TEAM provides estimates of the audience delivery of exterior Transit Advertising in local markets throughout the United States. The overall TEAM program was begun in the mid 1970's. TEAM V is the latest update/refinement.
For each Market/Showing combination, the standard TEAM report shows:
- Reach (in Percent and Number of Persons)
- Average Frequency
- Gross Rating Points and Total Exposures
- Exposure Frequency Distribution (in classes)
These audience statistics are reported for the Total Population 18+ and for selected demographic sub-groups defined by Sex, Age, Income, and Race.
TEAM is a statistical model. Estimates of the exterior Transit Advertising audience in each local market are based on:
- Characteristics of the Market
- Total Population and Demographic Composition
- Population Density
- Market Geographic Area
- Characteristics of the Transit Advertising Showing
- Number of Buses in the Showing
- Poster Placement (% of Posters placed on Left Sides, Right Sides, and Rears of Buses)
- Time-Period (One, Two, or Three Months)
The TEAM model is based on empirical research, in which Transit Advertising audiences in a sample of local markets were measured in respondent-based field studies.
Field Research Base
In the mid 1980's, the Transit Advertising Association sponsored field studies in a sample of five local markets. The sample markets (Chicago, Los Angeles, Memphis, St. Louis, and Washington, D.C.) were chosen to provide a reasonable cross-section by region, market size, and local Transit conditions.
In each market, randomly selected respondents were recruited to keep diaries of their exposure to exterior Transit Advertising (i.e., number and sides/ends of buses seen) over a 7-day period. Across the five markets combined, approximately 1,000 respondents completed 7-day Transit Advertising exposure diaries.
In so far as possible, the studies were designed to measure "actual" rather than "potential" advertising exposure. Respondents were instructed to record exposures only when a bus crossed their field of vision in the course of normal activity (i.e., no extra looking about), close enough that they could read the actual advertising placed on the bus.
All field studies were conducted by R.H. Bruskin, Inc.
Basic TEAM Model
There are three basic steps in the TEAM model, as follows:
- 1. Estimate Total Gross Rating Points
For any market, the first step in the TEAM model is to estimate the number of exposures the average bus in that market generates on the average day. This estimate is based on market-specific size and density, using a regression equation developed from the Bruskin study data. Not surprisingly, buses in larger, more densely populated markets produce a greater number of exposures than buses in smaller and less densely populated markets.
Separate exposure rates by poster location (e.g. Left Side, Right Side, and Rear Display) are derived by distributing total exposure rates to each poster location in the same proportion observed in the Bruskin studies.
Total Exposures and GRP for any Showing/Time-Period may now be gotten directly. For example, for a Showing that has posters on the Left Side of 50 buses for a Time-Period of 2 Months, Total Exposures is the product of:
- Exposures-per-bus-per-day for Left Sides, times
- 50 Buses in the Showing, times
- 60 Days in the 2 Month Time-Period.
Total GRP is Total Exposures times 100, divided by the Total Population of the Market.
- 2. Estimate Total Reach
Reach projections are made by the Gamma-Poisson model. This is the standard reach projection model for out-of-home media -- equivalent to the Beta-Binomial used in Print and Broadcast media -- and has been widely tested and validated.
For TEAM purposes, the Gamma-Poisson model has one parameter (parameter "a"), related to the variation of individual exposure rates. Given parameter "a", the Gamma-Poisson model provides the Reach (and the complete Frequency Distribution) for any GRP level.
The Gamma-Poisson parameter "a" is estimated for each TEAM market based on market-specific size and density, using a regression equation based on the Bruskin studies.
Note: Steps 1. and 2. completely specify all audience estimates for the Total Population 18+ in any TEAM market. Step 3. below deals with audience estimates for demographic sub-groups.
- 3. Demographic Groups
For each demographic group, factors were derived relating demographic-specific GRP and parameter "a" values to the corresponding Total Population values, based on Bruskin study data.
In each TEAM market, these factors are applied to the Total Population values estimated in Steps 1. and 2. to obtain GRP and Reach for each demographic group.
Comparability of TEAM V with Earlier Versions
The basic modeling approach has been the same throughout the TEAM program. Field studies in a sample of markets serve as a base for quantifying statistical relationships between Transit audiences and local market characteristics, and these relationships are then applied to the specific characteristics of each TEAM market to produce audience estimates.
Between TEAM II and III there was a change in the field base used for model building. TEAM I and TEAM II were based on field studies conducted in the mid and late 1960's; TEAM III, IV and V have been based on the more recent Bruskin studies.
Comparison of Bruskin with earlier study results shows remarkably little change in Transit audience over two decades. While the more recent studies show smaller differences between men and women (due to the increased incidence of working women), overall audience levels and demographic composition by age, income, and race are essentially unchanged. It appears that the individual choices and activities that lead to exterior Transit Advertising exposure -- presence on the streets and sidewalks along bus routes -- have been largely unaffected by the changes that have overtaken other media.
Apart from the change in field research base from TEAM II to III, changes from one TEAM version to the next have otherwise been cosmetic – mostly updates and refinements in the specific demographic sub-groups reported – with no change in the underlying model.