online commercials 2016 vast census data

STICHTING KIJKONDERZOEK
ONLINE
COMMERCIALS
2016
VAST CENSUS
DATA
MANDATORY CALCULATIO N AND REPORTING RULE S AND
RECOMMENDED PROCEDUR ES
``
STICHTING KIJKONDERZOEK
ONLINE COMMERCIALS 2016
VAST CENSUS DATA
MANDATORY CALCULATION AND
REPORTING RULES AND RECOMMENDED
PROCEDURES
AMSTERDAM 01-06-2016
This document is written by SKO.
© Author’s copyright. Any unauthorized use of the document subsequent to its publication in an
electronic or other medium or distribution or publication of said document to others who obtain
copies of it by any means without written authorization by SKO, shall constitute an infringement of
the Author's copyright
Pagina 1 van 19
TABLE OF CONTENTS
INTRODUCTION ..................................................................................................................................... 3
1.
MANDATORY CALCULATION RULES ........................................................................................... 7
1.1. VAST STARTS ....................................................................................................................... 7
1.2. VAST EVENTS ....................................................................................................................... 7
1.3. FURTHER CALCULATION RULES .......................................................................................... 7
2.
SUPPLEMENTARY MANDATORY CALCULATION RULES ........................................................ 10
2.1 NO ADDING-UP OF RESULTS ............................................................................................... 10
3. MANDATORY REPORTING RULES ................................................................................................ 12
3.1. ACKNOWLEDGMENTS .......................................................................................................... 12
3.1.1. SKO OC VAST CeNSUS .................................................................................................. 12
3.1.2. SKO OC VAST CENSUS COMBINED WITH RATINGS ................................................. 12
3.2. REPORTING LEVELS ............................................................................................................ 13
3.3
GEOGRAPHICAL ORIGIN ................................................................................................... 13
3.4
PLATFORM .......................................................................................................................... 13
3.5
TARGET GROUP ................................................................................................................. 13
3.6
RESULT TYPE ..................................................................................................................... 13
3.7
PERIOD ................................................................................................................................ 13
4. ADVISORY REPORTING RULES .................................................................................................... 15
4.1 RESULT TYPES AS USED BY SKO ........................................................................................... 15
5. APPENDICES ................................................................................................................................... 17
5.1. APPENDIX 1. LIST OF PUBLISHERS.................................................................................... 17
5.2. APPENDIX 2. LIST OF EVENTS AND TYPE OF COMMERCIALS ....................................... 17
5.3. APPENDIX 3. AVAILABLE FILES........................................................................................... 18
5.3.1. KANTAR MEDIA VAST DATA FILES .............................................................................. 18
5.3.2. NIELSEN CODING OUTPUT FILES ................................................................................ 18
Pagina 2 van 19
INTRODUCTION
The inventory of online video is growing, as it is the volume of online commercials played out. Online
commercials adopt different forms; pre rolls, mid rolls and post rolls. SKO measure scope are all
online video ad’s. The aim is to measure all video-campaigns booked and served by SKO-affiliated
publishers, agencies and advertisers according to SKO standards.
The basis of the SKO measurement is IAB’s VAST. VAST stands for Video Ad Serving Template. This
standard describes the format by which ad servers and video players exchange information about
video advertising. By using VAST, the complex delivery and serving process for online video
advertising becomes considerably simplified. More information on VAST templates is available on
the > website of IAB.
SKO Online Commercials VAST Census measurement (SKO OC VAST) measures starts and events of
online commercials at a census level. This is not a matter of respondent behavior registration; it is
about passive registration of play out, using server-based pixel tracking techniques.
SKO is the organization responsible for the audience research in The Netherlands. SKO’s goal to
realize a total measurement of all the viewing behavior in The Netherlands. For more information
on the Video Integration Model (SKO VIM) and the model for measuring and reporting Online Video,
please
consult
our
brochures
and
information
on
SKO’s
website.
See
https://kijkonderzoek.nl/online-kijkcijfers.
The objective of SKO’s Online Commercials measurement (SKO OC) is to delivery correct, accurate
and standardized data on online video commercials. These census data are an essential part of the
online video ratings. More information on the SKO Online commercial measurement is to be found
on SKO’s website. See https://kijkonderzoek.nl/over-online-commercials.
Since 2015, SKO has signed contracts and made additional contractual arrangements with video
content and video advertising providers as well, to ensure that the collection of ‘audience behavior’
(which is translated into an audience measurement) takes place in compliance with the applicable
rules laid down in Section 11.7a of the Dutch Telecommunication Act (the so-called cookie rules)
and the Dutch Data Protection Act (‘Wet bescherming persoonsgegevens’ or ‘Wbp’). The parties
responsible for compliance with the law by SKO, Kantar Media and the video content and video
advertising providers are clearly defined in these contractual arrangements. Finally, SKO also has
notified the processing of audience behavior to the Dutch Data Protection Authority
(Persoonsgegevens).
Pagina 3 van 19
On 24th of June 2014 SKO signed a two-year contract with Kantar Media Audiences for measurement
of online commercials. In January 2015 SKO renewed the contract with The Nielsen Company for
registration a classification of TV advertising, including now the harmonization of Online
Commercials.
Since June 1st 2016, Stichting KijkOnderzoek delivers VAST OC census data to its customers on a daily
basis. VAST OC data includes census data of online commercial (number of starts and standard
tracking events as defined in IAB’s VAST ) and online video-campaigns booked and served by SKOaffiliated publishers according to pre-defined standards.

The data contains online campaigns of the STER, RTL Nederland and SBS Broadcasting,
Sanoma and Disney. Later, data for BrandDeli will be released.

All premium campaigns tagged by this publishers are include in the data at this point.

Different types of advertising content are part of the measurement scope. This includes pre,
mid-, post- rolls and billboards

Events measured by SKO are: start, first quartile, midpoint, third quartile of total duration,
and complete play out of creatives

Based on observed User Agent the following devices are identified: pc (computers, desktops
and laptops), tablet, mobile (smartphones).

Events played abroad are included in our efforts.
SKO publishes guidelines that software agencies can use to edit and report VAST OC Census data
(including or excluding harmonized campaign information).
This report identifies three types of rules:
Mandatory calculating rules
•
Part of these guidelines consists of standard calculating rules that should be followed by
every user. In this way, we hope to prevent the publication of different, incomparable results in our
market. The SKO guidelines are intended to insure that the calculations of the various result types
are carried out in the same manner, regardless of the party carrying them out.
Mandatory reporting rules:
•
Apart from calculating rules, SKO also publishes reporting rules. Each user of aggregated
data who aims to publish or report results should adhere to these rules.
Pagina 4 van 19
Recommended/advisory rules:
•
Apart from mandatory calculating and reporting rules, SKO has developed advisory rules
that address auxiliary result types. With these rules, we aim to promote consensus in our field.
Appendix 1 offers an overview of the measured publishers and online video parties included in the
data. A description of the OC VAST result types and type of online commercials can be found in
Appendix 2. Appendix 3 contains a description of the available files.
Pagina 5 van 19
CHAPTER 1
MANDATORY CALCULATION
RULES
Pagina 6 van 19
1. MANDATORY CALCULATION RULES
Part of our guidelines consists of standard calculation rules that should be followed by every user.
In this way, we hope to prevent incompatible results from appearing in the marketplace. By
adhering to the SKO rules, parties can help create a level playing field with regards to calculating
and reporting.
1.1.
VAST STARTS
A starts regards the number of started creatives. A start event is used to indicate that an individual
creative within the ad was loaded and play out began. As with creativeView, this event is another
way of tracking creative playback.
1.2.
VAST EVENTS
Events measured by SKO correspond to the standard tracking events as defined in IAB’s VAST.

Start: an individual creative within the ad was loaded and playback began. See 1.1.

firstQuartile: The creative played for at least 25% of the total duration.

midpoint: The creative played for at least 50% of the total duration.

thirdQuartile: The creative played for at least 75% of the duration.

complete: The creative was played to the end at normal speed.
1.3. FURTHER CALCULATION RULES
Other result types based calculations are possible
Completion rate (25%, 50%, 75% or 100% Completed)
The percentage indicates the percentage of starts where 25%, 50%, 75% or 100% of the total running
time of the creative was played out.
In the VAST OC raw data, the count of events is provided for each combination of creative ID,
campaign and Ad server ID. The count indicates the amount of times the event was observed in the
time frame represented. The completion percentage can be computed for different selections based
on ad server, site name, creative, campaign, line item Id, Sales house, Agency or device type.
Pagina 7 van 19
X% Completed =
∑𝒄=𝟏
𝑪𝒐𝒎𝒑𝒍𝒆𝒕𝒆 𝐶 %
𝒛
∑𝒄=𝟏
𝑺𝒕𝒂𝒓𝒕𝒔𝒄
𝒛
Where C = 25%, 50%, 75% or 100% and z = 1, …, z starts in selection.
% Platform.
The type of device used to play out a commercial is a characteristic that is available in the raw OC
VAST Census data (Device Type). It is possible to distinguish between streams started on a computer
(desktop or laptop), a tablet or a mobile device (smartphone). The characteristic ‘null’ is also
available. Null category should be reported as “rest”.
The percentage PC, Tablet, Mobile and Rest indicates the ratio between the number of starts in these
categories and the total number of starts.
% Device can be calculated by dividing the number of starts (count) or other event per device by the
total number of streams (or event) of a started streams on all devices
% 𝑺𝒕𝒂𝒓𝒕𝒔 𝒅𝒆𝒗𝒊𝒄𝒆 =
𝑺 ∑𝒄=𝟏
𝑺𝒕𝒂𝒓𝒕𝒔𝑀𝑜𝑏𝑖𝑙𝑒
𝒛
𝒄=𝟏
∑𝒛
𝑺𝒕𝒂𝒓𝒕𝒔
%𝑬𝒗𝒆𝒏𝒕 𝒅𝒆𝒗𝒊𝒄𝒆 =
𝑺 ∑𝒄=𝟏
𝑬𝒗𝒆𝒏𝒕𝑀𝑜𝑏𝑖𝑙𝑒
𝒛
𝒄=𝟏
∑𝒛
𝑬𝒗𝒆𝒏𝒕
All result types % (completion rate, platform) can be computed for selections based on ad server,
site name, creative, campaign, line item Id, Sales house, Agency and device type.
Pagina 8 van 19
CHAPTER 2
SUPPLEMENTARY
MANDATORY CALCULATION
RULES
Pagina 9 van 19
2.
SUPPLEMENTARY MANDATORY CALCULATION RULES
2.1 NO ADDING-UP OF RESULTS
SKO KijkOnderzoek measures the actual watching behavior of the television panel (persons). SKO
Online Commercials VAST Census measurement (SKO OC VAST) measures starts and events of online
commercials at a census level. This is not a matter of respondent behavior registration; it is about
passive registration of play out, using server-based pixel tracking techniques. Streams played
abroad are included in our efforts.
The results of SKO OC VAST Census measurement and the SKO KijkOnderzoek ratings for TV
commercials should never be added up; these results are incompatible. When reporting OC Vast
results and SKO ratings, you should adhere to specific reporting rules (see 3.1.2).
Pagina 10 van 19
CHAPTER 3
MANDATORY REPORTING
RULES
Pagina 11 van 19
3. MANDATORY REPORTING RULES
This paragraph provides an overview of SKO’s reporting rules with regards to the SKO OC research.
As such, it may be regarded as an expansion of the calculation rules. The calculation rules in the
above have bearing on every OC data result report and combinations of such reports with ratings
research results. When reporting results, the following rules should be followed.
3.1. ACKNOWLEDGMENTS
3.1.1. SKO OC VAST CENSUS
When reporting results from the SKO OTV research, the following acknowledgement should be
included: ‘Stichting KijkOnderzoek OC VAST Census Measurement’ or ‘SKO OC VAST’.
Moreover, the reporting and result type, the period of year and the commercial harmonized
information should be mentioned.
Reports should also include the following note:
Currently, not all campaigns, platforms/devices of SKO partners and participants are included in the
SKO OC Census results.
3.1.2. SKO OC VAST CENSUS COMBINED WITH RATINGS
It is not allowed to report results that are based on accumulated SKO OC VAST numbers and SKO
KijkOnderzoek Online ratings. These numbers are not compatible.
However, these numbers can, in the same report, be reported separately. See also 4.1.1.
When reporting OC VAST results in combination with the results from the television and/or online audience
measurement, you should always include an acknowledgement of the source of the TV-ratings: ‘Stichting
KijkOnderzoek’ or ‘SKO Stichting KijkOnderzoek’. Reports should include the date, reporting type (type of
device, publisher), aggregation level (day, hourly), the period of year, the harmonized information for the
campaign. When reporting ratings, you should also mention the result type and the target group.
When reporting combined results from the television audience measurement, you must include the following
notes:
Results from SKO OC VAST Census and SKO KijkOnderzoek TV and Online should never be
accumulated; said results are incompatible.
Pagina 12 van 19
3.2. REPORTING LEVELS
SKO OC VAST Census Data is available on an overnight +1 basis and may be calculated from an
Overnight and several other levels.
SKO has formulated a link with commercials broadcasted on TV based on the fact that creative or
video is exactly the same to the corresponding creative for the TV spot (TVTID). The Online ID will be
linked to the TVID.
There is no link to the TV schedule for the TVTID in the SKO OC VAST Census Data. The link can be
obtained through the match with the Nielsen Online Commercial harmonized files.
3.3
GEOGRAPHICAL ORIGIN
SKO result data always includes streams that were started abroad.
3.4
PLATFORM
When starts are played out in an unspecified device, the device type reported is “null”. This device
should always reported as a so-called ‘Rest’ category.
SKO OC VAST census-results are always reported within all platforms. At this point, online
commercials starts and events played out via SMART TV apps cannot be distinguished in the device
type categories. They are classified as ‘pc’.
The platforms (devices) reported on must always be stated clearly.
3.5
TARGET GROUP
SKO OC VAST census data shall not include a target group indication or label.
3.6
RESULT TYPE
A result type name may only be used if the result type has been calculated according to the
mandatory calculation rules SKO OC VAST census data. Numbers or tables in publications need to
clearly state the result type in question.
3.7
PERIOD
The reporting period must always be stated clearly. It may be expressed as hour (s), year, quarter,
month, week or day (publication date). In case of deviating periods, you should always state the
starting and ending dates.
Pagina 13 van 19
CHAPTER 4
ADVISORY RULES
Pagina 14 van 19
4. ADVISORY REPORTING RULES
Apart from mandatory calculating and reporting rules, SKO has developed advisory rules that
address more miscellaneous result types. With these rules, we aim to promote consensus in our
field.
4.1 RESULT TYPES AS USED BY SKO
The result types described in this paragraph are used by SKO, either on the website or in publications
such as the annual report.
Pagina 15 van 19
CHAPTER 5
APPENDICES
Pagina 16 van 19
5. APPENDICES
5.1. APPENDIX 1. LIST OF PUBLISHERS
Code
1
2
3
4
5
6
7
8
9
Description
STER
RTL
SBS
DISCOVERY
VIMN
FOX
DISNEY
SANOMA
BRANDDELI
5.2. APPENDIX 2. LIST OF EVENTS AND TYPE OF COMMERCIALS
Basic VAST EVENTS:

start: This event is used to indicate that an individual creative within the ad was loaded and
playback began. As with creativeView, this event is another way of tracking creative
playback.

firstQuartile: The creative played for at least 25% of the total duration.

midpoint: The creative played for at least 50% of the total duration.

thirdQuartile: The creative played for at least 75% of the duration.

complete: The creative was played to the end at normal speed.
Campaign ID: every Advertisement is assigned to a campaign, and that campaign has some sort of
unique ID assigned to it in the database of the Agency that books it.
Creative ID: every unique Advertisement is assigned to a creative, and that creative has a unique ID
assigned to it in the database of the booking Agency.
Line ID: every Advertisement is assigned to a campaign, and that campaign has an unique order ID
assigned to it in the database of the Agency that books it. The purpose of this ID is only to further
split results of a campaign. The publishers use it as a unit of different orders in their booking systems.
Ad-url: For every campaign and every creative that has been played out, the URI for it, url-encoded.
For example:
http://media.rtl.nl/web/components/service/streamspots/2013/Henkel_Persil_DuoCaps_10s_100
92013.mp4.
Sitename: The publishers are asked to identify themselves in the tracking. For example
Pagina 17 van 19
“Google” or “YouTube” or “RTL” or “STER” .
Sales House: in the case the advertising space was sold by another party than the publisher, this
information also needs to be passed to the measurement.
Agency: The name of the Agency that is responsible for the booking of the advertisement.
Advertiser + Product : Simply name of the Advertiser and the product which the advertisement is
for.
CML TYPE Description
PREROLL
MIDROLL
POSTROLL
BILLBOARD
5.3. APPENDIX 3. AVAILABLE FILES
5.3.1. KANTAR MEDIA VAST DATA FILES
Kantar makes daily and hourly OC VAST census commercial data available for exploitation by media
planning software bureaus via SFTP. The output-files from Kantar Media Audiences are text files
containing

Day -1 data from 02:00 till 25:59 CET –(export.yyyymmdd.txt.gz)

Day -1 data with split per hour in the file (export.yyyymmddhourly.txt.gz)

Data every hour (export.yyyymmddhh.txt.gz)
Every day at around 03.15 CET Kantar Media Audiences generates a file with the aggregated results
from the previous day. Every hour at around 20 minutes past the hour, Kantar Media Audiences
generates a file with the aggregated results from the previous hour. The production of the file is
triggered by the successful completion of daily/hourly aggregation processes in the SKO Project.
5.3.2. NIELSEN CODING OUTPUT FILES
Nielsen will deliver every working day an ONLINE file (YYYYMMDD_HHMMSS.col), containing the
information received in the Input file and enriched with the fields filled by Nielsen (ONLINE ID,
duration, classification). Nielsen assigns every Ad- URL (which shows a working video ad) with a
distinctive ONLINE ID. When more URLs are related to the same online spot, then the ONLINE ID will
be the same.
Pagina 18 van 19
The ONLINE file is a separate file from the CLA file for TV. The format is created based on the CLA file
for TV.
Delivery of the files will be done by means of the FTP server from Nielsen. Delivery of all files will be
done in folder: media\tv\online-video.
The delivery from Nielsen to Kantar Media and the market is according to the following:
• Daily, on weekdays before 18:00PM.
• No deliveries in the weekends and bank holidays.
• One file per day for all online commercials.
The filename has the form YYYYMMDD_HHMMSS.col
At each delivery, both files include all ONLINE IDs created until the delivery day (=no incremental but
all data - so they contain historical data from the start date onwards).
Pagina 19 van 19