Nielsen Ad Intel™ Methodology by Medium

Nielsen Ad Intel™ Methodology by Medium

Nielsen Ad Intel™

2

7/28/2015

Table of Contents

Multimedia Coverage ...................................................................................................... 4

Network Television Methodology .................................................................................... 6

Network Clearance Television Methodology ................................................................... 8

Cable Television Methodology ........................................................................................ 9

Local Regional/Cable TV ............................................................................................... 11

Syndicated Television Methodology .............................................................................. 12

Syndicated Clearance Television Methodology ............................................................. 14

Spanish-Language Network Television Methodology .................................................... 15

Spanish-Language Cable Television Methodology ....................................................... 17

Spot Television Methodology ........................................................................................ 19

Network Radio Methodology ......................................................................................... 21

Local Radio ................................................................................................................... 22

National Magazine Methodology ................................................................................... 24

Local Magazine Methodology ........................................................................................ 26

National Newspaper Methodology ................................................................................ 27

National Sunday Supplement Methodology .................................................................. 28

Local Newspaper Methodology ..................................................................................... 29

Local Sunday Supplements Methodology ..................................................................... 31

Free Standing Inserts (FSI) Methodology ...................................................................... 32

Outdoor Methodology .................................................................................................... 33

National Internet Methodology ........................................ Error! Bookmark not defined.

Local Internet Methodology ........................................................................................... 36

Business-to-Business Publication Methodology ............................................................ 38

National Cinema ............................................................................................................ 39

Nielsen Ad Intel™

3

2/15/2017 /2015

Regional Cinema ........................................................................................................... 40

Digital ............................................................................................................................ 41

Nielsen Ad Intel™

4

7/28/2015

Multimedia Coverage

1. NETWORK TV Number of Networks 9 (ABC, NBC, CBS, FOX since 1/93; UPN & WB 1/95-9/06; PAX/ION since

2/99; MNT 9/5/06-10/28/09; CW since 9/18/06, ME-TV & Bounce-TV 4/1/13, COZI 1/26/15)

Data Source Nielsen

2. CABLE TV Number of Networks 100 networks Data Source Nielsen

3. SYNDICATION TV Barter Programs per month 200+ (since 1/93) Data Source Nielsen

4. SPANISH LANGUAGE NETWORK TV Number of Networks 6 Networks Data Source Nielsen

5. SPANISH LANGUAGE CABLE TV Number of Networks 7 Networks Data Source Nielsen

6. SPOT TV Number of Markets All 210 DMAs (since 9/25/00) registered commercials; 27 Hispanic Markets)

Top 75 DMAs (since 1/97); 50 Top DMAs (since 1/93) Data Source Nielsen

7. LOCAL REGIONAL/CABLE TV Number of Markets 51 DMAs Data Source Nielsen

8. NETWORK RADIO Number of Networks 2 Networks (Premier and WestwoodOne) Data Source Radio Networks

9. LOCAL RADIO Number of Markets 43 DMAs Data Source MMI Rating Source Nielsen Audio

10. NATIONAL MAGAZINE Number of Magazines 200+ Magazines (since 1/96) Data Source Nielsen Rating Source Mediamark Research Inc.

11. LOCAL MAGAZINE Number of Magazines 72 Number of Markets 42 Top DMAs (since 1/95) Data Source Nielsen

12. NATIONAL NEWSPAPER Number of Newspapers 3 (WSJ, USA Today, New York Times National Edition since 1/94) Data Source Nielsen

13. LOCAL NEWSPAPER Number of Local Newspapers 188 Number of Markets 75 Top DMAs (since 1/94) Data Source Nielsen

14. NATIONAL SUNDAY SUPPLEMENTS Number of Sunday Magazines 6 Data Source Nielsen

15. LOCAL SUNDAY SUPPLEMENTS Number of Local Sunday Supplements 6 Number of Markets Within 50 Top DMAs (since 1/94) Data Source Nielsen

16. FREE STANDING INSERTS (FSIs) Number of Markets 67 DMAs (since 1/93) Data Source Nielsen

17. OUTDOOR Number of Markets 180+ DMAs Data Source Plant Operators

Nielsen Ad Intel™

5

2/15/2017 /2015

18. NATIONAL INTERNET

Number of Companies Beginning with January 2006 data all advertisers are available. Prior to 2006 the data set was; the Fortune 1000 companies, Top 500 traditional media spenders, the Top 75 on-line spenders, the largest online publishers, and the top e-commerce sites (since 1/02); Fortune 500 companies (since 1/01)

Data Source Nielsen Online

19. LOCAL INTERNET Number of Markets 79 DMAs Data Source Nielsen Online

20. BUSINESS TO BUSINESS Number of Magazines 500+ Magazines Data Source IMS, Canada

21. NATIONAL CINEMA Coverage On-Screen and Slide Campaigns Data Source Cinema Companies

22. REGIONAL CINEMA Coverage On-Screen and Slide Campaigns Data Source Cinema Companies

23. Digital Coverage Available as of 12//28/2015. Crawlers visit ~5,000 US sites, which make up an

estimated 85% of all web traffic (US). Reported activity includes Desktop, Mobile Web & Tablet Web Display placements as well as Desktop Pre-roll & In-banner Video ads.

Data Source Powered with data captured by Pathmatics

Nielsen Ad Intel™

6

7/28/2015

Network Television Methodology

Commercial Occurrences

Nielsen Ad Intel utilizes a network of computerized monitoring sites to collect and identify commercial occurrence information. Using proprietary hardware and software designed specifically for this purpose, including our patented “passive pattern recognition” technology, the service has developed the capability to identify commercials based on their unique audio and video content.

Occurrence data collected from the monitoring sites is appended to other non-occurrence information, including ratings and expenditure data. All first-time seen broadcast data that has not been automatically identified through passive pattern recognition technology is processed by commercial analysts to assign brand identification and commercial duration.

A commercial is defined as any paid advertisement announcement, spot or message with duration of 10 to 120 seconds, aired by the television network and recorded by Nielsen Ad Intel. Monitoring is conducted 24 hours a day.

Reported advertising activity includes commercials aired during network telecast programs and regional telecasts. Regional commercials are identified using network broadcast log information when available.

Expenditure Data

Expenditure data displayed for commercial activity consists of Program or Daypart costs provided by the monitored networks wherever possible. Where broadcast data is unavailable, expenditure information is supplemented by other industry sources and may include data derived from proprietary models.

Based on input from market sources, expenditures for direct response commercials are computed at 50 percent of estimated expenditure rates for all other commercial types, to more closely represent their actual purchase price.

Commercial occurrences are pro-rated based on their relationship to a thirty second spot. The pro-ration is a straight pro-ration.

Gross Rating Points

The Nielsen National service provides program ratings data from the national People Meter panel on a weekly basis. Program averages (Average Audiences –based on average minute rating) are reported on an individual program basis although weekly multi-telecast program averages are used when individual telecasts are not available. The program rating is applied to each commercial within the program regardless of duration, pod, or position. In the case of regional commercials, the rating is prorated by the percentage of geographic coverage, which is provided by the networks.

The minimum reportable rating (before rounding) is .05.

Live = Household tuning or persons viewing to a program, quarter-hour or time period to a specific station or cable channel, including VCR record.

Live + SD = Households tuning or persons viewing to a program, quarter-hour or time period to a specific station or cable channel, including VCR record, and any DVR playback tuning/viewing in the same processing day.

Live + 7 = Households tuning or persons viewing to a program, quarter-hour or time period to a specific station or cable channel, including VCR record, and any DVR playback tuning/viewing within 7 days.

C3 = the term commonly used in the industry to describe ACM ratings that include time-shifted playback viewing within 3 days (75 hours) of the live program telecasts. ACM estimates represent average viewing levels to all minutes within a program that contain national commercial time.

Nielsen Ad Intel™

7

2/15/2017 /2015

ACM estimates are not produced for any programs that don’t have national commercial time within them. Promos and PSAs are excluded from ACM estimates.

Nielsen Ad Intel makes every effort to provide the most complete and accurate data possible, but does not warrant that the data will be complete or error free. Nielsen Ad Intel may experience some non- sampling errors that affect our ability to capture complete data. These situations include; remote site operational interruptions, the accuracy of our automated matching process, the quality of our manual & automated data processing & the accuracy of commercial logs provided by content distributors.

Equivalized TV Units, Impressions and GRPs Much like our existing CPME and CPPE, these new TV measures options will take into account the length of the commercial in comparison to a standard of 30 seconds. Although Nielsen does not endorse special formulations for GRPs and Imps, our clients are accustomed to analyzing competitive data in this manner.

The calculations are as follows:

120 second commercial - multiply by 4

60 second commercial - multiply by 2 (e.g., 100,000 impressions would equal 200,000 for a 60 second commercial)

45 second commercial - multiply by 1.5

30 second commercial - multiply by 1 (i.e., no adjustment)

20 second commercial - multiply by .67


10 second commercial - multiply by .33 The new measures will be available for selection in all summary and detail reports for TV except the Flowchart Link since it requires vendor action. Equivalized units will round to one decimal place.

Nielsen Ad Intel™

8

7/28/2015

Network Clearance Television Methodology

Within each market, Nielsen Ad Intel distinguishes between local spots and network clearances. For each network program, commercials from the local market are compared to commercials in the national database, thus identifying the local TV stations that did and did not air the “national” commercial, and flagged accordingly. Data is available from 5 am – 2 am.

Gross Rating Points

Commercial occurrences are linked to household and persons ratings from Nielsen’s Local service Viewers in Profile (VIP) ratings by report book month, station, day of week and quarter hour. Note, in metered markets, individual quarter-hour “metered” ratings are available for households. The VIP sweep data is available for all markets for the November, February, May and July diary panels. Data for October, January and March are also available in the appropriate markets.

Because of the timeliness of Nielsen Ad Intel data, commercial occurrences are available prior to the completion of the diary month. Therefore, ratings are estimated by applying the most recent sweep period (station share and market universe estimate) to the DMA HUT and PUT viewing levels of the prior year’s sweep period. The November estimate uses the May station share rather than July.

Actual VIP sweep data is processed and data is made available in Nielsen Ad Intel as early as two weeks after the measurement interval.

The minimum reportable rating (before rounding) for the market is .05.

Expenditure Data

Expenditures are not assigned to local market network clearances.










Nielsen Ad Intel™

9

2/15/2017 /2015

Cable Television Methodology


Nielsen Ad Intel utilizes a network of computerized monitoring sites to collect and identify program occurrence information. Using proprietary hardware and software designed specifically for this purpose, including our patented “passive pattern recognition” technology, the service has developed the capability to identify commercials based on their unique audio and video content.

Occurrence data collected from the monitoring sites is appended to other non-occurrence information, including ratings and expenditure data. All first-time seen broadcast data that has not been automatically identified through passive pattern recognition is processed to assign brand identification and commercial duration.

A commercial is defined as any paid advertisement announcement, spot or message with duration of 10 to 120 seconds, aired by the television station and recorded by Nielsen Ad Intel. Advertising activity is based on nationally distributed commercials aired during cable telecast programs. Monitoring is conducted 24 hours a day.

Paid Programming is excluded from reporting for this media type.

Expenditure Data

The source for nearly all Nielsen Ad Intel National Cable expenditure data is derived from SQAD, Inc.’s NetCosts system. The CPP data from NetCosts that Nielsen receives is created from actual purchases made by NetCosts contributors and reflects the combined costs based on their true weighting of upfront and scatter buys and is considered “gross.”

SQAD provides four types of CPPs for use in the calculation of Nielsen Ad Intel National Cable expenditures: a monthly averaged TVHHLD network specific by daypart CPP, a monthly averaged TVHHLD like-network by daypart CPP, a monthly averaged TVHHLD all sports network by daypart live sporting event CPP, and a monthly averaged TVHHLD general network by daypart live sporting event CPP

To compute a 30 second cost, Nielsen multiplies the appropriate SQAD provided monthly averaged daypart TVHHLD CPP by the monthly averaged TVHHLD Live stream rating of all telecasts of an individual program based on program start time. The result is a 30 second cost which is then pro-rated by commercial occurrence duration. A straight proration is applied. Direct response advertising is based off the same NetCosts CPP as a full charge unit; however, after the initial costing calculations, it is reduced by 50%.

Nielsen Ad Intel will continue to calculate expenditures for selected cable distributors - including ESPN De Portes (DSPA), Fox Sports En Espanol (FSE), Galavision (GALA), and MUN2 (MUN2) - using the previous methodology that relies on cable network participation in the form of monthly rate cards by program, daypart or a combination of both. For GOL and ESPD who do not supply us with ratecards the NetCosts Sports Category CPPs (Non-Live) is used.

Gross Rating Points (GRPs)

An individual telecast rating is used where available. If a telecast rating is not present, the Monday-Friday multi-day average will be used. If neither a telecast rating nor a Monday-Friday rating is available, the Monday-Sunday average will be used. This methodology applies to 2007 data forward.

Nielsen Ad Intel™

10

7/28/2015

If a network is not reportable at the program level, Nielsen Ad Intel will report half-hour time period ratings for the individual day of week where the spot aired. If individual day averages are not available, a Monday- Friday multi-day average will be reported. If neither an individual day rating nor a Monday-Friday rating is available, the Monday-Sunday average will be used.

Prior to 2007 the following methodology was applied; The Nielsen National service provides both program and half hour ratings data estimated from the national People Meter panel on a weekly basis. In order of preference, Nielsen Ad Intel utilizes program ratings based on: Monday-Sunday averages across all dayparts, Monday-Friday averages across all dayparts, individual telecast, and half-hour averages. Average audience estimates are based on the total national universe without regard to Cable coverage.

The program rating that is applied to each commercial is not adjusted according to duration.





C3 = the term commonly used in the industry to describe ACM ratings that include time-shifted playback viewing within 3 days (75 hours) of the live program telecasts. ACM estimates represent average viewing levels to all minutes within a program that contain national commercial time. ACM estimates are not produced for any programs that don’t have national commercial time within them. Promos and PSAs are excluded from ACM estimates.











Nielsen Ad Intel™

11

2/15/2017 /2015

Local Regional/Cable TV

Monitoring Criteria & Methodology

The Nielsen Ad Intel Local Cable Service measures local cable activity originating in select markets across the country. The data collected represent a measurement of local insertions identified between 5a and 2a on the specified cable networks within the measured headend of the selected interconnect in the market. Given the complexity of local cable advertising, including the multiple zoning options available in many markets, the Nielsen Ad Intel Local Cable Service is designed to deliver competitive insights on local cable advertising in the measured markets. This service should not be used to perform an audit of all insertions within the medium, or identify clearances of national advertisements.

The Nielsen Ad Intel Local Cable Service methodology consists of Full Discovery Market (FDM) and Automated Discovery Market (ADM) towers located within each DMA. Utilization of the two tower approach is to aid in the identification of interconnect-wide versus zoned commercial occurrences. The ADM tower is located where it can measure a different head-end for the same interconnect to guide users on zoned versus interconnect-wide advertising.

For National and Local Origination cable networks, occurrences captured on both the FDM and ADM towers at the same time, on the same network, have the potential to be interconnect-wide advertisements. These occurrences are tagged with an "I" indicator. For viewing sources classified as regional cable networks, occurrences appearing on both the FDM and ADM towers are identified as regional ads (R). For all networks, occurrences captured on the FDM tower at the time a different commercial is captured on the ADM tower are identified as zoned advertisements (Z).

The grand total of occurrences is defined as the total number of advertisements captured by the FDM tower only (zoned ads) plus the total number of advertisements captured on both the FDM and ADM tower (interconnect wide ads and regional ads).


Nielsen Ad Intel reports the commercial which aired on either the FDM tower only or that which aired on the FDM/ADM tower. Nielsen Ad Intel does not report those commercials airing solely on the ADM tower. The ADM tower is used only to provide insight as to whether a particular advertisement aired across zones within a particular DMA.

The Nielsen Ad Intel Local Cable Service does not provide expenditures or ratings associated with local cable insertions.

Nielsen Ad Intel makes every effort to provide the most complete and accurate data possible, but does not warrant that the data will be complete or error-free.

Nielsen Ad Intel™

12

7/28/2015

Syndicated Television Methodology



Occurrence data collected from the 12 transponders is appended to other non-occurrence information, including ratings and expenditure data. All first-time seen broadcast data that has not been automatically identified through passive pattern recognition is processed to assign brand identification and commercial duration.

A commercial is defined as any paid advertisement announcement, spot or message with duration of 10 to 120 seconds, aired by the television station and recorded by Nielsen Ad Intel. Advertising activity is based on commercials sold by the syndicator to air during bartered telecast programs. Monitoring is conducted 24 hours a day.

Expenditure Data

Effective with September 2006 data, expenditure information reported for National Syndication commercial activity is derived from data supplied by SQAD’s NETCOSTS system which reflect CPPs created from actual transaction level buys. Nielsen Ad Intel receives NETCOSTS Syndication CPPs each month. These CPPs are provided by Syndicator, Program Type and Program Category. For example: 20th Television - Entertainment - Drama Action. Nielsen Ad Intel makes every effort to accurately map Nielsen Ad Intel Syndication programs to SQAD program types and program categories.

NETCOSTS Syndication CPPs are then multiplied by the monthly average NSS TV Household Live stream program rating for each program which produces that program’s 30 second cost. In rare occasions where specific Syndicator/Program Category CPPs are not available, an average CPP by Program Type is used. These CPPs are created by averaging all programming contained in that Syndicator's program type categories. Commercial occurrence expenditures are then pro-rated based on their relationship to a thirty second spot. The pro-ration is a straight pro-ration.

Direct response commercials are discounted 50 percent, versus the other commercial types within the same program, to more closely represent their actual purchase price.

Nielsen Ad Intel includes commercials categorized as sponsorships or fee spots into all national TV media beginning with January 2006 data. This action results in an increase in units and dollars being reported for any national TV media that utilizes this type of advertising. These occurrences are typically 10 seconds in length and air towards the end of a program.

Expenditure methodology for fee commercials is as follows: Nielsen Ad Intel will assign 1/3 of a standard 30 second commercial rate to fee spots, then pro-rate the occurrences by duration. For example, if a 30 second spot is $300; a 10 second fee spot will start off at $100 and then be pro-rated by duration so that the end expenditure will be $33.

Gross Rating Points (GRPs)

The Nielsen National service provides program ratings data estimated from the national People Meter panel on weekly basis. Program averages (Gross Average Audiences – the accumulation of airings in all markets for all stations in the period for which a particular episode aired, usually over the course of one week) are reported on an individual program basis although weekly multi-telecast program averages are used when individual telecasts are not available.

Nielsen Ad Intel™

13

2/15/2017 /2015

The program rating that is applied to each commercial is not adjusted according to duration.
















Nielsen Ad Intel™

14

7/28/2015

Syndicated Clearance Television Methodology

Within each market, Nielsen Ad Intel distinguishes between local and syndication clearances. For each syndicated program, commercials from the local market are compared to commercials in the national database, thus identifying the local TV stations that did and did not air the “national” commercial. For daily and weekly programs, five weeks of national episodes are used to ensure that those stations tape delaying programs match to the correct national episode. In the case of specials, a thirteen-week window is reviewed.

Once the correct national episode is found, the local market commercials are matched based on parent company and brand description.

Data is available from 5 am – 2 am.

Gross Rating Points

Commercial occurrences are linked to Nielsen’s Local service Viewers in Profile (VIP) ratings by report book month, station, day of week and quarter hour. Note, in metered markets, individual quarter-hour “metered” ratings are available for household data. The VIP sweep data is available for all markets for the November, February, May and July diary panels. Data for October, January and March are also available in the appropriate markets.

Because of the timeliness of Nielsen Ad Intel data, commercial occurrences are available prior to the completion of the diary month. Therefore, ratings are estimated by applying the most recent sweep period (station share and market universe estimate) to the DMA HUT and PUT viewing levels to the prior year’s sweep period. The November estimate uses the May station share rather than July.

Actual VIP sweep data is processed and made available in Nielsen Ad Intel as early as two weeks after the measurement interval.


Expenditure Data

Expenditures are not assigned to local market syndication clearances.










Nielsen Ad Intel™

15

2/15/2017 /2015

Spanish-Language Network Television Methodology

Nielsen Ad Intel identifies advertising activity in the following Spanish-Language networks; Azteca, Estrella, Mundo Fox, Tr3s (10/07-6/27/10), Telemundo, Unimas, and Univision.




A commercial is defined as any paid advertisement announcement, spot or message with duration of 10 to 120 seconds, aired by the television station and recorded by Nielsen Ad Intel. Monitoring is conducted 24 hours a day.

Reported advertising activity is based on commercials aired during network telecast programs. Local advertising purchased from a Spanish-Language network’s affiliated station and live in-show product endorsements by show hosts and other personalities are excluded.

Expenditure Data

Expenditure data reported for commercial activity is derived from program costs provided by the monitored networks. If a network does not provide broadcast data, expenditure information is supplemented by other industry sources.



Gross Rating Points

The Nielsen National service provides program ratings against the Spanish Language Network TV distributors. Nielsen Media Research is including Univision (starting 12/26/05), Telemundo (starting 1/30/06) and Telefutura (starting 2/27/06) in their standard pocket-piece side-by-side with broadcast networks. In Nielsen Ad Intel, Univision (UNIV) Telemundo (TELE) and Telefutura (TF) and Azteca (AZA) are distributors classified as Spanish Language Network TV. All three are currently rated against the Hispanic Sample. The classification will not change in Nielsen Ad Intel; however, the users will now have the option to report Univision, Telemundo and Telefutura (Azteca is not included at this time) against either the Hispanic Sample or the standard National People Meter. Azteca will be introduced in the later part of 2006.

Program averages (Average Audiences – based on average minute rating) are reported on an individual program basis although weekly multi-telecast program averages are used when individual telecasts are not available. In the case of ‘regional’ commercials, the program rating is prorated to reflect the

Nielsen Ad Intel™

16

7/28/2015

percentage of geographic coverage as is provided by the networks. The program rating is applied to each commercial within the program regardless of duration.
















Nielsen Ad Intel™

17

2/15/2017 /2015

Spanish-Language Cable Television Methodology

Nielsen Ad Intel identifies advertising activity on; Discovery en Espanol, Discovery Familia, ESPN Deportes, Fox Deportes, Galavision, GolTV (2009-2/2013), NBC Universo, Tr3s (6/10- 12/14), and Univision Deportes.


Nielsen Ad Intel utilizes a network of computerized monitoring sites to collect and identify program occurrence information. Using proprietary hardware and software designed specifically for this purpose and technology known as “computerized pattern recognition,” the service has developed the capability to identify commercials and other broadcast material based on their actual unique audio and video content.

Occurrence data collected from the monitoring sites are merged with non-occurrence information, including ratings and expenditure data, and the results are maintained on a centralized computer database. All broadcast data that has not been automatically identified through computerized pattern recognition are processed by Commercial and Data Analysts using standardized identification procedures.

A commercial is defined as a paid advertisement announcement, spot, or message with duration of 10 seconds or more, aired by the Spanish language cable network and recorded by Nielsen Ad Intel during the measurement period.

Advertising activity is limited to nationally distributed commercials aired during cable telecast programs. Locally sold advertising are excluded.

Expenditure Data

Expenditure data displayed for commercial activity consist of program costs provided by the monitored network wherever possible. Where broadcast data are unavailable, expenditure information is supplemented by other industry sources and may include data derived from proprietary models.

Based on input from market sources, expenditures for direct response commercials are computed at 50 percent of estimated expenditure rates for all other commercial types, to more closely represent their actual purchase price.


Gross Rating Points

The Nielsen National service provides program ratings data estimated from the national panel on weekly basis. Nielsen Ad Intel uses the total program average audience estimates with M-F multi-day program averages, matching by program code and telecast number. Average audience estimates are used and ratings are calculated on the total National Hispanic universe. The program rating is applied to each commercial within the program regardless of duration.


Live = Household tuning or persons viewing to a program, quarter-hour or time period to a

specific station or cable channel, including VCR record.

Nielsen Ad Intel™

18

7/28/2015














Nielsen Ad Intel™

19

2/15/2017 /2015

Spot Television Methodology

Monitoring Criteria

Effective September 25, 2000, Nielsen Ad Intel has been capturing local television advertising in 210 DMAs. In 131 markets, including the top 100, all commercial activity is reported. For the rest of the US, Nielsen Ad Intel reports on “registered” commercials, defined as those that have the opportunity to match a commercial airing in any of the Full Discovery Markets® or on National TV.

Within each DMA those stations that achieve a three share of the market for three consecutive sweep periods are considered eligible for monitoring. Stations are evaluated on an annual basis to determine if this criterion has been met.




A commercial is defined as any paid advertisement announcement, spot or message with duration of 10 to 120 seconds, aired by the television station and recorded by Nielsen Ad Intel. Public service announcements are excluded. Monitoring is conducted from 5 am – 2 am.

Expenditure Data

Expenditure data reported for commercial activity is derived from household cost-per-point information supplied by Service Quality Analytics Data, Inc. (SQAD). The cost-per-point information is based upon actual spot television buys placed by advertising agencies and media buying services.

The expenditures reported in Nielsen Ad Intel reflect a computation using a blended cost-per-point from SQAD’s data and Nielsen’s local market ratings data. The metered markets use Live +7 ratings and the diary markets use a Live +24 number.

Expenditures for local commercials with durations under 30 seconds are factored as follows:

1-9 seconds 15%

10-14 seconds 50%

15-19 seconds 65%

20-29 seconds 75%

The factors above will not be applied if multiple commercials air within the same pod for brands with the same subsidiary.

Local Spanish-Language Expenditure Data Expenditure data reported for local Spanish-Language commercial activity is derived from adults 25-54 average cost-per-point information supplied by Service Quality Analytics Data, Inc. (SQAD). The cost- per-point information is based upon actual spot television buys placed by advertising agencies and media buying services.

Nielsen Ad Intel™

20

7/28/2015

Effective January 28, 2002, expenditures for local Spanish-Language stations with reported activity as of 1997 (broadcast year) for which there are Hispanic ratings and Hispanic adults 25-54 SQAD CPP are calculated using the following calculation: the Hispanic A25-54 rating multiplied by Hispanic adults 25-54 SQAD cost per point. For markets in which Hispanic ratings do not exist, the local quarter hour rating is used to calculate the expenditure.


Gross Rating Points

Commercial occurrences are linked to the Nielsen Local service Viewers in Profile (VIP) ratings by report book month, station, day of week and quarter hour. Note, in metered markets, individual quarter-hour “metered” ratings are available for household data. The VIP sweep data is available for all markets for the November, February, May and July diary panels. Data for October, January and March are also available in the appropriate markets.

Because of the timeliness of Nielsen Ad Intel data, commercial occurrences are available prior to the completion of the diary month. Therefore, ratings are estimated by applying the most recent sweep period (station share and market universe estimate) to the DMA HUT and PUT viewing levels to the prior year’s sweep period. The November estimate uses the May station share rather than July.

Actual VIP sweep data is processed and made available in Nielsen Ad Intel as early as two weeks after the measurement interval.


Live + 7 = Households tuning or persons viewing to a program, quarter-hour or time period to a specific station or cable channel, including VCR record, and any DVR playback tuning/viewing within 7 days. This is only applicable for the LPM and set-metered markets.

Live + 24 = Households tuning or persons viewing to a program, quarter-hour or time period to a specific station or cable channel, including VCR record, and any DVR playback tuning/viewing within 24 hours. This is used for the diary markets.










Nielsen Ad Intel™

21

2/15/2017 /2015

Network Radio Methodology

Nielsen Ad Intel receives monthly summaries from participating radio vendors; Premiere and WestwoodOne (formerly Dial Global). In July 2014 Cumulus and WestwoodOne merged and the data for Cumulus will be reported under WestwoodOne.

All types of programming are included from January 2001 forward for Cumulus Media and Premiere; Westwoodone reports all programming from November 2001 forward. Prior to these time frames, exclusively short form programming is available.

Data is based on a broadcast calendar and is assigned to the first day of the calendar month.

Sub-networks are “rolled up” and reported at the network level.

Units and commercial duration are not reported. The monitored networks provide expenditure data for commercial activity.

Nielsen Ad Intel™

22

7/28/2015

Local Radio Occurrences

Nielsen Ad Intel reports data collected by Media Monitors Inc. MMI identifies all pre-recorded advertising that are: 30 seconds or longer on stations rating a 2.0 or better share (as rated by Nielsen Audio) in all monitored markets. Coverage includes all announcements airing in a market whether the announcement is bought locally or airing as a network/syndication clearance, for those stations which are network affiliated. Airplay is detected and identified by a patented and automated recognition process.

Currently 43 markets are monitored 24 hours a day, 365 days per year. On average there are approximately 20+ stations monitored per market. Local Radio markets are as follows: Atlanta, Baltimore, Boston, Charlotte, Chicago, Cincinnati, Cleveland, Columbus OH, Dallas, Denver, Des Moines, Detroit, Ft. Myers, Greenville-Spartanburg, Honolulu, Houston, Indianapolis, Kansas City, Las Vegas, Los Angeles, Louisville, Miami, Milwaukee, Minneapolis, Nashville, New York, Orlando, Philadelphia, Phoenix, Pittsburgh, Portland, Providence, Raleigh-Durham, Sacramento, Salt Lake City, San Antonio, San Diego, San Francisco, Seattle, St. Louis, Tampa, Washington DC, and West Palm Beach.

Expenditure Data

Expenditure data reported for radio advertisements is computed by applying SQAD’s Spot Radio Service cost-per-point for Adults 18+ (average level) to Nielsen Audio ratings for Adults 18+. Spot radio cost-per- point estimates are received quarterly by daypart for each measured market. Each station's DMA level ratings per daypart, supplied twice a year by Nielsen Audio, are multiplied by the cost-per-point by daypart to estimate weekly expenditures. In an effort to provide more accurate Local Radio expenditures, Nielsen Ad Intel will prorate each commercial occurrence using commercial duration beginning with data captured from January 2007 (broadcast month) forward. Base expenditures are calculated at 60 second durations, a straight proration will be applied to those occurrences deviating from a 60 second duration. Local Radio expenditures are reported on a weekly summary basis.

Expenditures are available/updated monthly and reported four weeks after the reported calendar month’s end.

Ratings

Reported ratings data are provided by Nielsen Audio. Diary, Average Quarter Hour, DMA level data from Nielsen Audio’s Spring & Fall Expanded Summary Dataset Surveys are received at Nielsen Ad Intel twice a year. With this data, Nielsen Ad Intel is able to report on 45 local radio age/sex demographics.

Local Radio ratings are calculated using Nielsen Audio provided per market impressions and universe estimates. Occurrences are matched to Nielsen Audio impressions data via quarter, DMA, call letters and daypart; then divided by the market UE. By using DMA level impressions that are only released twice a year, Nielsen Ad Intel must replicate a quarter from the prior quarter. For example:

The Fall 2007 Survey is applied to Q1 & Q2 2008 occurrences The Spring 2008 Survey is applied to Q3 & Q4 2008 occurrences

Audience estimates are derived by Nielsen Ad Intel based on Nielsen Audio’s copyrighted and proprietary audience estimates. It is not an estimate produced by Nielsen Audio. Questions regarding Nielsen Audio Methodology should be directed to your Nielsen Audio representative.

On Monday July 08, 2013, the January through May 2013 ratings data was enhanced to report ratings using Portable People Meter (PPM) data versus previously reported Diary ratings. Thirty nine Local

Nielsen Ad Intel™

23

2/15/2017 /2015

Radio markets’ ratings and expenditures will be updated to reflect the Nielsen Audio PPM™ data source, while four markets remain using Diary ratings methodology: Des Moines, Greenville, Honolulu and Louisville.

This enhancement provides data that is collected electronically using inaudible codes that identify the source of a broadcast, such as a radio station. In the Nielsen Audio PPM service, "station" can also refer to digital broadcasts, including HD Radio®, HD multicasts and Internet streams. Audience estimates generated from each monthly survey are used as the buy/sell currency for radio stations and advertisers/agencies.

Ratings data is made available monthly and reported four weeks after the reported calendar month’s end.

Reportable Local Radio Dayparts

1. Overnight M-S MID-6AM 2. AM Drive M-F 6AM-10AM 3. Daytime M-F 10AM-3PM 4. PM Drive M-F 3PM-7PM 5. Evening M-F 7PM-MID 6. Weekend S-S 6AM-MID

Nielsen Ad Intel makes every effort to provide the most complete and accurate data possible, but does not warrant that the data will be complete or error free.

Nielsen Ad Intel™

24

7/28/2015

National Magazine Methodology

Source: Nielsen Ad Intel

Subscriptions: National advertising activity is monitored in 200+ consumer-oriented magazines on a continuous basis through mailed subscriptions. All issues, including special issues, double issues and supplements, are analyzed throughout the year by examining subscription copies obtained from three markets across the United States. Magazine analysts record all reported data elements about the issue and all specifics for national ads meeting the 1/6

th of a page minimum ad size.

Magazines included in the Nielsen Ad Intel database typically have a minimum paid circulation of approximately 225,000 (as verified by the ABC - Audit Bureau of Circulation). This roster is reviewed and updated annually to reflect additions requested by clients. There are instances where magazines not meeting the minimum circulation requirement are added based on client demand.

Magazine subscriptions that Nielsen Ad Intel secures are considered national editions. By default, mailed subscriptions received may be the 21+ national editions if a 21+ national edition is offered by the publication. Ideally magazines that have a true national edition are requested, secured and coded. If no true national edition exists, three regional editions of the magazine will be reviewed. Ads appearing in all three regional editions will be considered “national” ads and reported as such. Regional ads are not reported.

Occurrence Detail Reported:

Publication Name

Publication Parent Co

Publication Date (Issue date) - Monthly publication occurrences will be reported on the first of the month.

Total Pages (in issue)

Physical Page Number

Ad Size (as decimal)

Ad Position

Ad Type

Bleed Indicator

Section

Color

Parent Company

Subsidiary

Brand

Product

PCC

Adjacent Ad

Opposite Ad

Right Ad

Left Ad

Expenditures

Nielsen Ad Intel™

25

2/15/2017 /2015

Types of Ads: Spread: An advertisement designed occupy two facing pages as a single unit of space (no

separation by a gutter or inside margin).

Center Spread: An advertisement occupying the two center pages of a publication, where there is a continuous gutter.

Gatefold: Typically, a four page sheet (though it is not limited to four pages), creased and folded approximately half way of its width, so as to bind one end and open up like a gate. These ads open beyond the covers of the magazine.

Insert/Outsert: An advertisement that is bound into a publication and must be at least two sided. They are typically printed on stock that may be different in size, shape and weight from the publications normal stock. Inserts may be easily removed from a magazine for reference or informational purposes. Scent Strips are currently coded for this type of ad. Pop-up ads are coded as an insert type ad.

Tearout/Liftout: A single page card that serves as a business reply card. Typically, these cards are 5x7 in size and are perforated near the edge at the binding.

Multiple Page: Refers to advertisements that occupy more than one page, but not considered spread ads.

Spread/Tearout: A spread that contains a perforated tearout between the two pages (order form, coupon, sample).

No Type: A generic term referring to any advertisement that is shown on a single page and does not fit the criteria of any of the other designated types. The size of the ad may vary within the page.

Bleed: An ad that has inked impression extending to the trimmed edges of one or more sides of the page. Ads that bleed into the gutter are considered non-bleeds.

Advertorials are not counted towards national ad space. For example, if there were a one page Cover Girl ad followed by a five page advertorial on skin care, the ad would be reported while the five pages of advertorial would be maintained only in the total number of pages in the magazine, but not specifically reported as an advertorial.

Any non-traditional type of advertising (if found to be national) such as polybags are coded with the ad type of miscellaneous. These ads are not able to be scanned to produce a reportable image.

Expenditure Source: Magazine rate cards, SRDS, and Nielsen Ad Intel

Expenditure Methodology: National magazine advertising occurrences are processed through an expenditure calculation that uses published advertising rates for each magazine. Rates are obtained from published rate cards or rates provided to Nielsen Ad Intel by SRDS. Rate information is updated twice a year in January and July. Costs reflect discounts earned by corporations (ultimate parent companies), by accumulating the advertising activity of all brands within the corporate structure.

Discount levels recognize the total advertising activity of the previous twelve months and these discounts become the basis for determining unit costs for the next six months. Twice a year discount levels are recalculated (based on the previous twelve months of activity) to best capture spending fluctuations that might occur as brands significantly increase or decrease their magazine spending.

If a publication’s media kit provides a 2-Color rate that will be used in expenditure calculations. If a 2- Color rate card is not provided, the 4-Color rate card is used.

Nielsen Ad Intel™

26

7/28/2015

Local Magazine Methodology

Occurrences

Local magazines are monitored as published (monthly, semi-monthly, weekly, etc.) from mailed subscriptions. Magazine advertising analysts enter all pertinent information, including advertiser name, brand name, and product classification code. All advertising supplements as well as special sections and inserts/outserts are captured.

Prior to data of November 2011, Local Magazine occurrence and expenditure data was provided to Nielsen Ad Intel by AdSpies. From November 2011 forward, occurrence and expenditure data is gathered directly by Nielsen Ad Intel.

Expenditure Data

Expenditure data reported for advertisements is derived from each magazine's published rate card. Reported expenditures are estimated based on the advertiser's actual activity to date including ad size and eligibility for frequency discounts.

Nielsen Ad Intel™

27

2/15/2017 /2015

National Newspaper Methodology

Occurrences

Newspapers are monitored seven days a week (or as many days as published) from mailed hardcopy and electronic subscriptions by Nielsen Ad Intel. Newspaper analysts measure each display advertisement to determine total column inches of each ad insertion. Display ads must be 1/16 of a page or larger to be reported. All pertinent information including name of the advertiser, product category, total column inches, date of the ad insertion, section of the newspaper that the ad was located and primary promotion/event (if any) carried in the ad is collected. This data is incorporated into Nielsen Ad Intel standard coding guidelines.

National newspaper advertisements are defined as those advertisements that appear in the New York Times National edition and ads that also appear in at least three regional editions of either the Wall Street Journal or USA Today. The three national newspapers are monitored every publication day.

Prior to data of November 2011, newspaper occurrence and expenditure data was provided to Nielsen Ad Intel by AdSpies. From November 2011 forward, occurrence and expenditure data is gathered directly by Nielsen Ad Intel. The following additional changes also took effect with data of November 2011:

Reporting of movie theater ads.

Employee recruitment display ads that fall within the Classified section of the newspaper are reported provided that they meet the minimum size requirement. Previously, employee recruitment ads were only captured if they were outside of the Classified section.

Expenditure Data

Expenditure data reported for advertisements is derived from the national run of press rates as stated by the newspapers. Each advertiser's expenditure estimate is based on the advertiser's actual accumulated volume over the previous twelve months of activity. Annual column inch volume discounts earned are reflected in costing estimates for each advertiser. Each advertiser's column inch volume accumulation is monitored and recorded by Nielsen Ad Intel. Frequency discounts are also applied.

Nielsen Ad Intel™

28

7/28/2015

National Sunday Supplement Methodology

Occurrences

National Sunday Supplement advertising and expenditure data are reported for the following six National Sunday Supplements:

American Profile’s Community Table

Athlon Sports

New York Times Magazine

Parade

Relish

Spry

USA Weekend (stopped publishing 12/28/14)

DASH (merged with Relish)

LA Los Angeles Times Magazine (stopped publishing 6/3/12)

Display ads must be 1/16 of a page or larger to be reported. Our team of analysts record all pertinent information, including: Advertiser, Product Category, Newspaper Name, Paper Section, Event Code, Publication Date, Expenditures, and Size of Ad.

Prior to data of November 2011, National Sunday Supplement occurrence and expenditure data was provided to Nielsen Ad Intel by AdSpies. From November 2011 forward, occurrence and expenditure data is gathered directly by Nielsen Ad Intel.

Expenditure Data

Expenditure data reported for advertisements is derived by utilizing each supplement's four-color rate card and applying an earned frequency discount. Parade and USA Weekend offer both national and local advertising, thus a two-step discount structure is used for national advertisers. First, national advertising earns the “national participation” discount, and secondly, earned frequency discounts are applied.

Nielsen Ad Intel™

29

2/15/2017 /2015

Local Newspaper Methodology

Occurrences

Newspapers are monitored seven days a week (or as many days as published) from mailed hardcopy and electronic subscriptions, which include inserts, pre-prints or wraps. Insert information is supplied by a team of Nielsen representatives spanning many markets through home delivered subscriptions, ensuring that inserts are complete for participating newspapers.

Newspaper analysts measure each displayed advertisement to determine total column inches of each ad insertion. Display ads must be 1/16 of a page or larger to be reported. All pertinent information including: name of advertiser/brand, product category, total column inches, date of the ad insertion, section of the newspaper within which the ad appeared and primary promotion/event (if any) carried in the ad is collected for each newspaper monitored.

Prior to data of November 2011, Local Newspaper occurrence and expenditure data was provided to Nielsen Ad Intel by AdSpies. From November 2011 forward, occurrence and expenditure data is gathered directly by Nielsen Ad Intel.

The following additional changes also took effect with data of November 2011:

Pre-Prints:Pre-prints include a variety of circulars that are bundled with a given newspaper. These ads are typically sold via a different process as compared to the run of press ads seen within the newspaper itself. Pre-print activity is reported under its own unique distributor. Pre-print distributors have the same name as the primary distributor, but are appended with the letters "PP". For example, "Newspaper ABC" include run of press advertising and "Newspaper ABC PP" includes pre-print activity only. This change is applied to back data from 2005 to present.

We-Prints: We-prints are similar to pre-prints, but are generated as part of the press run of the newspaper. These ads are typically subject to the ad rates of the newspaper itself. Beginning with data for November 2011, we-print ads are included as part of the primary newspaper distributor, using run of press rates. We-print activity is assigned a unique section, allowing users to break this data out separately in reports if desired. Our previous methodology priced we-prints at the pre-print rate. This methodology enhancement will not apply to back data.

Additional Product Categories: Beginning with November 2011 data, Nielsen Ad Intel began coding movie theater ads. In addition, employee recruitment display ads that fall within the Classified section of the newspaper or magazine are reported provided that they meet the minimum size requirement. Previously, employee recruitment ads were only captured if they were outside of the Classified section.

Guidelines for Local Retailer Ads: Local retailers are assigned as the primary brand in all ads for the retailer, even if a specific product is featured. Prior to 2011, the primary brand in some of these ads was coded to the product advertised within the ad, rather than the retailer. In cases where the ad features a particular product, that product is coded as a secondary brand. Secondary brands are coded for specific categories only, including jewelry, watches and cosmetics.

Expenditure Data

Expenditure data reported for advertisements is derived from the local rate card of each newspaper (three categories; display/retail, classified and preprint). Rate cards under Nielsen Ad Intel are updated annually, including any changes to discount tables and mechanical requirements per newspaper section.

The cost of each ad insertion is estimated based on the following:

Costing Retail & Classified Display Advertising - The sources for newspaper expenditure data are the local retail and classified annual bulk-discount rate cards (Sundays and weekdays). National, nor combo (AM & PM editions) nor category-specific rate cards are applied to local newspaper activity. All costs are estimated at black and white rates. No premium charges are applied.

Nielsen Ad Intel™

30

7/28/2015

Annual column inch volume discounts earned are reflected in costing estimates for each advertiser. Each advertiser's expenditure estimate is based on the advertiser's actual accumulated volume over the previous twelve months of activity, or the total number of months the title has been measured by Nielsen Ad Intel, if less than twelve.

Each advertiser's column inch volume accumulation is monitored and recorded by Nielsen Ad Intel.

Effective 1/1/2009 – Pre-print inserts have the following expenditure methodology: (Ad Size in Pages * Rate) * Circulation / 1000 = Estimated Cost

Estimated costs are based on the number of pages (instead of ad size) of a preprint/insert, which allows Nielsen Ad Intel to utilize more specific rate detail provided in the newspaper’s rate cards & media kits. Frequency discounts still apply. Ad sizes for inserts are reported in column inches; however, the cost is based on number of pages and frequency discounts. Adjustments are not made for partial runs. As a result, estimated costs for preprints/inserts are inflated for any partial run material.

Prior to January 2009, expenditures for preprints and inserts were estimated based on the published rate for a tab size unit. The appropriate cost-per-thousand (CPM) rate was calculated against the newspaper's full circulation. Adjustments were not made for partial runs. As a result, estimated costs for preprints/inserts will be inflated for any partial run material.

Nielsen Ad Intel™

31

2/15/2017 /2015

Local Sunday Supplements Methodology

Occurrences

Newspapers are monitored from hardcopy mailed and electronic subscriptions. Local Sunday Supplements are newspaper sections that are categorized as Magazine (Local) by Nielsen Ad Intel. The definition for this section is a local lifestyle and general interest magazine that is published by the newspaper, most commonly found in major market Sunday newspapers.

Analysts measure each display advertisement to determine the total column inches of each ad insertion. Display ads must be 1/16 of a page or larger to be reported. All pertinent information including name of advertiser/brand, business category, total column inches, and date of the ad insertion is collected for each supplement monitored.

Prior to data of November 2011, Local Sunday Supplement occurrence and expenditure data was provided to Nielsen Ad Intel by AdSpies. From November 2011 forward, occurrence and expenditure data is gathered directly by Nielsen Ad Intel.

Expenditure Data

Effective January 2006, rates applied to Local Sunday Supplements, also known as Magazine (Local) sections, are supplied by each publication and specific to the rate card for this section of the newspaper. Color rates are also applied. Prior to 2006, the local retail, black and white rates were applied. Reported expenditures are estimates based on an advertiser's cume inches and projected to an annual contract rate (rolling 12 month average). The annual contract rate is applied to the current reporting month’s estimates. If the advertiser's cume YTD inches do not project to a contract line or inch minimum, the newspaper's open line rate is applied. Frequency discounts are also applied.

Nielsen Ad Intel™

32

7/28/2015

Free Standing Inserts (FSI) Methodology

Occurrences

Coupon data consists of FSIs for all packaged goods, credit card svcs, and all types of restaurants (including ice cream and donut shops). FSIs are gathered from newspaper subscriptions in 67 markets across the continental US. The coupon advertising analysts analyze each coupon and collect all pertinent information. This data is incorporated into Nielsen Ad Intel standard coding guidelines.

A coupon is defined as a paid manufacturer offer appearing in a Sunday newspaper's free standing insert supplement, circulated by a coupon distributor and coded by an Nielsen Ad Intel analyst.

Expenditure Data

Expenditure data reported for FSIs reflect costs provided by the coupon distributors.

Nielsen Ad Intel™

33

2/15/2017 /2015

Outdoor Methodology

Occurrences

Outdoor advertising plants located throughout the continental U.S. supply Nielsen Ad Intel with their advertising data. Prior to January 2002, only billboard and poster advertising was reported. After this date, all outdoor advertising, such as transit, kiosk, and furniture, is included.

Nielsen Ad Intel data analysts collect all pertinent information, including: brand name, advertiser, ad type, product category, date, and expenditure data from the plants' sales logs and incorporate this information into Nielsen Ad Intel standard coding guidelines.

Expenditure Data

Expenditure data reported for advertisements is supplied directly by the outdoor plant operators.

Most of the outdoor companies providing Nielsen Ad Intel with their data do not provide detail level information, and therefore this element is not reported.

Nielsen Ad Intel™

34

7/28/2015

National Internet Methodology

Occurrences

Nielsen Ad Intel receives weekly activity from Nielsen’s AdRelevance Service. The universe of coverage for Internet data consists of all reportable sites and sub-sites captured via Nielsen’s probing technology which utilizes Nielsen’s Cross-Platform Homes Panel. Nielsen Ad Intel reports data from ad supported websites only. Micro-sites, built as extra sections on an advertiser' site (e.g., Lucky Charms game promo on the General Mill's site) are not tracked nor reported. Weekly Internet spending is reported by brand and site/sub-site combination. All activity will be assigned to the Monday of each broadcast week. Occurrence level detail is not available.

AdRelevance metrics are based upon display advertising that is bought and/or sold on a cost-per- thousand (CPM) basis. Therefore performance based advertising is excluded from tracking or it is assumed that the ad unit was purchased on a CPM basis. AdRelevance also does not report sponsorships, promotions, sweepstakes, partnerships, barter impressions, performance based and affiliate advertising or co-op campaigns. Spending for paid search campaigns is not included in our reporting. Nor do we measure pre-roll/streaming advertising or any ads that appeared within an application (i.e. AOL Messenger). In-house ads (advertisements that promote a brand/product owned by the same company as the site itself) are not assigned expenditures or reported in Nielsen Ad Intel resulting in significant decreases in reported activity for these sites.

Internet advertising activity is loaded to Nielsen Ad Intel on a broadcast weekly basis for data from two weeks prior. Each weekly load includes advertising activity that has been reviewed and verified by Nielsen Ad Intel. Any data for the current week that has not yet been verified will not be loaded until the following week. This can include activity for new brands that require additional research before they are reported.

Expenditures

CPM data used to calculate internet expenditures is from SQAD’s WebCosts service. SQAD’s contributing agencies provide their negotiated/purchased rates and related display ad data to SQAD on a confidential scheduled basis. Some of the display ad data attributes include: Site name, negotiated price, time span, ad size and creative type. Agencies do not provide the advertiser's name or any other information that would expose the product.

AdRelevance aggregates SQAD’s site level CPMs, then aggregates that data to AdRelevance site genres. Site genre CPMs are multiplied by site level impressions data to calculate estimated costs.

The estimated dollar value for a given campaign, site, industry or advertiser is calculated by using statistical models analyzing estimated CPM values against a site’s genre, ad dimension, technology and delivery format. WebCosts CPMs and related factors are reviewed and updated twice a year.

Estimated Expenditures = (Impressions * (Site Category Rate * (delivery type factor and/or

dimension factor) + technology factor)

The above expenditure, traffic source and impressions methodology improvements were implemented in May 2013. Reprocessing took place to reflect these changes on data from September 26, 2011 forward.

Impressions

Internet impressions are reported for Persons 2+, beginning with data of September 21, 2009. Internet impressions represent the number of times a banner ad was requested and potentially viewed. Impressions are calculated by multiplying Ad Frequency by Total Page Visits. Ad Frequency is the average number of ads per page across all URLs that make up a site. Page views are modified at the

Nielsen Ad Intel™

35

2/15/2017 /2015

site level not at the category/genre level which more accurately reflects number of site visits. Hybrid page views are projected based on the NetView panel data.


Nielsen Ad Intel™

36

7/28/2015

Local Internet Methodology

Occurrences Nielsen Ad Intel receives weekly activity from Nielsen’s AdRelevance Service. The universe of coverage for Internet data consists of all reportable sites and sub-sites captured via Nielsen’s probing technology which utilizes Nielsen’s Cross-Platform Homes Panel. Nielsen Ad Intel reports data from ad supported websites only. Micro-sites, built as extra sections on an advertiser site (e.g., Lucky Charms game promo on the General Mills’ site) are not tracked nor reported. Weekly Internet spending is reported by brand and site/sub-site combination. All activity will be assigned to the Monday of each broadcast week. Occurrence level detail is not available.

Where applicable, sites and sub-sites within the AdRelevance database are assigned a “Primary DMA”. This process associates individual sites with the market in which the site is most likely to be viewed. Sites that are of interest to the general population are not assigned Primary DMA. All sites will be reported under the National Internet media type. A sub-set of the National Internet medium will be reported under individual local markets as Local Internet. Since Local Internet is a sub-set of National Internet, data should not be summed across these two media types.

AdRelevance metrics are based upon display advertising that is bought and/or sold on a cost-per- thousand (CPM) basis. Therefore performance based advertising is excluded from tracking or it is assumed that the ad unit was purchased on a CPM basis. AdRelevance also does not report sponsorships, promotions, sweepstakes, partnerships, barter impressions, performance based and affiliate advertising or co-op campaigns. Spending for paid search campaigns is not included in our reporting. Nor do we measure pre-roll/streaming advertising or any ads that appeared within an application (i.e. AOL Messenger). In-house ads (advertisements that promote a brand/product owned by the same company as the site itself) are not assigned expenditures or reported in Nielsen Ad Intel resulting in significant decreases in reported activity for these sites.

Internet advertising activity is loaded to Nielsen Ad Intel on a broadcast weekly basis for data from two weeks prior. Each weekly load includes advertising activity that has been reviewed and verified by Nielsen Ad Intel. Any data for the current week that has not yet been verified will not be loaded until the following week. This can include activity for new brands that require additional research before they are reported.

Expenditures

CPM data used to calculate internet expenditures is from SQAD’s WebCosts service. SQAD’s contributing agencies provide their negotiated/purchased rates and related display ad data to SQAD on a confidential scheduled basis. Some of the display ad data attributes include: Site name, negotiated price, time span, ad size and creative type. Agencies do not provide the advertiser's name or any other information that would expose the product.

AdRelevance aggregates SQAD’s site level CPMs, then aggregates that data to AdRelevance site genres. Site genre CPMs are multiplied by site level impressions data to calculate estimated costs.

The estimated dollar value for a given campaign, site, industry or advertiser is calculated by using statistical models analyzing estimated CPM values against a site’s genre, ad dimension, technology and delivery format. WebCosts CPMs and related factors are reviewed and updated twice a year.

Estimated Expenditures = (Impressions * (Site Category Rate * (delivery type factor and/or

dimension factor) + technology factor)

The above expenditure, traffic source and impressions methodology improvements were implemented in May 2013. Reprocessing took place to reflect these changes on data from September 26, 2011 forward.

Nielsen Ad Intel™

37

2/15/2017 /2015

Impressions

Internet impressions are reported for Persons 2+, beginning with data of September 21, 2009. Internet impressions represent the number of times a banner ad was requested and potentially viewed. Impressions are calculated by multiplying Ad Frequency by Total Page Visits. Ad Frequency is the average number of ads per page across all URLs that make up a site. Page visits are calculated by projecting the total number of page views within the panel to a total US audience.


Nielsen Ad Intel™

38

7/28/2015

Business-to-Business Publication Methodology

Data Collection

Nielsen Ad Intel trade publication data is received monthly from IMS (Inquiry Management Systems) based in Canada. IMS has been measuring competitive trade information since 1979. Through this partnership, Nielsen Ad Intel is able to offer detailed B2B occurrence information in over 800 trade publications.

Publication subscriptions are from the Northeast USA, and data is available from January 2006 forward. There are no ad size minimums applied to B2B data as all display ads are recorded and reported.

IMS data entry analysts are organized in teams, which allows for market familiarity and a high level of quality control. The data entry screens employed by IMS data collection analysts contain the history of each advertiser in the database for a period of up to 10 years. This contributes greatly to accuracy and consistency in the data collection process. Once this data is received, Nielsen analysts harmonize IMS descriptions into standard Nielsen Ad Intel brand/parent hierarchy structure.

Reportable Data Elements

Publisher, publication, parent company, subsidiary, brand description, PCC, issue date, page number, size, genre, ad type, premium position, section, creative headline, and expenditures.

Expenditure Data

Rate cards from each publication, secured annually, are used in IMS’ expenditure process. The 1x, 1 page, black and white rate for each publication is the only rate used and all expenditures flow from this one rate.

For example, when IMS records a 1/4 page ad, the system uses the 1x, 1 page, black & white rate and divides by 4. No color premiums or premium position cost factors are used.

Data Availability

Trade publication occurrence and expenditure data is available monthly, 12 weeks after the reported month’s end.

Nielsen Ad Intel makes every effort to provide the most complete and accurate data possible, but does not warrant that the data will be complete or error free.

Nielsen Ad Intel™

39

2/15/2017 /2015

National Cinema

Occurrences

Nielsen Ad Intel receives Advertiser/Brand campaign flight level detail from contributing Cinema companies on a monthly basis. The universe of coverage for National Cinema data consists of on-screen commercial and slide campaigns appearing on 100% of the markets available in a Cinema company’s inventory.

It is estimated that each National Cinema commercial runs approximately 4 times per day per screen for the length of the campaign flight. Commercial content runs during the “preshow” or “pre-feature” program before the featured motion picture. Total screen potential for any given National campaign, taking current contributing companies into account, is approximately +34K screens.

Reportable data elements for National Cinema are: Advertiser, Brand, Date, Duration, Ad Type, Distribution, MPAA Rating(s) (G, PG, PG13, R, etc), Expenditure, Impressions, and Creative Title.

Expenditures

Expenditure data is provided by contributing Cinema companies and is rate card based.

Expenditure data for each National Cinema campaign flight will be prorated according to the following industry rules:

75% of the expenditure for a specific campaign flight period is associated with the Friday-Sunday airings;

25% of the expenditure for a specific campaign flight period is associated with the Monday – Thursday airings.

Impressions

National Cinema impressions data are reported for Persons 2+. Impressions represent the number of times a campaign was potentially viewed. Impressions for National Cinema activity are provided per campaign flight and are contracted impressions supplied by each contributing cinema company.

Impressions data for each National Cinema campaign flight will be prorated according to the following industry rules:

75% of the impressions for a specific campaign flight period is associated with Friday-Sunday airings;

25% of the impressions for a specific campaign flight period is associated with Monday – Thursday airings.


Nielsen Ad Intel™

40

7/28/2015

Regional Cinema

Occurrences

Nielsen Ad Intel receives Advertiser/Brand campaign flight level detail from contributing Cinema companies on a monthly basis. The universe of coverage for Regional Cinema data consists of on-screen commercial and slide campaigns appearing in two or more markets but less than 100% of the markets available in a Cinema company’s inventory.

It is estimated that each Regional Cinema commercial runs approximately 4 times per day per screen for the length of the campaign flight. Commercial content runs during the “preshow” or “pre-feature” program before the featured motion picture. Total screen potential for any given Regional Cinema campaign varies depending on the cities which they aired.

Reportable data elements for Regional Cinema are: Advertiser, Brand, Date, Duration, Ad Type, Distribution, MPAA Rating(s) (G, PG, PG13, R, etc), Expenditure, Impressions, and Creative Title.

Expenditures

Expenditure data is provided by contributing Cinema companies and is rate card based.

Expenditure data for each Regional Cinema campaign flight will be prorated according to the following industry rules:

75% of the expenditure for a specific campaign flight period is associated with the Friday-Sunday airings;

25% of the expenditure for a specific campaign flight period is associated with the Monday – Thursday airings.

Impressions

Regional Cinema impressions data are reported for Persons 2+. Impressions represent the number of times a campaign was potentially viewed. Impressions for Regional Cinema activity are provided per campaign flight and are contracted impressions supplied by each contributing cinema company.

Impressions data for each Regional Cinema campaign flight will be prorated according to the following industry rules:

75% of the impressions for a specific campaign flight period is associated with Friday-Sunday airings;

25% of the impressions for a specific campaign flight period is associated with Monday – Thursday airings.


Nielsen Ad Intel™

41

2/15/2017 /2015

Digital

Nielsen Ad Intel Digital Service & Methodology Overview

Nielsen Ad Intel Digital data available in the Ad Intel application is powered with data captured by Pathmatics. The cloud based service operates with customized crawlers located in multiple proxy locations across the US. The online advertising measurement system mimics actual online users by automatically adjusting to traffic patterns to reflect user behavior. The crawling technology continually probes Web traffic 24 hours a day 7 days a week to capture ad data. It is constantly developing to keep current with the online and mobile advertising environment. The Ad Intel interface enables users to easily access and customize data; providing users with key insights to online and mobile marketing strategies and ad campaigns as well as in-depth cross media reporting on National and Local levels.

Ad Collection and Occurrences

The data within Nielsen’s Ad Intel Digital service consists of all reportable sites and subsites captured via the probing technology. The service reports data from ad supported websites only. Reported activity includes Desktop, Mobile Web & Tablet Web Display placements as well as Desktop Pre-roll & In-banner Video ads. “Crawlers” (machines in the cloud) gather data that is presented within the Ad Intel application. These crawlers visit ~5,000 US sites, which make up an estimated 85% of all web traffic (US), in much the same way as a real user visits a site. Crawlers employ Pathmatics’ PathSource™ technology to keep track of exactly what happens on each page load of the site. To accurately mimic user behavior, crawlers are launched from ~60 different geo-locations throughout the US. Micro-sites, built as extra sections on an advertiser' site (e.g., Lucky Charms game promo on the General Mill's site) are not tracked nor reported. Data is updated weekly in broadcast week increments. Digital estimated spend and impressions data is reported by Advertiser and site/subsite combination. Occurrence level detail is not available. Ad Intel Digital does not report sponsorships, promotions, sweepstakes, partnerships, barter impressions, performance based, affiliate advertising or co-op campaigns. Nor do we measure mid-post roll/streaming advertising or any ads that appeared within an application. Volume ad thresholds are run in real-time during the ad collection process. As the prober is traversing sites, it is calibrated to expect X number of ads per site/page. Some parameters are: Region, Device Type, Overall, Service Identification, etc. The service probabilistically clears collected cookies at a rate which matches user behavior, thereby giving the data a sample representative of actual market behavior. There is no demographic targeting during collection and there is no demographic reporting. The crawler collects cookies while loading hundreds of page views, and then clears them and starts over.

Direct vs Non-Direct Placements

The service maintains a list of recognized patterns that represents the entities that are known to be involved in digital buying situations, direct and non-direct. This list is inclusive of major adtechs, platforms and services which is continually reviewed and updated. When the crawler reaches a page, it will analyze its contents in full, and finds patterns that match these companies. Pathmatics’ PathSource™ technology allows Nielsen Ad Intel Digital to break out the first and last hop of an ad path a creative takes during the serving process. In addition to knowing what creatives ran on SiteABC.com, Ad Intel can report the different ways they were rendered up on the site.

Site & Traffic Analysis

The service crawls approximately 100-125 pages for each domain each day. URL's are updated and refreshed on a monthly basis. Very large scale sites (Yahoo, MSN, AOL, etc.) are divided into multiple domains first and then the top 100-125 pages are identified therein. The current methodology for determining traffic to sites which are measured as well as how often to measure them is derived from industry sources. Multiple entries for the homepage of a site are added to make sure it is crawled more than other URLs. The service uses in-depth metrics for each domain that provides the average number of URLs a typical user navigates to each time they visit a domain. When the system crawls, those are the URLs utilized to ensure the same ‘real user’ ad experience is mimicked, captured and reported.

Nielsen Ad Intel™

42

7/28/2015

Service Specifications

The crawler technology utilized impersonates multiple browsers and operating systems such as: Firefox, Chrome, IE, PC, Safari, etc. It has the capability to track dynamic, static, flash, and html5 creative and covers both HTTP and HTTPS. The technology supports iFrame and multiple nested iFrame (there is no limit to the number of nested iframes that the crawler can capture). This same technology is used to track creatives across Desktop, Mobile Web and Tablet Web. Desktop and Mobile Web & Tablet Web

· Desktop includes desktop web from the largest OS and Browsers

· Mobile Web +Tablet Web includes Apple and Android Devices (it currently does not include in-app)

· Desktop, Mobile Web +Tablet Web include all IAB standard and rising units, as well as other commonly seen ad sizes observed by the crawlers that fall outside of IAB sizes. (Custom units and skins are not currently included)

· Desktop and Mobile Web + Tablet Web data includes image, gif, html5, flash, and rich media.

· All tracked sizes include standard display, overlays, and interstitials. Typically, any ads that appear without custom user interaction.

Video

· Video data is desktop-only including pre-roll, instream and in-banner (it does not include TV, mobile web, tablet web, mobile in-app, tablet in-app, OTT, etc.)

· All video sizes and resolutions are tracked

· Video formats include WEBM, FLV, MP4

· Video includes solely Autoplay videos

Projected Impressions Methodology

Internet impressions are reported for Persons 2+. Internet impressions represent the number of times an ad was requested and potentially viewed. Impressions are calculated by multiplying Ad Frequency by Total Page Visits or Views. Ad Frequency is the average number of ads per page across all URLs that make up a site. The estimated impressions number reported is not the raw input number; but rather, a projected number. The number of impressions for a specific ad creative on a particular site on a particular day is calculated via a two part process. First, the number of actual impressions a crawl of a single page on a particular site on a particular day represents is calculated. This is computed by the following formula: SiteDateImpressionsPerCrawl = ((Vendor Source) Monthly Page Views for Site) / (# of days in month) / (# of pages crawled on this site on this date) Example: The engine crawled SiteABC.com 100 times on March 12, 2016 and the page views received were 6.2 million page views for SiteABC.com in March, 2016. Each page crawled on SiteABC.com on March 12, 2016 is "worth" 2,000 impressions (6,200,000 / 31 / 100). Then, to compute the number of impressions for a given creative on a site and date, the system counts the number of times it observed a creative on a site on a date and multiplies that by the previously computed SiteDateImpressionsPerCrawl value. CreativeSiteDateImpressions = (# Occurrences of creative on a site on a date) x SiteDateImpressionsPerCrawl Example: The crawl engine observed a specific Nike "Just Do It" creative on SiteABC.com 12 times on March 12, 2016. The estimated impressions of this creative on this site on this date are 24,000 (12 occurrences x 2,000 impressions/crawl). * Homepage weights: The service weights the publisher homepages higher than other pages. This means that home pages receive a higher percentage of the universe estimates since they are weighted higher.

Nielsen Ad Intel™

43

2/15/2017 /2015

Estimated Spend Methodology

Ad data from the crawler is combined with an estimated CPM from a multivariate model to generate an estimated CPM (on a per impression basis.) This CPM is then multiplied by an impression estimate to generate a spend estimate for that specific occurrence. Once spend estimates for every creative occurrence have been calculated, they can be aggregated by creative, advertiser, publisher, adtech vendor, etc. and displayed in the data. This process occurs during every night’s data processing and model parameters can be adjusted nightly based on newly received data. Ad Intel Digital metrics are based upon display and video advertising that is bought and/or sold on a cost-per-thousand (CPM) basis. Therefore performance based advertising is excluded from tracking or it is assumed that the ad unit was purchased on a CPM basis The multivariate model uses the below inputs for both direct and indirect ad placements. The variates differ by buy type. Creative sales channel (direct, programmatic, network, etc.) Creative location on the page (x,y) Creative size Creative format (display, video, etc.) Publisher domain Publisher category Publisher rank This model will use as its training data, two broad data sets: a programmatic/indirect CPM data set and a direct rate-card data set. The programmatic/indirect data set is built with CPMs seen actual crawling and programmatic clearing prices from actual buying in the market. The Direct rate-card data set is built from a variety of partnerships and industry sources. These data sets are continually updated allowing for model updates as market forces change.

Ad Size Guidelines

The list of ad sizes and types captured by the service is dynamic and evolving. As new sizes are captured, they are automatically added to the system. Some examples of the ad sizes currently captured on each platform are noted below. Desktop:

· 970 x 90

· 728 x 90

· 720 x 300

· 300 x 1050

· 300 x 600

· 300 x 250

· 180 x 250

· 160 x 600

· 120 x 60

· 1272 x 328

· 970 x 250

· 970 x 66

· 940 x 45

· 924 x 50

· 336 x 280

· 336 x 79

· 333 x 100

· 300 x 100

· 180 x 90

Nielsen Ad Intel™

44

7/28/2015

· 180 x 450

· 163 x 90

· 120 x 600

· 468 x 60

Mobile/Tablet:

· 120 x 20

· 168 x 28

· 216 x 36

· 300 x 250

· 300 x 50

· 320 x 50

· 320 x 480

Digital Glossary

· MEASURES: Spend - The estimated cost to place the ad on a site. Impressions - The point in which an ad is viewed once by a visitor, or displayed once on a web page. The number of impressions of a particular advertisement is determined by the number of times the particular page is located and loaded. CPM – Cost per thousand. Reflects the cost per 1000 estimated views of an ad.

· PLATFORM: Indicates what device or vehicle was used to deliver a digital ad to the end user. Current Options are Desktop, Mobile Web and Tablet Web. Desktop – Ads reported on this platform were rendered on PC and MAC devices. Mobile Web – Ads reported on this platform were rendered on Smartphones, Androids, and iPhone devices. Tablet Web – Ads reported on this platform were rendered on Android Tablets, iPad devices.

· AD TYPES: Display – Advertising that incorporates text, logos and pictures or images positioned on a website. Captured ad include but are not limited to IAB sanctioned types and sizes. Display will be reported on all platforms. Pre-roll Video – A pre-roll ad is a promotional video message that plays before the content the user has selected. The video advertisements are often repurposed television ads. Initially, pre-roll video placements are reported solely on Desktop. In-banner Video - A Video delivered as part of (inside of) the display ad creative for a given placement rather than initiating the use of a Video player.

· AD SIZE: Display placements are reported in width and height pixels. Video placements are reported in second durations.

· AD TECHNOLOGY: Describes the method(s) used to produce and create the digital ad creative. Current options are Image, Flash, HTML, HTML5, Video and Content. Content = Desktop advertising run in units/widgets Content Networks. This includes traditional Content Networks like Taboola and Outbrain, some sites' internal Native solutions (i.e. Yahoo or Buzzfeed), and Native providers like Nativo and Sharethrough that have custom units that run across numerous publishers. Each of these networks is added to our tracking solution based on industry trends and customer demand. These ads are run in a different technological implementation than traditional Display ads.

· BUY TYPE: Indicates the buying model used to purchase a particular digital ad.

Current options are Direct, Programmatic and Programmatic Unknown.

Direct Buy occurs when an advertiser or agency contacts a publisher or ad network directly to purchase inventory.

Nielsen Ad Intel™

45

2/15/2017 /2015

Programmatic Buy means using technology to buy and sell digital ads, through an SSP, Ad Exchange, Ad Network, DSP and/or Trading Desk. Programmatic encompasses RTB done in open ad exchanges, and also includes guaranteed direct buying, private exchange buying and preferred deals. RTB occurs within both private and open ad exchanges. It is when ad inventory is bid on by agencies/advertisers and sold by publishers through an auction model in an online ad exchange. It is possible for publishers to expose their unsold inventory to the RTB ecosystem to auction it off indirectly to the highest bidder

Programmatic Unknown is for any ad impression not confidently identified. For these instances, estimated spend is calculated using indirect CPMs.

· AD SERVICE COMPANY: Details the name of the company(s) that participate(s) in serving digital ads from placement to page. This list is dynamic and updated as warranted.

· AD SERVICE TYPE: The role an ad service company plays in delivering digital ads from placement to page. Possible options are Ad Exchange, Ad Network, DSP and SSP. Ad Exchange - an automatic platform for selling and buying online advertising inventories where space buyers (advertisers, media agencies, retargeting networks) and suppliers (networks, publisher) meet. An ad exchange allows automation for buying / selling phases and campaign implementation. Ad exchange compares supply and demand in real time and can display campaigns in real time impression by impression or by blocks of impressions. Real time bidding is the norm on ad exchanges Ad Network - a company which aggregates advertising inventories from a large number of publishers’ websites. The advertising network is a commercial and technical intermediary between advertisers and publishers. Ad networks are a single buying opportunity for media buyers and an opportunity to access advertising market for small publishers or to sell remnant inventory for premium / standard publishers. DSP - short for Demand Side Platform, is a technological platform that give media buyers access to ad exchanges and ad networks in RTB environment. A DSP is connected with multiple sources of ad inventory and allows buyers to bid for single ad impressions according to targeting and price restrictions and well defined goals. It allows media buyers to bid on display, video, social, and mobile ads. SSP - stands for Supply -or Sell- Side Platform. A SSP is a technological platform that allows publishers to access programmatic selling on ad exchanges and to optimize ad revenues and ad selling costs.

· PUBLISHER/DISTRIBUTOR GENRE: Each distributor is associated with a category the best describes the content presented within the distributor. Examples categories include: Entertainment, Sports, and News.

· MARKET (Distributor origin): Each distributor is associated with a market origin either National or local DMA. National and DMA designations are maintained under 'Market' for Digital. Geographic breakdown is determined by home origin of the site. Within the Digital media type, sites will continue to be associated with a DMA origin when warranted. To expose the market origin, users must select 'Market' as a reportable data element. Digital will not be broken out under National or Local as individual media types When National is selected, all distributor sites are returned regardless of distributor home origin.

· OCCURRENCE TYPE: House Ads are self-promoting ads placed on home sites. Users will have the ability to filter on Occurrence Type to remove House Ads from reporting.

· PROXY LOCATION: This is the geographical location/city where the ad was captured during the ad collection process. Reporting Proxy Location may give insight into geo-targeted placements. Proxy location reporting will be released in a later iteration of the digital service.

· AD VISIBILITY: Describes where on the page an ad was seen. Choices include: Above the fold and Below the fold. Ad visibility will be released in a later iteration of the digital service.


Documents

Nielsen Ad Intel™ Methodology by Medium