Research

For the purposes of the audience research, an annual establishment survey is held for SKO. Since 2011 the survey is commissioned in collaboration with the independent audience measurement services for other media (see Collaboration). The research is conducted by Kantar TNS on a continuous basis. The MSS is used as a source for census information on changing demographics, television, online devices and other media equipment as well as media use in the Netherlands.

MSS was annually conducted until 2017 in a nationwide representative sample of net 6,000 private households and 5,100 individuals. The sample preparations for the audience research panel is based on the results of the establishment survey as well as census data (Golden Standard calibration tool provided by the the Market Research Association in collaboration with the National Statistics Agency). Universes are used for weighting and panel control purposes. The MSS also provides with addresses for recruitment of the television panel households.

TV ratings are estimated by means of a panel of private homes, representative of all particular Dutch households (including non television households). The sample of the audience research consists of approximately 1,250 households, including at least 2,800 individuals aged three years and older. Households are representative of the Dutch population on a number of characteristics, such as region, household life cycle, education of the head of the household and ethnicity. Panel recruitment is carried out by GfK by means of a 65-cell matrix based on these household characteristics.

Continuous controls on cell and panel representativeness are carried out by GfK, SKO and its Technical Committee.

Households are only selected for panel membership if all household members are willing to participate. Once a household is selected, each member of the household is surveyed on a large amount of characteristics. At least once a year they are surveyed.

Panel membership lasts for a maximum of six years. Panel households do not receive paid incentives in order to participate in the research. They are compensated for the yearly costs of electricity used by the data collection on the home (the meters). They participate in a three-month lottery.

Each (Smart-) television-set and tv equipment (set top box, media centre, HDR, DVD, VCR) in the household's residence is connected to a meter device. The metering system consists of a remote control, with a button for identification of each household member and a device to register viewing, including a display able to show brief instructions, placed nearby the television set.

The viewing of all household members is registered by the system. Registration starts automatically at the moment the television set is switched on. All household members aged three years and older, identify themselves as viewers by pushing a button on the remote control when he or her starts watching. When their viewing stops, the same button is used. 

The viewing of panellists through the television sets and equipment is registered by the meter device on a second by second level. Every night after 2 am, all the viewing registrations in all household meter devices are automatically uploaded to the GfK central computers for channel detection and processing.

The meter device is equipped with a measurement module to detect the channels that are being viewed in the household. The main configuration of the meter consists of a modules for Enhance Audio Matching (and until 1-1-2018 Metric Line watermarking).

Enhanced Audio Matching (EAM) is a technique that uses audio matching for channel identification. EAM-measurement is not dependent on broadcasters or television reception types (digital or analogue, cable or antenna, with or without decoder). The EAM module on the meters samples the audio in all tv-sets in the household by taking into account all potential audio sources (tv-set or other tv-connected equipment). The digital samples are stored in the meter and later compared (matched) with the audio obtained from all the content broadcast by channels that are referenced. The audio is recorded on the so called channel reference sites, at the central GfK computers.

Viewing behaviour is ready to be reported after identification of channels by audio matching, data processing and after validation and weighting. Daily viewing figures are available the next morning.

Every night all information on viewing statements is checked on completeness and validity. The quality of the viewing data is controlled by means of different checks. One check is done by identifying situations when the television-set appears switched on during a long time and there are no viewers registered. Information is passed through to panel management, which will contact the household and verify the situation. Data of this household is not used until the problem is solved.

Despite the high daily response of panellists (more than 98% of the households is reached on a daily basis), weighting is necessary to compensate daily fluctuations in the sample. Weighting guarantees representative results based on education, household size, social class, age, gender, position in household, region and combination of these characteristics.

Viewing data can be related to programs and spots schedules. Commissioned by SKO, the Nielsen Company registers, harmonizes and classifies, on a daily basis, all broadcast information of a number of channels – the so called Full Audit-channels. The exact starting- and ending times of all broadcasts are verified and logged, in case unexpected changes in programs and spots schedules may occur. Nielsen audits for each commercial break whether spots have been broadcasted. They also classify and harmonize characteristics of commercials, billboards and channel promotions. Broadcast data are also processed during the night and available the next morning.

Each day, around 7.30 am, the complete viewing and broadcast data of the day before are available to SKO partners, participants and other clients. Several viewing reports are available to the public on a daily basis on the website of SKO.

Several software systems are used to process the data and calculate applicable results such as ratings, market shares and number of viewers. Stichting KijkOnderzoek (SKO) controls the results and certifies the software suppliers. An overview of certified software suppliers can be found here.

The organizations responsible for the official radio, print, television and internet audience measurements in the Netherlands (NLO, NOM, SKO and VINEX) founded the Media Standard Survey (MSS) in 2010. The survey started in 2011 aimed as source for census information on media equipment and use in the Netherlands.

The collaboration between the official media audience measurement services in the Netherlands is a unique initiative. However MSS is not the only example of an agreement between the Dutch media audience measurement services.

For example, all audience media measurement services in the Netherlands use the same Lifestyle classification developed by the Netherlands Public Broadcasting (NPO) since 2010. Since 2013, the audience measurement services, their partners and the SCP (Sociaal en Cultureel Planbureau) commissioned a Time Budget Study (Media:Tijd), which is used as a hub or basis to fuse the different media audience currencies (CrossMedia:Time).

Please find below a description of the audience research methodologies.

Methodological description Dutch TAM

Appendix 1: Social status classification
Appendix 2: Business-to-Business classification
Appendix 3: SKO Commercial targets
Appendix 4: Cell matrix
Appendix 5: Variables in weighting procedure
Appendix 6: Channels in he raw data
Appendix 7: Output SKO program files
Appendix 8: SKO guidelines on the use of program titles
Appendix 9a: Country codes
Appendix 9b: Football clubs codes
Appendix 10: Footbal registration
Appendix 11: SKO Methods output spot files
Appendix 12: Branche classification
Appendix 13: Exceptions to grouping rule for program segment
Appendix 14: Calculation of spot reach using a probability model
Appendix 15: Period sample criteria
Appendix 16: SKO genre classification
Appendix 17: Promo classification
Appendix 18: Lifestyle targets (Dutch)

Calculation and reporting rules

Appendix 1: Definitions
Appendix 2: Calculation of spot reach sing probability model
Appendix 3: Period sample criteria
Appendix 4: Exceptions to grouping rule for program segments
Appendix 5: Adjoining continuous viewing statements in the TSV data
Appendix 6: Time shifted viewing until 28 days after broadcast (until 31-12-2017)

Calculation and reporting rules Online TV & Video