IGF 2010
VILNIUS, LITHUANIA
15 SEPTEMBER 2010
SESSION 84
1630
HOW TO MEASURE COMMUNICATION AND MEDIA
IN DIGITAL CONVERGING ERA



********
Note: The following is the output of the real-time captioning taken during Fifth Meeting of the IGF, in Vilnius. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the session, but should not be treated as an authoritative record.
********

>> ISMO SILVA: Good afternoon, ladies and gentlemen.  And welcome to this last workshop of the day on how to measure communication and media on digital converged era.  My name is Ismo Silvo.  I'm coming from Finland, public service broadcaster, and I am chairing this panel in this workshop.  
This workshop has been put together by such organisations as OECD, the European Broadcasting Union, EBU, and the European Audiovisual Observatory.  They very much are concerned about the fact that very little, especially on media content, communication, cultural content, is known about on Internet platform.  
Many of us here on this side of the table come from the broadcasting world, media world, which is well measured, well established, lots of information around.  We are all heading towards Internet platforms, and we are very concerned that we know very little about -- actually about Internet, what is happening on Internet platform in terms of business, particularly content wise, even good reliable statistics concerning consumption, people's use on media and cultural content, audiovisual content on Internet platform is not well known.  
So we want to promote the idea that we should, since we all are in this Internet platform world, we should invest better and more in knowing what's happening on Internet world.  So the main thrust of our main motivation is to know more about Internet or media or know more about media on Internet and, certainly, get more transparency on the Internet world.  
There's lots of politics being made on Internet, lots of regulation is discussed about.  There's lots of business being bought on Internet.  And we have never, in the media world, known so little about our environment than we now know of Internet.  So this is our main motivation.  
And I have here colleagues from various organisations who have been working on this topic, and we want to -- we have structured this panel so that we have two presentations that lead into this theme and into these topics, and then we have panelists who will comment from their experience and comment on their backgrounds, how they see the situation.  
And we want to be interactive, as all these workshops in this conference.  
So, first we'll start with a presentation from OECD, Frederic Bourassa, who is statistical expert in the OECD.  OECD has been doing lots of studies on telecommunication and also on, in particularly, on traditional media.  But they are also concerned about this media on the Internet era.  So Frederic is talking about or giving us an introduction on definition of the international framework for measuring good indicators that relate to media on Internet.  
So Frederic, the floor is yours.  

>> FREDERIC BOURASSA: Thank you, Mr. President.  Is it possible to put the slides on, please?  Thank you.  Just a word about the OECD.  One I will mention in the OECD in this presentation, I will mainly talk about the work done in my division, the Information Communication Consumer Policy division, so we are not part of the statistical directorate, so we are part of the science, technology, and industry directorate.  It's a bit different.  
Most of our experience in the ICCP Division, most of our expertise is on collecting data on telecommunication, Internet, and ICT in general.  
So we decided to join forces with the European Broadcasting Union and the European Audiovisual Observatory to organise this workshop because we feel that we could be complementary on the convergence measurement.  Even if we at the OECD are not specialists in the audiovisual content, we increasingly need data to measure convergence.  So you will find on this slide three questions which I will use to structure this short presentation, hopefully.  
So the first question is why do we need to measure the convergence of communication networks and audiovisual content?  
What are the indicators we need to collect?  
And how to proceed?  
Is it possible to flip -- to go to the third slide?  Okay.  
So why we need to measure convergence.  Here I listed three reasons, among many other valid reasons.  First one is to assess the economic impact of convergence on traditional telecommunication and broadcasting sector.  The second one is to measure the changes in individual and household access to individual content.  And the third one is to provide the policymaker data on the development of convergence and for them to benchmark themselves with other countries and to access their policy efficiency, which is one of the EOCD mandates.  
Indeed, we collect data mainly for policy analysis, and this is our main motivation, to measure convergence because we use today treat separately telecommunication policies and broadcasting policies.  
Convergence is, of course, for access technology, but it is also for -- there's also a convergence for the legal frameworks, like for instance, some countries include VOD in their broadcasting regulations.  Some don't.  And there's also convergence in regulatory bodies like for the Ofcom in the UK.  
So second question, which indicator do we need?  I selected an almost realistic list of indicators.  It's data that we can currently -- almost -- get from official sources like regulators, communication ministries, statistical agency, or annual reports.  
At the OECD, we collect telecom and Internet data according to similar categories, which are infrastructure indicators.  So this is a necessary condition to access audiovisual content.  You need a good Internet capacity, high speed bit caps, and if you want the audiovisual services, you need fiber.  Also, we need to know to have a look at the fixed broadband and wireless broadband penetration if you want to access mobile TV, for example.  We need reasonable access prices, and we can have a look to triple play offers.  And we began to collect data on the number of subscribers to Internet-based services, like video on demand or IPTV.  
The second type of indicators we can collect is obtained through surveys, so it's called individual and household ICT surveys.  It's surveys done in cooperation with Eurostat.  The current surveys we have already contain a question on audiovisual content access by the Internet, but it's not sufficient to have a good picture, so many other questions will have to be added to surveys if we want a good picture of convergence.  
And the third category is financial data.  Financial data is increasingly difficult to get, as there's more and more bigger global players in the audiovisual market, following many integration waves, and there's also new entrants in the market, so it's very hard to gather broken-down data from these huge media groups.  So this is one difficult one, the financial data.  
We've seen it in the OECD while trying to collect data on Internet providers' revenue investment.  It's very difficult.  We didn't manage to get good data.  
And the last category is other indicators that are harder to collect, and we would need some assistance from other countries or other organisations.  First there's the Web audience.  Our colleague will talk about that later.  The availability of services by platforms.  Market shares, which are very difficult to get as well.  And yes, probably many others, and later we are totally open to -- to proposals.  
So we proceed.  When we want to develop an indicator, we have first to set a common core list of the indicator we want to collect.  We have to agree on the indicator's definition.  We have two good definitions which are not too narrow, not too broad.  We have to find a wide international consensus on the list of indicators and definitions.  At the OECD, it's a long process.  We have to ask our working parties, the members from our -- the delegates from our member countries.  It can take a long time.  I will show you later.  And then fourth step, we collect the data.  
So as an example, I will talk about the last indicator that the OECD developed.  It's the OECD mobile broadband indicator.  So just switch to a timeline to show you how long it could be.  And even though this one was developed very quickly.  The first agreement -- yeah.  The first agreement to create this indicator was -- was done in December 2008.  In February 2009, a workshop was organised in Lisbon due to the very strong leadership of Portugal, so there was an agreement on methodology and definition.  One year later, after several rounds of discussion, the final definitions were available.  And in April, the OECD decided to use the definition for their own wireless broadband indicator.  And in May-June 2010, we made the first data collection for this data.  But harmonization problems were too important, so we decided not to publish them.
So the next -- the next collection will be November 2010, and hopefully we will be able to publish harmonized wireless broadband data.  So this is how long it takes to develop an indicator, and this one was developed quickly.  
Just to end, I will show you that our first -- most recent try to collect audiovisual service data, it's a questionnaire we sent to our non-European countries.  Basically, the same data that the European Audiovisual Observatory is publishing, we used the same questions, and we sent this questionnaire to non-European countries.  And in the last part of the questionnaire, there's a few questions which are interesting for convergence issues, like the number of on-demand audiovisual media services, the number of enterprise operating mobile TV services, and number of IPTV services as well.  So thank you very much, and we will have your question later.  

>> ISMO SILVO: Thank you, Frederic.  Would it be right to say from your point of view, from OECD point of view, this is really, really undeveloped area on statistics, this media on Internet?  Or do you see some other industry sectors that are so undeveloped at the moment?  

>> FREDERIC BOURASSA: From our point of view, it's undeveloped because it's new, the convergence is new.  We used to collect data separately, telecom on one side, broadcasting on the other side, so yeah, to measure the convergence is quite new.  

>> ISMO SILVO: All right.  Thanks, Frederic.  And our next speaker also leading us into this general theme attics is Dr. Andre Lange, Director of Market Department in European Audiovisual Observatory in Strasberg, and the Observatory is one organisation that Europe has put together in order to know more about audiovisual sector in a harmonized manner.  So Andre, floor is yours.  

>> ANDRE LANGE: Thank you.  It's more good afternoon.  I will just very rapidly introduce European Audiovisual Observatory.  I suppose a lot of people of you have never heard about our organisation, which is a pan-European public organisation based in Strasberg.  We are financed by Member States and European Union represented by the European Commission.  We are working in the classical field of the audiovisual sector, film, television, but more and more, of course, we are trying to understand what succeeds with new technologies and, in particular, the Internet.  
Internet is only one of the new technologies, you have cable, satellite, DVD, Blu-ray, everything is arriving at the same time for the audiovisual sector, and it makes a lot of problems.  
We collect facts and figures, statistics, but also legal information, and we have various databases accessible on the net and various publications.  
Now, Ismo was my first Director.  I arrived exactly today 17 years ago, and we were just joking on the fact that in the early years, we were discussing Internet was arriving and it's just a phase.  It will be forgotten very soon.  
Now we can list all the impacts of the Internet on the audiovisual sector.  Industry people will tell you that the main one is, of course, the development of piracy, peer-to-peer technology.  There has been the development of the retail online business important for DVD.  Development of user reason generated content, which are also not only user content, but also the industry has invested in platform like YouTube.  We have audio services integrated in -- offer.  We have the development of IPTV, connected TV, new services, Web TV, VOD, capture TV, mobile Internet, including a lot of video programmes.  We have competition on the advertising market between the traditional broadcasters and new players on the Internet.  
We have the role of Internet in the promotion of films, of TV programme, of schedules.  We have the role of Internet in the promotion of film and TV heritage.  We have also online filmography.  I think IMTV is one of the most popular websites on the Internet.  And we have more and more cooperative process of production using Internet and broadband.  
Just to give you one area of what we are trying to do, we are trying to monitor the development of the various on-demand services, and just for pure VOD, commercial VOD services in the European Union, June 2010, we have identified 170 services operating, which is -- we did make a first census end of December 2008, there was almost 160.  A lot of services already died, some new have appeared, an it's a very fast-moving activity.  
However, I think there is a bit of overestimation at this time of the importance of VOD on Internet.  According to the estimates provided by one of our providers, Green Digest, in 2009, online revenues for TV or film on Internet were still less than 20% of the transactional revenues including near video on demand or video on demand on other platforms than Internet, IPTV, cable, satellite, meaning that Internet is, for the commercial services, still relatively marginal, mainly because people still prefer to watch film and TV on the TV screen rather than on the PC screen.  
Now, our key word is transparency.  It was quoted in the workshop.  We need transparency for various reasons, of course, for the investors, they need to know where they are going with the investment and strategies.  We need transparency for various categories of stakeholders.  For example, on the importance of knowing what the potential audience of a service is for a writer or a distributed.  
We need, of course, advertising industry is keen to obtain more data on audiences.  The public needs data for various activities, first of all, to recognize freedom of expression, to recognize current state of cultural diversity.  There is a need of structural business statistics to calculate the importance of the sector in the GDP.  There is also issue of the protection of intellectual property rights.  There is a question of protection of protection of consumer.  There is a situation of analyzing the -- and defining the policies of the film and multimedia production and support to the industry.  
And of course, cities and consumer also need to be informed on the system, and this will be developed by a former speaker.  
I will not provide all our shopping list, partly prepared OECD shopping list presented by Frederic, but of course, we need a lot of data that currently either does not exist, either are just collected by private consultant company, and this type of monitoring at the national level by public authorities in this field is really, really reduced and just starting in the best of cases.  
Now, there are various challenges.  The first one is the rapid obsolescence of the classification used by the classical statistician, by the national statistic institute.  I was involved myself six or seven years ago in the working group of the European Commission official statistical authority, Eurostat, and we did define new categories for the reform of the European classification of economic activities.  Seven years ago, it's already archaeology in this field, and the categories are already, for me, not sufficient.  
We need analysis of the evolution of the value chain.  For example, there is a lot of debate in the industry about the importance of the Windows release.  At one moment you release your film at VOD.  We have to underline the scarcity of public resources to collect data industry, in particular my organisation, we have absolutely no means to work on this fee currently.  And there is also one important issue of transparency.  It was already quoted by Frederic of transparency of the financial statements of the companies.  We are familiar with both U.S. and European financial reports published by the media groups, and I should say in Europe we are very much in delay if we compare with the quality of the information you can obtain from U.S. companies.  I do not say that they are perfect, but there is still a lot of work to be done in Europe on this.
I think we are facing three paradox with Internet and transparency.  The first one is when you ask Europe, the private companies, to provide more information, more detailed information on the activities.  We cannot do this because we are bound by the rules of publication of information of the Stock Exchange Commission, and we cannot give you more details.  It's a bit funny that rules which are supposed to protect investors are used to refuse to provide more information.  
Second paradox is that Internet is supposed to facilitate the circulation of information.  I think my general assessment would be that in this sector, it has increased the asymmetry of information.  It has increased the difference between data poor and data rich.  
And the third paradox is now young people view the lack of transparency of the audiovisual industry to justify the practices of individual pirates.  
I'd like to quote a Nobel Prize economist.  If there is market failure on the information market, there is a role for public bodies to try to answer to this market failure, and it's what we are trying to do.  
In Europe, one important step was done in 2007 by the European Union with the adoption of the directive on the audiovisual media services, which now regulates on-demand services, including VOD on Internet.  It's progressively implemented in the national Member States.  It's a very slow process, and this process implies that Member States will have to report to the Commission on the development of on-demand services.  It will be very slow.  It will be very basic information.  But it's better than nothing.  And the Commission is promoting quite a lot of studies on the aspect of on-demand services.  
One other development interesting in Europe is that since three years now, the National Statistic Institute obliged to collect business statistics on the audiovisual sector, meaning that we will have more on the audiovisual sector, but it is a very slow process and not very harmonized.  There is very few national initiatives on monitoring process.  I think one of the most interesting is the observatory.  The observatory of VOD created by the French film agency, the CNC, in cooperation, of course, with the provider of services, and where it allows them to provide quite a lot of very interesting information on the situation of the market.  
EBU has made some sense of the services of capture TV provided by the members, and at the observatory, we are trying to monitor the development.  
Now, to conclude, my proposal would be that the authorities, of course in cooperation with the stakeholders, should define a set of concrete and other rules to implement transparency in the audiovisual market, including, of course, Internet aspect, to create a developing field between the various stakeholders.  
I think we probably have to reflect on obligation of publication of disclosing of a minimum set of standard information useful to all the industry.  
I think also that the public institution activity that is collected on the Internet should increase the cooperation.  That is what we are trying to do with OECD, with the European Commission, and with the EBU.  Public bodies should have a role in democratizing access to the information.  It's what we are already do.  We process data, publish at a very expensive price by private consultant, and we make it available at most accessible price.  Not everything is free.  But I think it's very important to have this possibility for the last portion of public to access to data which are elaborated in general commissioned by the U.S. major companies or big advertising companies.  They should be made available to all sectors.  
Also I think private companies should be requested to be more transparent on the methodology, on the concept, on the wording.  Very often we see data, prospective data, we do not know how they've been elaborated.  We do not know was the content of this aggregated in figures.  
So this is my personal proposal.  It's not a proposal of my organisation.  And if you wish to know a bit more on what is succeeding on the VOD market in Europe, I invite you to look on our last report.  Thank you.  

>> ISMO SILVO: Thanks, Andre.  Your conclusions and proposals seem that unless there's lots of top-down activities, public authorities asking and demanding and regulating information from the actors, nothing happens.  You have long experience on collecting market data in the traditional media sector working for transparency.  How much bottom-up willingness for transparency do you see in this market that is openly for openness and transparency?  Will it happen?  

>> ANDRE LANGE: Well, I think there are right orders asking more and more and asking more and more clarification and transparency to know what is the potential separation of the work, the real audience of the work, the ones that have been distributed on the new platforms.  
I think there is also a need and a push from some big stakeholders.  Several months ago I was involved in a panel in Belgium with one of the leading telecommunication providers in Belgium, and he was showing figures -- not figures -- graphs without figures, trends but not data.  And well, I asked him why don't you provide the precise data?  He said I will do it only if the regulatory body asks us to do it because our competitor will not do it.  So only the public authorities can oblige to both the two basic providers, only the mediation of the state can oblige them to be more transparent in common interest to know what the real market.  I think the push will come also from the industry.  

>> ISMO SILVO: All right.  Thank you.  Thank you, Andre.  And now we turn to our next speaker.  That is -- who is head of Strategic Information Service in the EBU, European public broadcasters union.  Alex Shulzycki.  Please, give us comments from the public service point of view, public service broadcaster's point of view.  Alex.  

>> ALEX SHULZYCKI: Thank you very much.  Can you hear me above the din?  I think -- I don't know what's worse, this background noise or those vuvuzelas from the World Cup or whatever they're called.  
Anyway, let's face it, we are here.  There are a lot of different stakeholders.  We have a lot of different positions.  And sometimes those positions are going to be diametrically opposed, and there's not going to be a compromise.  What's going to happen in the end?  Policymakers are going to decide, and they're going to decide on what they view is the real situation.  And the real situation is defined on how we measure certain aspects of the Internet.  
So I'm going to continue to promote this cause of better, more transparency, better measurement, because at the end of the day, it's got very important consequences for all of us.  
Just a quick overview of what do we know about the Internet in terms of current measurement systems.  And essentially, there are three ways of doing it.  There is a site-centric based system, where the server in your company or your organisation is tracking every little detail of what's happening, the page views, where people are going, how long, very, very granular, good information.  What it's missing is we don't know who these people are.  We don't have demographics.  And at the end of the day, they're just cookies.  We don't really know if they're people.  
The other system is a user-centric system, and this is run by companies like Nielsen and comScore, and essentially, they have large panels in countries that take a representative sample and try to predict on the population what people are doing.  The good thing is we know who these people are.  We have demographic information.  But those are the two basic ways of measuring the demand side of the Internet.  
And then the third way, of course, is all of the advertising information.  Very, very detailed, and I think that industry is probably the most progressive right now in measuring what is happening on the Internet in terms of how ads are displayed.  The real issue now is video advertising.  How do we do it?  How do we measure it?  
So that's essentially where we are.  And on the consumption side, we're looking at three kinds of measures, basically, unique monthly visitors, how many people are visiting your Web site unduplicated; reach, and reach is one of the common things if you come from audiovisual side, you can understand reach; and then time spent, minutes.  Those are the three basic categories of how the Internet is measured on the demand side.  
What I want to argue here is that there are things happening, there are structural changes happening to the Internet or Internets now, as people are starting to refer to it, that is creating an urgent, urgent need to identify and develop indicators, and especially in light of two phenomenon.  I'll just make a quick -- two quick points.  One is on the market concentration issue, and another on mediated access.  
Let me tell you what I mean.  But the importance here is that policymakers need these tools, these indicators, these measures to make informed decisions.  So if you have a disagreement with your fellow stakeholders, hopefully theoretically the decision is made on objective, clear measures, and not on anything else.  
Market concentration.  In terms of the ad industry, we know from our friends at Internet Advertising Bureau that the top 50 companies control 90% of Internet ad revenue.  The top 10, the top 10 companies control 71%.  Now, that's a certain -- that's a certain level of concentration, whether it's -- it seems high.  Whether it's high or low needs to be benchmarked, but these measures need to continue and then be put into a framework of competition law, antitrust regulation, et cetera.  
Another measure of concentration that we've been working on -- and we use comScore data.  What we have found, our basic conclusion, 99% of all measurable properties in Europe have less than 1% reach.  We have duplicated this for other populations, individual countries, a more global scale, but that's another idea, another area to go into in terms of how to measure concentration.  
When you look at these top properties, media companies figure very, very prominently, and it's becoming more and more so the case.  The top Internet companies are the traditional media companies.  And this -- there's a lot of reasons for this.  Mostly it's because of the audiovisual content that they control.  
You've all seen that Cisco study that says by 2014, 90% of all global Internet traffic will be video.  And I don't know we have to argue that most of that video is going to come from professional originators, the traditional media companies.  By the way, half of that traffic is going to be high definition, according to Cisco.  
So we are going to see in terms of concentration mergers, alliances, acquisition within, among, between media Internet companies, and I will go out on a limb.  I think we are going to see a big one.  I think in the next 18 months, two companies, we will see an AOL/Time Warner, but this time done correctly, this time done with adding value to both companies.  But I don't think that's out of the question that we could see something like that coming up.  
So convergence, we'll accelerate concentration, and at first, I think these issues are going to fall within two jurisdictions.  One is the competition law and antitrust regulation, and we see that happening already.  And the other with national media laws, especially how broadcasters are governed.  
And again, the policymakers, the regulators, the antitrust authorities will need the right tools, the right measures to make informed decisions.  And just to give you an example, the basic -- in competition law, the basic first step is to identify the relevant market.  With converging markets, this becomes more and more difficult, more and more challenging.  
The next step is to assign market shares.  We don't have any market share for the Internet.  We can talk about reach, minutes, visitors.  But what is a universally accepted market share?  I don't think we're even there yet.  
So the imperative to measure market concentration convergence is very strong, I feel.  
The second trend that I mentioned is mediated access.  And three forms of this:  Gatekeepers, paid content, and network discrimination.  I think we're seeing the rise of proprietary commercial fiefdoms, platforms that are going to result in closed controlled territories within the Internet.  How is this happening?  Applications on smart phones.  Smart phones and mobile Internet, the most rapidly growing part of the Internet.  Internet connected platforms with widgets, et cetera.  Internet connected TVs, a very strongly growing market.  Finally, apple TV, Google TV, which is both hardware and software.  When these are rolled out, this will add to this phenomenon.  And what's the measurement issue here?  The data coming from these platforms is proprietary.
It's a closely guarded secret what people pay, the number of subscribers, what your expansion plans are.  So this will become a measurement issue in terms of the proprietary nature.  
Paid content.  All major media companies are looking at this seriously or are already implementing it.  Rupert Murdock famously said I will rather have fewer that pay, and the effect of all of this we will see.  But the best study I've seen is by The New York Times, which conjectured what would happen if we made our content free, and the result is 90% -- they would lose 90% of their unique monthly visitors if people had to pay for it.  
So that's -- paid content is the strongest form of mediation, of access.  
In the end, the infrastructure will have to mirror this change in market structure, and if you were in the previous -- the very well-attended previous session on net neutrality, you know the issues involved, but there are some strong indications that this is becoming an empty concept.  We see the Google/Verizon proposal that essentially exempts wireless Internet.  In the UK there is very strong consideration now for putting laws in place, allowing ISPs to use network discrimination policies to create new business models.  And we have the arguments from the telecoms.  These are very strong economic arguments, and they're very strong technical arguments, and I think there are going to be so many exceptions to what we think of as true network neutrality that it will become meaningless.
It is very difficult for policymakers to stand between eager buyers and sellers, those that want to sell a premium service and access, and those who want to buy it.  
So I'll just -- I'll finish up here.  If we're moving toward excessive concentration, if we're moving towards severely mediated access, we need to measure and we need to monitor to make informed policy decisions.  And we need to consider new policies.  Again, well informed new policies.  And my constituency is public service broadcasting, and it is -- I think the public sector, the public sphere in general is important because they are the -- they get washed away by these trends.  They want to -- their imperative is to provide free access on an equal basis to all.  So these new policies we might consider, I don't know, an Internet version of must carry.  I don't know.  I'm not making any formal proposal here.  But any new policies and determining how ongoing policy is made requires better monitoring, better measurement.
I'll stop there.  Thank you.  

>> ISMO SILVO: EBU member organisations, the big public service companies, have moved on Internet towards Internet with speed during the last years.  How pleased are you, Alex, with them being able to provide you transparent information on their own activities on Internet?  Or is it just a bane within even the EBU family to get some comparable data and understand what is happening amongst the members?  

>> ALEX SHULZYCKI: You hit the nail on the head with the word comparability.  In fact, in a way, it's overmeasured in some cases.  In any national territory, you will have Nielsen figures, you'll have comScore figures, you'll have an independent research body giving out figures.  You'll have site data.  Even when there's a consensus within a national territory, if we pull that all together from other countries, it's not comparable.  That's the bottom line.  
So we see -- we see a willingness from the individual broadcasters to provide information, but then it's difficult to give them back something and to put them into the context of comparability.  
By the way, public broadcasters in Europe are very strong on the Internet.  The reach is average between 20% and 25%, which, in light of the first statistics that I gave, is very -- it's very strong.  And I think it's essential that policymakers continue to have public content impactful in that way.  If it gets washed out and marginalized, that's our biggest fear.  

>> ISMO SILVO: Right.  Well, now we have our next speakers come from the public service organisation, public service broadcaster of Holland, NPO, and they are telling us how they measure Internet services in Holland.  We have here Jeroen Verspeek, who is Head of Audience Research Department in NPO, and then Marije, who is specialized researcher on Internet issues.  So the floor is yours.  

>> JEROEN VERSPEEK: Okay.  Maybe it's easier to get the remote.  I will try to start with a view of the topics I will shortly talk about.  First, tell a bit about the market situation in the Netherlands, then give an overview of the measurement instruments we use.  Next, we will get into the methodologies and indicators we use on the Internet.  A lot of them are about the indicators Alexander just mentioned.  And although we will talk a bit about -- a shift from focus of concept to time spent, I will also get into research we started on the next step for us.  That's from time spent to needs met.  
And apart from getting insight in the reach of our services, we also measure our performance to the public in what we call the impact monitor.  And finally, I will show some examples on the monitor.  
So first, this sheet gives you an idea of the situation in the Netherlands.  We have a fairly advanced convergent media market with high digitization.  Although countries in Europe -- some countries have a high penetration on digital TV, we have a high supply of TV channels because historically we have a high cable penetration.  But when it comes to Internet penetration, we rank among the leading countries, and mobile Internet, as you can see, is quickly following this trend with the 25% penetration.  
So because nearly everyone is online in the Netherlands, the use of Internet services in mostly every sector and part of society is common.  In nearly every aspect of our lives, the Internet plays an important or supporting role, at least for the majority of the population.  And when it comes to media services, the last five years or so, Catchup TV  has become an important part of the strategy at NPO.  
We actually launched our Catchup TV service nearly at the same time YouTube became a success and two years ahead of BBC's iPlayer.  
So historically, we are radio and television broadcaster.  In just a decade, the instruments we use at our Audience Research Department have more than doubled.  Originally, we just focused on old-fashioned radio and television measurement, but in the meantime, having just two continuous audience measurements for NPO proved to be insufficient.  So we had to collect other data, as you can see on this slide.  We had to collect data from all sources available, and continuous instrument instruments have become part of our tools.  
We can use a lot of sources to measure Internet and mobile Internet quantitative, especially for mobile Internet.  None of these sources are yet 100% reliable.  Internet in the Netherlands is measured user centric since a couple of years with a panel called STIR consisting of 10,000 respondents.  And we combine these outcomes with Web analytics from a service from NetStat and now comScore division.  
As an addition, we also use server locks and data of suppliers such as DM delivery or YouTube.  And for mobile Internet, our audience research is still in its early stages.  Partly it's measured with Site Stats as well, partly with supplier and operator data, and we can use server logs as well.  But we still lack genuine reach monitor or measurement instrument.  
As a group of people using mobile Internet is still relatively small, this is also in the Netherlands still in its infant stages.  
Next I'll zoom into the methodologies and indicators for the Internet, and it's partly what was talked about by Alexander.  Last year the Maya project released a white paper which called for transparency, consistency, and standards, and the WAA in the Netherlands formed a task force for online media, and they published a white paper which aims to give this transparency for the Dutch market.  But I guess it can be used as a model for all markets.  
In the paper, we try to establish transparency by explaining the different methodologies and the causes of differences in output.  By offering a model for standards, we hope to achieve consistency in the use of each methodology as well because we don't think when it comes to these methodologies there will be one methodology or one system.  
These, in general terms, are the basic difference between the three tools for Web indicators.  In the Netherlands, for us the most confusion lies in the difference between the Web analytics and the STIR, the panel data.  And the Web data indicates that all three research tools are relevant, but they search each its own role for the different parts of the market.  
For us, as a public media -- public service media organisation, a focus on user-centric and site-centric indicators is most important.  We focus on public, so we use indicators such as unique visitors reached, also visitor profiles, to compare them with our TV and radio public, and viewing length.  This causes the use of different key performance indicators for each tool, and they are named here.  
But as you saw on the last slide, these indicators aren't exactly the same indicators we use for radio and television markets.  So when it comes to comparing our radio and television indicators with the new Internet indicators, it's still a bit of comparing apples with oranges, or as we say, apples with pears.  Some cultural differences.  
As we mentioned at the beginning, although we are in the middle of a shift of focus from context to time spent, we also have started doing research in the next step, and that's from time spent to needs met.  What do people use the Internet for, and what are the differences, for example, for different age groups?  
This sheet gives an overview of the different needs related to using the Internet at various moments during the day.  More detailed, it gives a closer look at needs met by the Internet for different age groups.  As you can see, where young people mostly have a need for information when surfing the Internet, they also have a need for togetherness, joining the social media later in the afternoon and the evening, and a need for entertainment during the evening.  
And that's different for older people, for which information is much more their key need.  It's -- Internet for them is less about the other needs.  So this kind of research is an altogether different kind of research when you compare it to the other indicators that gives us the idea of the role of Internet compared to our TV and radio services.  
Next, an idea of the impact monitor, which isn't a monitor only for the Internet, but in this monitor, we want to display our performance as a public service media broadcast for all our services, radio, television, Internet, mobile, all the platforms we are using.  
Apart from getting insight in the reach of a service, we also want to measure our performance to the public in the way our content is delivered.  We don't just want to reach our audience.  We somehow also want to know if these messages we display come across.  So this is the total grid.  And the two final sheets will give some short examples of outcomes of this grid.  
This is the first example for streaming video sites, where we compare our Catchup TV service with, of course, the major competitor, YouTube.  And at NPO, we always try to benchmark our performance to the performance of our main competitors.  And this short overview sees the reach of YouTube next to the reach of our Catchup TV.  But it also shows that in terms of reliability and in terms of interaction, there's a difference between both services.
They finally, an example on performance of our news site, called NOS.nl, compared to the major newspaper competitor, telegraph.nl, and the map only provider new, it's called now.nl.  Over here, you'll also see a difference in performance in interaction, but also in the banners as well, which we, as a public broadcaster, apparently have a higher score than both commercial operators.  
So this is a short case of what we do or try to do in the Netherlands.  

>> ISMO SILVO: Looks very impressive attempt, Jeroen, but just to be clear, we are still living the times, aren't we, that when you get one minute use of Internet or video and Internet from your tools, you cannot compare it with one minute of TV video use.  So there are different minutes.  

>> JEROEN VERSPEEK: Not yet.  We are still struggling to collect the minutes, actually, but when a person is one minute in front of an audio box or a television set or an Internet screen, the computer screen, you start -- you can start comparing them.  But then again, the needs, the underlying needs might be different.  But whenever a news broadcast is viewed one minute on a television screen or viewed one minute on an Internet screen, they start to become comparable.  

>> ISMO SILVO: All right.  Well, this is interesting and important because they tend to relate usually to the same content, the content that has come from television and the content that comes from on-demand service.  So somehow the minutes need to be at least some way comparable.  

>> JEROEN VERSPEEK: Yeah.  Sometimes it's the same content.  Sometimes it's short form of a longer form television probably.  

>> ISMO SILVO: Well, minutes and contents are uncomparable, but then we have an example from a news room from a converged content, news content, and our next speaker is Professor Antonio Granado, who is teaching in the new University of Lisbon, but his daily job is as an online news editor of the Portuguese public service broadcast.  So give us an example from a converged content.  

>> ANTONIO GRANADO: Thank you.  Okay.  Thank you very much.  One of the things that I would like to talk here about is about journalism and about the quality of journalism and the quality of journalism is being impacted by convergence.  Convergence is one of the words that is dominating the speech about the media.  When we, in news rooms, talk about convergence, we talk about several things.  We talk about the convergence of media groups, as Alexander told you.  We talk about the convergence of news rooms themselves.  We talk about the convergence of platforms in this apparatus.  And we also talk about another type of convergence, which is the convergence between producers and consumers of information because that convergence is also happening on the media.  
And the quality of news is being impacted by these kinds of convergence.  Convergence in a certain way for journalism means less diversity of information.  You've seen in the last few years lots of transformations in newspapers, in televisions, in radios, in online media all around the world that changed into smaller news rooms, into less participation in world events, et cetera.  
Something that happened last week -- last year, for instance, that was unthinkable, just two years or three years ago, the Washington Post closed its bureaus in New York, in Chicago, and in Los Angeles.  So every news room is taking resources from doing journalism and using them or not using them, just firing them.  
The convergence also, in news rooms, means more work for journalists.  Journalists are doing more work every day, and also, if we want to talk about journalism as the fourth state, I think in this new media world, the idea of the media as the fourth state is no longer an idea that we -- that we can agree upon.  And that's -- that's bad for people like me that are journalists, but also that is very bad for democracy because journalism is not doing what it should do for democracy.  
Media owners, that's what I see in several countries, media owners see convergence as a way to spend less and to reduce workforce, and the fact is that in all news rooms, all news rooms are shrinking.  And if we think a little about what happened in the 1990s, for instance, I think that we can say that news rooms are not bigger than they were in the '90s, in the beginning of the '90s.  But we can say that news rooms are doing much more than they were doing in the '90s.  
I worked in a newspaper that was born in the '90s, and we were doing, in the beginning of the '90s, we were doing daily newspaper every day, and that's all.  And now the newspaper is doing lots of other things.  It's doing online media.  It's doing video.  It's doing audio, et cetera, et cetera.  And we are -- and we were less journalist and we are less journalist than we are in the '90s.  
There's a very interesting speech by the editor of the New York Times.  It was done in November to his news room.  You can watch it on the Internet if you want to.  And he says a very important sentence.  With less, you can only do less.  And that is what is happening in news rooms all around the world.  
And when we are talking about convergence here, we are talking about the ways to measure it shall to measure convergence.  We also have to think ways of measuring the quality of information in a converged newsroom.  
It is necessary to study the impact that convergence has brought to journalism, has brought to the work of journalists, to the daily routines of journalists because I'm absolutely sure that very much is changing in the way people get their information and in the way that newspapers get their information.  We need -- we probably need to have pilot studies in news rooms, some research done in news rooms to see exactly what is happening to journalism, how the role of journalists is changing in this convergence scenario.  We also need to do transnational content analysis of what is happening to new media in this convergence scenario because we need to know exactly if -- what is the role of journalists in the next few years.  And if we still -- if we still can trust journalism to bring us the news and to be -- to fulfill its role as a watchdog, as we are used to see the role of journalists.
The consumer habits are also changing, a very strong pace.  The mobile access to news is increasing a lot.  People want news immediately.  They don't want to wait till they are in front of their television or in front of their computer.  And there is also much more public participation.  Participants want to participate in the news process.  And that is profoundly changing the news that gets into people.  According to a recent study that you've probably heard about, in 2014, 91% of Internet use will be on video.  And we are talking about video here and the audiovisual industry, and also that will be changing a lot how people get the information, how journalism will perform in the next few years.  

>> ISMO SILVO: Thanks, Antonio.  Alex just said in his presentation that he would like to see at least reach and time and number of visits being measured.  You add on his list quality and impact.  And you add a lot of gray hairs to those who want to do harmonization, I think.  Do you have a view how to measure impact?  

>> ANTONIO GRANADO: It's very hard to measure the impact of that information.  That's not a quantitative analysis.  It's more of a qualitative analysis.  It has to be done in order for people to understand what is the role of journalism and if journalism is fulfilling its traditional role or not.  

>> ISMO SILVO: I tend to agree personally.  But now let's have -- turn to our last but not least speaker, Professor Preben Sorensen, who is representing here those who we want to measure and from whom we want to get indicators, that is viewers and users of our services.  Preben is representing or is here in his position as a Chairman of the  Alliance of Europe even Listeners and Viewers Associations.  So Preben, how do you want to be measured and weighed?  

>> PREBEN SORENSEN: I will try to do my best.  I think it's important when I try to represent listeners and viewers, customers occasion, that we should try to collect, present, and evaluate information that will be helpful and relevant to an ordinary man or woman in the street.  
I must also underline the precision of our organisation, at least, which is also underlined in the latest directive from the new commission saying that public service broadcast must have all possibilities to use all the new things that were coming with the new media and so on on all platforms and useful for the society.  That's also important for us.  
Anyway, in starting resource provision in the media field, one looks necessarily at where the money comes from, from public, from state, commercial sources, et cetera, and where it is invested.  The purpose of this is to establish some understanding of so-called fitness of a given relationship between provision and consumption.  
Yes, but on the consumption side, I think it's important to look carefully at the relationship between media habits and media needs.  So far as needs are concerned, one obviously has to be conscious of the dangers of being prescriptive or generalizing definitions.  It's clear that habits do not truly and follow fully needs.  Habits, in more concrete terms, although they enter into resource provisions in the laws of demand and supply, are themselves influenced by the practical -- practicalities of content, the prize of services, et cetera.  The consumer is not always the king and can, indeed, become the victim.  
Socioeconomic and socioculture terms.  These dynamics can lead to contentious -- I hesitate to say the word unfair outcomes in terms of actual media provision for population groups, which is why investment, cost, and benefit categories need to be subdivided for research purposes.  
For example, in the digital age, there can be citizen consumers determined by age, class, or other factors who may do relatively little use or benefit from investment in certain source of platforms offering participation opportunities, whether in line or recorded, linear or not linear, university broadcast or on-demand content.  
In relation to audiences generally or to particular groups, how realistic and credibly can judgment be made about the quality and value about particular programmes' content demand or information streams?  
Against this background, federal decision making or bench making -- marking process for rating need and adjusting resource provision, can consensus be established among stakeholders about appropriate indicators, or is everything just up in the air and down to the market?  Who can reasonably be expected, not just of the regulators, but of the leaders in politics and business, who set the parameters of regulatory action?  What are the respective roles of, A, solid data, and B, policy priority in establishing the existence and extent of public needs and then responding to and implementing it?  
How well understood in this context is the impact of technological convergence who may be difficult in racial or different socioeconomic groups?  What would be the research in this?  In an increasing global community environment, how important is it to rely on if not the actual institutions then the media decision-making values in different countries?  If not actual consistent, can comparity at least be achieved?  So again, I find it best to think about information that would be of value to the ordinary man or woman and also what information would be of value for the whole civil society.  
Professor Vincent Porter, our member of the Advisory Council -- Advisory Committee to the European Audiovisual Observatory, I have some proposals we can discuss.  For instance, what is the annual per capita expenditure per household or head spent on viewing and listening equipment?  Such figures should include computers for accessing online TV programmes.  
What are the annual consumption costs of electricity or whatever source to run this equipment plus, if possible, capital depreciation costs?  
What are the annual costs to access TV and radio services?  This should include TV license costs for public service broadcasts, subsequent costs for other services.  In addition, this will need to include costs incurred in order to support them.  Although it may not be possible to distinguish between the cost incurred for electronic communication services and incurred for online radio and TV services.  
Beyond this, there are cost incurred by viewers to way pay-per-view services.  In one sense, these are a subset of a subscription cost which might be further divided into basic rate subscription costs and premium rate subscription costs and pay-per-view costs which are likely to be introduced as a means of paying for video-on-demand services.  
On the other hand, revenue per user, or AAPU, which -- these costs across their network.  Thus, it is not a direct measure of consumer expense for different level of services, but rather a measurement or rate of return of the company's investment.  
The companies must also claim that this measure is commercially confidential, but the same cannot be said for public service broadcasters.  The former, therefore, might have to be estimated by a form of sampling technology, which might be maintained in order to achieve from commercial companies already in the field.  
Commercial community revenues for TV and radio services, both broadcast and online, should also be measured, should be subdivided into revenues from product sponsorship and product placement.  It may also be necessary to subdivide this further, especially with the assistance of revenues from domestic companies and revenues from nondomestic companies.  
Beyond costs incurred, we also need to know about the time spent by viewers in accessing audiovisual services.  Broadcasts that currently seem to measure total time spent watching TV or listening to radio, but we shall now have to add time spent accessing online services.  Some of these will be catchup services, but we also need to know time spent watching video-on-demand services.  
At this point, we also ought to consider including time spent watching DVDs or CDs, as these also are taking a share of the viewers or listeners day, and as the audiovisual market becomes increasingly fragmented, they increasingly compete with online services for our attention.  
It's important also that the discussion only concern online services, as most people -- as in most people's lives, the time spent using online services increasingly competes with that spent using offline services.  
Beyond this, there are how this time is divided between TV channels, broadcasters, the difference between audience share and audience rates.  For instance or for example, the audience rates during the course of a week.  It should be possible for ISPs to monitor these electronically, although it may raise privacy issues if they do.  This is something we can discuss perhaps later.  
There is, finally, also the issue of which regulatory body is responsible for all these services.  It should be possible here to establish a degree of correlation for the regulatory body which monitors the activities of audiovisual services, such as video on demand and the body which collects the data on the use by listeners and viewers of each of these services.  
We know that some regulatory bodies do this by a formal procedure.  Others ask service providers who keep audio tapes or videotapes of their programmes for a period of, say, six months after their first broadcast.  
I would also like to support fully somewhat the proposals from Andre Lange about the public authorities and stakeholders need to define a set of concrete basic rules to implement as transparency of the AV market and create a level playing field, as it was said.  And I also support -- there has just been a new proposal from the Cultural Committee of the European Parliament saying that EAO -- I'm sorry -- the European Audiovisual Observatory -- shall have means to go on and find out these things much more strong in the future.  I think that's a good idea also to take up in the Parliament.  And of course, of the Commission.  Thank you.  

>> ISMO SILVO: Thank you, Preben.  I take that your main message is that also viewers and listeners' associations want more complete information on the usage of Internet.  You don't see any privacy issues here concerned?  Have you been discussing this?  Traditionally, media has been studied very carefully, but if you take the same carefulness in studying Internet, are there any privacy issues concerned?  

>> PREBEN SORENSEN: I think it's possible to find matters to do this thing.  I think we already heard some exact samples here how we can do it.  I would also like here to underline what was said of our friend from Portugal that also, as listeners and viewers and our organisations, we see this development fewer and fewer journalists.  I mean, we have to discuss the technique also for the new media in the future.  

>> ISMO SILVO: All right.  It's high time we have questions from the audience.  I know it's been a long day here.  I think.  

>> I have been involved in the Internet policy integration in India for about 13 years now.  One of the things I would like to mention is we heard about OECD and Council of Europe frameworks for ICT measurements, and also one has looked at organisations like ITU, Task Force many years back after the visas, et cetera, and then, of course, the network readiness index, economic index unit, et cetera.  Most of these are based on numbers like so many computers per household, so many mobile connections, et cetera.  
Now, one of the things which I would say from developing countries' perspective mostly, but even from other countries' perspective, you should look at this type of framework to look at four different dimensions.  
One is instead of just using the term subscriber, which has been traditionally used in terms of fixed telephones and cable connections, et cetera -- in fact, even ITU world telecom communicators, which released in the month of May in India this year, they started using for some of the parameters the term "subscription" rather than "subscriber." So I may be one subscriber, but I have two connections, I am a single subscriber, but I still account for two subscriptions.  That's one aspect of it.  So subscriber/subscription, one differentiation there.  
Second is the number of users.  Now, in case of mobile phone -- in case of even fixed phones in a country like India, the number of users was typically much more than the number of fixed lines, which is also the case elsewhere, because in a single household or office or institution you would have that.  But in a country like India, because of the whole revolution right in the mid-'80s of the public officers, so the number of users was significantly higher than the number of fixed-line connections, especially from this public access point of view.  
And the other thing that we need to look at is in case of Internet connection, especially so in case of broadband connections, and that too, when you have anchor institutions like schools or colleges or in hospitals, et cetera, those types of places, the number of users to the connection ratio is significantly different.  So for example, the office -- in a big office you may be having a thousand people, but probably you'll have only one or two Internet connections.  So in terms of subscriptions, it will be two, but in terms of the number of users, it will be significantly more.  It could be in thousands oftentimes.  It could be hundreds that could be in schools.  
The other thing I would say is that especially in case of Internet, again, giving an example of India, and of course, I'm sure it also will apply in other countries, is there are a whole lot of people who, despite the fact that you may have other type of devices, you may have computers in local languages, software applications in local languages, et cetera, would still not be in a position to consume those types of content on the Internet or applications directly by their own usage because of other type of challenges, apart from accessibility, which is one thing which is being discussed in some workshops, there will be other challenges.  
So I would say that when we are looking at these ICT measurements, we should look at a framework of subscriptions, subscribers, users, and beneficiaries.  And beneficiaries could be even people who have not even touched something.  So for example, if my driver, he doesn't know how to use a computer, he is he is illiterate, but if he has something to do and I assist him in that endeavor, then he still becomes a beneficiary.  In the country like India, we are now in the process of establishing 600 thousand kiosks across 600,000 villages.  These types of things will actually have still -- a whole lot of people will be only beneficiary and may still not qualify as users.  Thank you.  

>> ISMO SILVO: Well, thank you very much for your interesting positions.  I don't know if any of our experts will want to comment on this.  Somehow in my ears, you're approaching, at least, with the beneficiaries question, the same aspect as the public service broadcasters are trying to find out what is their impact.  

>> FREDERIC BOURASSA: At the OECD, we switch to use subscriptions now instead of subscribers, like the ITU.  We did that at the same time.  It's true that we do not collect any more the number of users.  It's harder to have an accurate data because if you talk about user, you will have to talk about frequencies of usage.  And when you talk about subscription, well, it's a contract.  You sign a contract.  It's easier to.  But I think there's some studies made on usage of Internet, ICT at schools, done in other directorate in OECD, like in Education, in the Education directorate.  So they have a definition of user with the frequency added and with the type of usage as well.  It's very complicated to have good data on that.  

>> ISMO SILVO: Marije wants to add something.  

>> MARIJE ANDELA: Yes, to add on this discussion about users and subscriptions and all sorts of definitions, we conquer the same problem in Holland because normally all stakeholders, all use of data use Web analytics data.  That's the best reliable source normally.  And it all measures cookies, so browser sessions.  And it's nothing like people behind the screen, how many people we're actually reaching.  It's a totally different kind of story.  So that's why in Holland a couple of years ago we introduced a panel, the STIR  panel Jeroen talked about, and that measures exactly the number of people sitting behind a computer desk.  And we also get a lot of demographic profiles about those people.  
So for example, we, with the amount of websites, we exploited Dutch public broadcasting company, we reach 7 million people monthly in Holland, and if we measured it with the Web analytics tool, it's like 14 to 15 million.  So that's almost the amount of people there live in Holland, so it can't be right.  So that's why I think I can advise countries who want to know more about the people they are actually reaching to establish some kind of panel measurement tool.  

>> RENALDAS GUDAUSKAS: Thank you.  I am Renaldas Gudauskas from Vilnius, Lithuania.  My question is about another indicators of measurement, what we know and you presented today.  But my question today is about conclusions that we can make from these quantitative measurements.  What kind of qualitative conclusions we can make.  To whom of audiovideo area or industry of audience we can make -- make these conclusions from these numbers?  I think you understand my question.  Thank you.  

>> ISMO SILVO: Thank you for this very, very interesting and topical question.  This is exactly what we are here trying to discuss.  Anybody want to start?  Andre?  

>> ANDRE LANGE: I think that, of course, data makes sense only if you know what they mean or they've been elaborated on with definition and in which context.  And I think I will just give one example.  Alex referred to a problem of the definition of the relevant market in case of analysis done by competition authority in case of debate on abuse of dominant position.  It was a very precise case.  And this was underlined by Screen Digest two years ago when there was a debate in UK on the possibility for BBC and other broadcasters to launch the so-called Congru project, common platform between broadcasters to create common viewing services.
The Competition Commission in UK opposed to this, arguing that this would reinforce the position of BBC and other broadcasters on the VOD market when BBC was already in a dominant position with BBC iPlayer.  But as underlined by Screen Digest, they forgot to count the iTune stores in the calculation of the VOD market.  And according to Screen  Digest analysis, iTunes in UK is relating on demand market.  Always been very good in never using the word VOD, but if you take it from a purely definition of VOD, according to the European directive on on demand services, so iTunes stores with film and video services is VOD.  
So if you don't take a clear definition, precise definition, you don't take into consideration part of the market, and then you take a decision that kind of project was prohibited.  

>> ISMO SILVO: So at least one answer, part of answer to your question, is that perhaps those who make policies, policymakers and regulators, would be in the first row to benefit in this.  But we are here call for more wider transparency so that the industry itself will generate willingness for transparency, definitely.  I think at the moment, when we're seeing what happens on media-based Internet market, never has media industry done so much big talk and development project with so little information.  But on the other hand, it may be just one way to avoid regulation, one way to keep the regulators away, to be nontransparent.  
Anyone want to comment on this question further?  Okay.  

>> Maybe for us, not as a regulator, but as public service media organisation, for us it's about content on the one hand and the impact it gets and for the public on the other hand and the type of profiles the content generates.  
So we try to always make comparisons for our old-fashioned radio services and television services compared to Internet services because we have only one budget, and we can only dedicate resources to a small part of the programme.  So we want to follow our public, and whenever the public is more online and our content online, which is more impact, there is a reason for us to change our policy and to move the resources and maybe the journalists and the creative makers to the online world.  If the impact and the reach and the pace in which it moves to online is slow, then there's no need for us to redistribute the resources more quickly.  

>> ISMO SILVO: Preben.  

>> PREBEN SORENSEN: Yes, I would mention the results or the way we do it in little Denmark, where I am from.  Our public service radio, they get license fee in condition that every year they make a so-called public service report, and that's not a report about economy.  Of course, it has to do with that also, but it's a report about how they had measured the quality and also other conditions set in a normal four-year contract for this public servicing.  For instance, you have to show so much drama, so much news, so on and so on.  And a part of that is also the online service and other services through this public service station.  
So I know it's difficult to measure quality, but you can try to do it at least to have some measures and try to compare with the reality through the year, have you done what you should do in accordance to these rules?  

>> ISMO SILVO: Any further questions or comments?  

>> Can I just add one thing on this topic?  The overwhelming majority of measurement of the Internet now is used to enhance advertising, to sell advertising.  That's the primary reason for measurement and indicators.  And it's used operationally.  
Now, we heard for public service it's used operationally but in a different sense.  But I would agree with Andre that it's -- ultimately, it's for policymakers, and I think in the medium-term, it's going to be the antitrust and competition authorities who need this information because they're going to be making very important decisions on mergers and acquisitions that have a tremendous impact on the general public.  So that's what I would say.  

>> ISMO SILVO: The next question is there.  

>> Yes.  A couple of questions.  The first one is the fragmentation of the audience that is due to the multiplatform, a rise to the point where the significance of numbers is still meaningful or we are losing ones that we are going to minimum numbers on the mobile access, on the catchup TV, on the normal TV, et cetera, et cetera, there is a level below the one we cannot understand anymore the real trends.  
And the second point is I want to have a view from the panelists from the pact that -- on the fact that in some countries, like for instance, in Germany, the broadcasters are forbidden to enter into the online services.  There is a ceiling in the resources that they can spend and do the online.  Do you think that this makes sense in this kind of environment that you described today?  

>> ISMO SILVO: All right.  Two questions.  One related to the possibility of measuring small, fragmented audiences, and then one related to policy on public service broadcasters on Internet platforms.  

>> Just from my point of view on both of those, the fragment -- the system that we are most familiar with is comScore, but Nielsen is very similar, and their cutoff is 15.  It means that in a panel, let's say one of their larger panels has 50,000 people.  If they don't get 15 -- at least 15 responses, which is I don't know what percent, it's not counted.  So there is -- there is some, let's say, invisible part of the Internet.  
But what I think -- this phenomenon just extends the long tail.  As I said early on, most -- the top ten, the top hundred, top thousand properties are the real Internet.  The rest is a wilderness that nobody ever goes to.  So if there's -- if this long tail is extended further, we wouldn't be surprised.  
And of course, on the German issue, where the tendency to restrict public broadcasters from certain online or new media activity is very -- is very troubling, and frankly, it comes mostly from lobby from commercial broadcasters and others who stand to benefit from the absence of free quality content on the Internet.  

>> ISMO SILVO: All right.  Any further comments or questions?  

>> Just on the second question, just to follow what Alexander said, I think there's -- the politics of not letting public broadcaster go into the Internet is somehow menacing its future.  Because if you're not there, you won't be able to live in, let's say, five years.  

>> ISMO SILVO: So we agree that this particular question relates to bigger issues than measuring indicators alone.  
All right.  It seems that we are coming to the end of our workshop, so if that is the case, I just want to say at the end that -- that the intention is to produce a report on this workshop on the presentations and discussions and make it available first on the Internet site of this Forum, but then later on the Internet sites of the European Audiovisual Observatory, OECD, and the EBU.  And at least these three organisations are committed to furthering transparency and conferability of data related to audiovisual content on Internet platform.  So any one of you interested in joining or helping in this very challenging task, please contact these organisations.  
So thank you very much for your interest, and thanks to our speakers and panelists for their contributions.  
(Applause)


********