics and profiledatabases for behaviors ranging from leisure time preferences to shopping to eatingto favorite magazines and TV shows, all of which can help craft advertisingmessaging and media strategy. Components of the PRIZM system can be grouped bythe stage of customer analysis, as shown in the following table.1

Customer Analysis StagePRIZM Component UsedCoding customer recordsHousehold-level codingGeodemographic codingFill-in for uncoded recordsComparing coded customer records totrade areaCurrent-year segment distributionFive-year segment distributionWorkplace segment distributionDetermining segment characteristicsand behaviorsNeighborhood demographic profilesHousehold demographic profilesNielsen Financial Track profilesNielsen Insurance Track profilesNielsen Convergence Audit profilesMediamark Research Inc (MRI)profilesR.L Polk profilesSimmons profilesAdditional profiles as created by theNielsen Link Partner NetworkData DescriptionIn developing PRIZM, Nielsen assembled a database that included more than890,000 household records from sources that include the proprietary NielsenFinancial Track, Nielsen Insurance Track, and Nielsen Convergence Audit surveys,the GfK Mediamark Research & Intelligence Survey of the American Consumer,R.L. Polk’s vehicle registration database, and the Nielsen Homescan consumerpackaged goods panel. Each of these records included demographic and behavioralmeasures, and the behavioral data included measures of both penetration andvolume. For example, data is available not only about whether a household owns amutual fund (penetration) but also about the value of the mutual fund (volume).Most important, every record in the file had demographic data reported by thesurvey respondents themselves. This database was regarded as an unprecedentedbenchmark for other data sources, including the compiled list data that wouldultimately be used to append PRIZM to customer records.When implementing PRIZM on partner files, segment assignments depend on thepartner compiled list data. The unique models built for each partner are designed toproduce a distribution of assignments that mirrors the distribution produced by theNielsen MultiSource Aggregation and Distributional Alignment (MADA) process.MADA is a proprietary methodology for assessing national distributions, whichbegins with data from the annual Nielsen demographic update and is informed byadditional data from the Nielsen Financial Track survey, Epsilon Targeting, Valassis(formerly ADVO), infoUSA, TARGUSinfo, and TomTom . This combination ofdata sources provides Nielsen a unique competitive advantage in its segmentationassignment methodology, due to the unparalleled breadth and depth of address-levelinformation. By combining data from multiple vendors with data from the annualdemographic update Nielsen can make PRIZM single assignments at the ZIP 6,2

ZIP 4, block group, and ZIP Code levels, allowing better fill-in for records that donot get a household-level assignment.MethodologyPRIZM culminated two years of research and development in a groundbreakingmethodology that allows marketers to seamlessly shift from the ZIP Code levelthrough the block group, ZIP 4, or ZIP 6 level—all the way down to the individualhousehold level—using the same set of 66 segments. This single set of segmentsaffords marketers the benefits of household-level precision in applications such asdirect mail, while at the same time maintaining the broad market linkages, usability,and cost-effectiveness of geodemographics for applications such as market sizingand site selection.New Statistical TechniquesIn 1980 and 1990, Nielsen statisticians rebuilt the original PRIZM by essentiallyrepeating the same steps they performed when Nielsen pioneered geodemographicsegmentation in 1976. They aggressively analyzed the data, isolated key factors, anddeveloped a new clustering system. The development of each new system providedan opportunity to evaluate and implement improvements as they became available,but the underlying segmentation technique was clustering.Since the 1970s, the most popular of the clustering techniques has been K-meansclustering. The final number of clusters desired is specified to the algorithm (this isthe origin of the “K” in K-means) and the algorithm then partitions the observationsinto K-number of clusters as determined by their location in n-dimensional space, asdictated by demographic factors. Membership in a cluster is determined by theproximity to the group center, or mean, in space (hence the origin of the “mean” inK-means).For any type of clustering process to work well, the statistician must correctlyidentify the important dimensions before implementing the clustering process. Formarketing purposes, obvious drivers are age and income. However, appropriatelevels for each of these critically important dimensions still need to be chosen. Forexample, does the dimension of income create better differentiation at 35,000 or 50,000? How does choosing between these two values of the same dimensionchange the clustering outcome? These choices are important, because when theclustering iterations end and yield an answer, marketers are left with clusters ofhouseholds that have been organized by their proximity to each other by thedemographic metrics that were chosen. This answer may or may not be meaningfulto the original task of creating groups that differ in their behaviors—in large partbecause behavior measures were not incorporated into the clustering technique itself.With PRIZM, Nielsen broke with traditional clustering algorithms to embrace a newtechnology that yields better segmentation results. PRIZM was created by aproprietary method developed by Nielsen statisticians called Multivariate DivisivePartitioning (MDP). MDP borrows and extends a tree partitioning method thatcreates the segments based on demographics that matter most to households’behaviors.The most common tree partitioning technique, Classification and Regression Trees(CART), involves a more modeling-oriented process than clustering. Describedsimply, statisticians begin with a single behavior they wish to predict and start withall participating households in a single segment. Predictor variables, such as income,age, or presence of children, are analyzed to find the variable—and the appropriatevalue of that variable—that divides the single segment into two that have the greatest3

difference for that behavior. Additional splitting takes place until all effective splitshave been made or the size of the segment created falls below a target threshold.In the example that follows, the CART process starts with all of the surveyrespondents in one segment for the behavior of interest—in this case, owning mutualfunds. Of this particular respondent pool, 10% report owning mutual funds. Next,the CART routine searches for the demographic variable—and the value of thatdemographic variable—that creates the two segments that are most different on thebehavior of interest. Our example shows that dividing the first group by an incomeof 50,000 yields two segments—one with mutual fund use of 3% and the secondwith mutual fund use of 18%. We can divide the second segment again, with theresult that a split based on an age of 45 yields two more segments—one with mutualfund use of 30% and the other with mutual fund use of 12%.If the process stops here, we have a segmentation system with three segments—onewith 3% of its members owning mutual funds, a second with 12% of its membersowning mutual funds, and the third with 30% of members owning mutual funds.However, this resulting segmentation system does not provide useful informationabout any other behaviors—it’s optimized only for owning mutual funds. This is oneof the limitations of the CART technique: it generates an optimal model for only asingle behavior. Because PRIZM is a multi-purpose segmentation system,optimization across a broader range of behaviors is necessary. Nielsen made severalmodifications to the CART process, resulting in the proprietary MDP technique.These modifications extended the basic CART process to simultaneously optimizeacross hundreds of distinct behaviors at once. This advancement allowed Nielsen totake full advantage of the nearly 10,000 behaviors and hundreds of demographicpredictor variables available at different geographic levels, including the householdlevel. The MDP process was run hundreds of times, with varying sets of behaviors,predictor variables, and a number of other parameters, to ensure that the resultingsegments represent behaviorally important groupings.4

New Data SourcesIn addition to a unique statistical technique, Nielsen employed an unprecedentednumber of data sources and data levels in the development of Nielsen PRIZM.Geodemographic data, the mainstay of previous segmentation systems, includedCensus 2000 demographics and ZIP 4-level demographics summarized fromcompiled lists.For the first time, Nielsen also used household-level demographics in thedevelopment process of PRIZM. To each of the 890,000 customer records in thedevelopment database already coded with Census 2000 demographics, summarizedZIP 4 demographics, and custom Nielsen measures, Nielsen appended compiledlist—household—demographics from the Epsilon Targeting TotalSource Plus file.The resulting database was used to design and evaluate systems built with fourdifferent sources of data: Self-reported household, compiled list-based household,ZIP 4, and block group.Nielsen Urbanization MeasuresA distinctive feature of PRIZM is the Nielsen urbanization model. Multiplerefinements to the urbanization model for the latest PRIZM release allow it toprovide a better contextual framework than earlier models had. The result of theseimprovements was the identification of five distinct urbanization classes; however,PRIZM development showed optimal performance by using the following fourclasses:Urban areas (U) have population density scores (based ondensity centiles) mostly between 85 and 99. They include boththe downtowns of major cities and surrounding neighborhoods.Households within this classification live within the classichigh-density neighborhoods found in the heart of America’slargest cities. While almost always anchored by the downtowncentral business district, these areas often extend beyond citylimits and into surrounding jurisdictions to encompass most of America’s earliestsuburban expansions.Suburbs (S) have population density scores between 40 and 90,and are clearly dependent on urban areas or second cities. Unlikesecond cities (defined below), they are not the population centerof their surrounding community, but rather a continuation of thedensity decline from the city center. While some suburbs may beemployment centers, their lifestyles and commuting patterns willbe more tied to one another, or to the urban or second city core,than within themselves.Second Cities (C) are less densely populated than urban areas,with population density scores typically between 40 and 85.While similar to suburbs in their densities, second cities are thepopulation centers of their surrounding communities. As such,many are concentrated within America’s larger towns andsmaller cities. This class also includes thousands of satellitecities, which are higher density suburbs encircling majormetropolitan centers, typically with far greater affluence than their small citycousins.Town & Country areas, collapsed into a single urbanizationcategory (T), have population density scores under 40. Thiscategory includes exurbs, towns, farming communities, and a5

wide range of other rural areas. The town aspect of this class covers the thousands ofsmall towns and villages scattered throughout the rural heartland, as well as the lowdensity areas far beyond the outer beltways and suburban rings of America’s majormetros. Households in the exurban segments have slightly higher densities and aremore affluent than their rural neighbors.Demographic PredictorsThe resulting PRIZM model, selected for its consistently outstanding performancecompared to Nielsen’ benchmark systems, incorporates the Nielsen urbanizationmeasure; household characteristics, such as affluence, age, and family composition;and neighborhood characteristics, such as housing stock and home ownership.The 66 segments are numbered according to socioeconomic rank (which takes intoaccount characteristics such as income, education, occupation and home value) andare grouped into 11 lifestage groups and 14 social groups. Social groups are basedon urbanization and socioeconomic rank. Lifestage groups are based on age,socioeconomic rank, and the presence of children at home.Nielsen PRIZM ComponentsAs customers have come to expect from Nielsen, PRIZM users are afforded the fullcomplement of supporting data and documentation. From posters to profiles, fromdemographic detail to descriptions for each segment, PRIZM will continue to leadthe industry in helping marketers reach a deep understanding of their most importantcustomer groups.Among the current offerings for PRIZM:Household-Level Coding: Household-level is the new andpowerful extension of PRIZM—the same segment codes asthe geodemographic system that differentiate betweenhouseholds in the same neighborhood. Using the customer’sname and address, household-level demographic data isappended and used in an algorithm to determine thehousehold’s PRIZM code.Geodemographic Coding: PRIZM codes can be appended tonearly every address in the U.S. using the File Enhancementmodule of Nielsen MarketPlace Net, our desktop geocoderNielsen PrecisionCode, or a PRIZM directory file. Codescontinue to be easy to append, with nearly 100% coverage forgeographies ranging from ZIP 6—a near-household levelassignment pioneered by Nielsen—to ZIP 4 to block group toZIP.Directory License: Segmentation best serves marketers whenit provides an organizing framework for how the entirecompany thinks about its customers. To make PRIZM asaccessible—and cost-effective—as possible, thegeodemographic levels can be licensed for unlimited use bymeans of a directory flat file that can be delivered to yourtechnology users.6

Segment Distributions: With a coded customer file, you candetermine your own customer penetration using the PRIZMsegment distributions. The distributions provide counts “onthe ground,” by segment, for standard geographic levelsacross the entire U.S. Distributions are available for Country,State, DMA, CBSA, County, Tract, ZIP Code, and blockgroup (ZIP 4 is available only as a single assignment).Segment distributions are also useful for ranking markets andestimating demand.Profile Databases: What makes segmentation more than ascore is the means to assess what your customers are like,where they live and shop, and how best to reach them. Profiledatabases provide descriptive detail to create a fullythree-dimensional view of your best customers. Lifestyleprofiles are available for both Simmons and MRI; automotivedetail comes from Polk; and industry-specific behaviors arecaptured by Nielsen surveys—Nielsen Financial Track forfinancial behaviors, Nielsen Insurance Track for insurancedata, and Nielsen Convergence Audit for communications andenergy.Link Partners: Extending the legacy of PRIZM 62 andMicroVision, Nielsen intends to make PRIZM availableeverywhere. Our network of partner companies makesPRIZM available on virtually all industry-standard databases,direct marketing service bureaus and primary researchvendors. Marketers can utilize the power of PRIZMcontinuously from market research, analysis, and targeting,through media strategy and evaluation.Nielsen MyBestSegments: MyBestSegments.com provides therich demographic and behavioral detail to answer the question“What are my customers like?” MyBestSegments.com linksdirectly to the Nielsen ConsumerPoint datamart, which makesdata for PRIZM available online and makes updatesautomatic.Nielsen MarketPlace NET: ClaritasMarketPlace.com deliverssite, market, customer segmentation, and profiling analysessolutions tailored for different user types. This browser-basedsolution is ideal for casual users who are better suited to aneasy-to-use, Internet-based solution—a complete marketingtoolbox of products and resources in one convenient location.Perform site, segmentation, or profiling analyses; accessreports, maps, lists, and more. Additionally, NielsenMarketPlace NET can be upgraded to include NielseniMARK Online.7

Contact InformationNow that you understand how PRIZM was developed, we invite you to explore thebenefits of PRIZM for your own customer targeting applications. For moreinformation about profiling your customers using PRIZM, please contact youraccount team.8

Nielsen PRIZM DemographicsNielsen provides a series of demographic descriptors used to classify the segmentsacross core dimensions. While demographics form the basis for every segmentassignment, not every segment falls neatly into only one category for eachdemographic.Interpreting Nielsen PRIZM DemographicsThe following sections provide the broad range of predominant values for each ofthese demographic descriptors. These values were created with household-leveldemographics, which provide the greatest possible precision due to a purerdistribution. This is especially true for variables such as age and presence ofchildren, which vary the most between households living in the same neighborhoods.To further assist comprehension of these demographic descriptors, the set of rulesused to place each segment into one of the categories is provided below. Please notethat there are situations, as with Predominant Household Income Class, wheregroups are subdivided until a point is reached where the “least commondenominator” for the remaining segments is simply not anything else. The numbersin the order column denote the sequence in which the rules are applied, whichimpacts how a segment is classified. The numbers in the PNE column present thenumber of segments in each grouping.Finally, each segment is assigned a demographic caption, which covers the essentialdimensions of income, age, and family composition. For example, Segment 27,“Middleburg Managers,” has a demographic caption of “Midscale Older withoutKids.” The tables below provide the specific meaning of each descriptor.When reading the heuristics tables below, please keep some key thoughts in mind.First, the definitions of each category are hierarchical. The key is that the ranges arenon-exclusive, so the rules must be followed in a specific order for a segment to becategorized. For example, the rule “1 Age 35 where age 35 50% and notAge 25-44”should be read as “Segments with the descriptor “Age 35” aresegments where 50% or more of the households in the segment are under 35 yearsold, and the segment does not meet the definition of the “Age 25-44” range.”Further, the rule “3 Age 25-44 where age 25-44 65% and not age 35-54”should be read as “segments with the descriptor “Age 25-44” are segments where65% or more of the households in the segment are between 25 and 44 years old, andthe segment does not meet the definition of the “Age 35-54” range”.As you can see by looking at the definitions this way, the rule “7 Age 65 ” has noexceptions, it is simply assigned to segments where more than 70% of the populationis over 65. As you move through the ranges in the order in which they are created,you see that each range is defined as long as the previous range definition is not met.This is important because the segments designated using rule “1 Age 35” are notnecessarily under 24 years old. Rather, they just do not meet the definition of rule “3Age 25-44”.Finally please note the order in which heuristics are applied, as that also helpsdetermine the segment descriptor.9

Predominant Urbanization ClassUrbanization Class has eight categories from Urban to Mix. Urbanization is aprimary driver of Nielsen PRIZM and you will see consistent urbanization classdescriptors from year to year. Urbanization is not a primary driver behind NielsenP YCLE or Nielsen ConneXions which means descriptors may change annually forthose two segmentation systems.Segment Descriptor1 Urban2 Suburban3 Second City4 Town5 Rural6 Town/Rural7 Metro Mix8 MixHeuristicwhere Urban 100%where Suburban 100%where Second City 100%where Town 100%where Rural 100%where Town Rural 75%where Urban Suburban Second Cit

Nielsen MultiSource Aggregation and Distributional Alignment (MADA) process. MADA is a proprietary methodology for assessing national distributions, which begins with data from the annual Nielsen demographic update and is informed by additional data from the Nielsen

Related Documents:

PRIZM NE replaced the PRIZM system (now referred to as “PRIZM 62”) and MicroVision. PRIZM NE will continue to lead the industry in helping marketers reach a deep understanding of their most important customer groups. Among the current offerings for PRIZM NE: Household-level coding House

The VALS Segments 432 Geo-Lifestyle Analysis (NIELSEN PRIZM ) 434 PRIZM Social and Life Stage Groups 434 Sample PRIZM Segments 435 An Application of PRIZM 436 International Lifestyles 436 Summary 437 PART THREE CASES Case 3-1 Patagonia's Eco-Fashion Push 444 Case 3-2 Domino's Reform

In the PRIZM lifestyle segmentation system, every U.S. neighborhood is defined according to sixty-two distinct types or “clusters.” Each U.S. neighborhood is assigned to one of these PRIZM clusters according to the current year’s demographic projections. The 62 PRIZM clusters

Nielsen PRIZM Lifestage Groups PRIZM household segmentation –66 segments –Based on a variety of demographic, consumer behavior, media preferences, socioeconomic and urbanization attributes. Nielsen has condensed the 66 PRIZM codes down to 11 Lifestage gro

Card images are solely for the purpose of design display. Actual images used on cards to be determined. BASE GIRAFFE PRIZM TIER II BURGUNDY CRACKED ICE . PRIZM TIER III SHIMMER PRIZM. HOBBY EXCLUSIVE PARALLEL FOTL EXCLUSIVE PARALLEL. PRIZM. BASEBALL. 2021 MLB PLAYERS TRADING CARDS · 1ST OFF THE LINE

PRIZM Coding Applied to Survey Database Using Nielsen’s PRIZM classifications, we identified the five most common segments in the population of ISD 622’s voters. These 5 segments represent 45% of the electorate 22 ISD 622 Data Survey Results: Effect of Information on PRIZM Segments Able to ide

Sep 01, 2011 · segment bisector bisect In the Geo-Activity, M is called the midpoint of AB&*. The of a segment is the point on the segment that divides it into two congruent segments. A is a segment, ray, line, or plane that intersects a segment at its midpoint. To a segment means to divide the segment into two congruent segments. M is the midpoint of AB&*.

Korean and Chinese belong to different language families. In terms of their linguistic structures, they are extremely dissimilar. Beginning Korean: A Grammar Guide 2 Autumn 2004 Finally, hangeul is uniquely associated with the language, literature, and people of the Korean peninsula. No other community uses the hangeul system for graphically representing the sounds of their language. Given the .