The new Nielsen Netview panel has gone live – everyone chuffed

So the upgraded Nielsen Netview panel has now been officially switched on and the results are causing a lot of high fives and delight in the world of the publishers.

It was commonly agreed the Netview grossly underestimated the audiences of most publishers, which was due in part to the inadequacy of the panel size and nunerous other issues.

Nielsen worked to correct the issue, and in consultancy with the IAB and other industry bodies increased the amount of people on the panel and invested in an upgrade of meter technology. The result is the new data which does tell a better story for the wider industry.

It’s hoped that this will provide more accurate measurement in the area of reach, engagement and user information.  I’ve had a play around this morning and one thing is for certain, all sites are going to report increased overall reach across their networks which will appease them no doubt.

Let’s hope all publishers stop using Market Intelligence figures now the Netview panel is more robust (given Market Intelligence claims there’s anywhere between 40-50m Internet users in AU)

Below are the July figures for 8 of the main players, in brackets is the increase on June figures (which were the old panel) 

ninemsn – 9.718m (up 26%)
News – 5.597m (up 34%)
Fairfax Digital – 5.317m (up 49%)
Bigpond – 5.402m (up 44%)
Google – 12.374m (up 22%)
Yahoo! – 6.851m (up 41%)
ABC – 3.237m (up 48%)
Facebook – 7.7m (up 50%)

My takeout is this – sites which have a high proportion of daytime usage (ie news, sport, some social networks) will see big benefits from this. Sites which are more popular out of work hours will see a minimal change with the new panel.

The key thing I want to see as an agency guy is more robust measurement around consumption – time spent, repeat visits, user information.

To be honest, adding another 20/30/40% of users won’t do much for any of these sites in terms of a commercial gain. They already had such large user figures so adding another 1-2m on top of 4-5m really doesn’t make any difference. Many sites were reporting massaged MI, Google Analytics or internal figures anyway, as many didn’t like how Netview was reporting their numbers.

Regardless, most of the initial spin on these figures will be around audience numbers as that is the pissing contest many of the main publishers think they are competing in.

Personally I am very interested in the differences that will be displayed between home and home and work numbers – ie, what sites are popular at what times of the day and how we can use this info to integrate digital better into other media strategy.

The key questions still remains around engagement and impact. From a brief look, this appears to have improved significantly. I would elaborate more but my Netview account appears to have gone back to the old June numbers as I was using it.

Anyway – the official launch is tomorrow.


29 responses to “The new Nielsen Netview panel has gone live – everyone chuffed

  1. Wasn’t the knock on Google Analytics that they were over-reporting numbers by 20%? To me, this just strengthens my belief that you don’t need to pay Nielsen five figures per annum to tell you what you can get from the big G for free.

    If Google published monthly country-specific Web traffic rankings, they could run Nielsen out of the Web business altogether.

  2. Paul,

    You can’t really compare GA to Nielsen here because this article is about their NetView panel.

    The panel estimates actual people, not “browsers”. Google doesn’t have an audience measurement service (to de-duplicate home and work usage, discount cookie-related inflation, provide demographics etc).

    Also, don’t forget Google is still a publisher and is included in these rankings. I wonder how other publishers would feel about putting a competitor in charge.

  3. The main significance of the Netview upgrade is that we have an audited panel (which is a first in itself) that gives us robust people-based audience data. We now have visibility of the demos behind 950 online brands (compared to old 250) in one place. This is the most accurate data Australia currently has and supported by several industry bodies.

  4. How does a panel of approx. 7,000 people (maybe 2,500 homes at best) show a good representation of visits by Australians to millions of websites? As mentioned previously this benefits a handful of large trafficked sites which we already know about.

    A panel is a great measurement tool for a few free to air TV channels but for online it falls short and really does nothing to show visitation to vertical sites.

    As a publisher of 11 vertical news websites I am always looking for a way for agencies to see our sites. Hopefully the new ABA digital audit service works in this area. It looks like it will.

    At least it is improving and until there is a better option we should applaud this upgrade.

  5. Kerry, I would beg to differ. Our (comScore) panel of 30,000 Australians is the most accurate in this market. We have not actively courted AU and the bodies, but are starting to.

    What can a 30,000 panel do for you? We report down to 3,388 properties, and 4,297 web domains. 950? Puh…leez….

    You will start to see more of us Down Under…. and we don’t just track urbanites in a couple of main Australian cities either.

  6. Karl,
    Nielsen are working towards a ‘hybrid’ system where the tagged sites in MI will plug into the panel data to fill out the missing niche/long-tail sites and any other volumetric info that will be more accurate with the census data.

    It is probable that the ABA-audited sites will also plug in to hybrid – so that’s another place for audited sites to be verified and seen.
    btw thanks for the mention.


    re: comscore. For the benefit of anyone who couldn’t attend the panel discussion yesterday – the MFA expressed concerns about comscore passing the grade while it is limited to measuring an audience aged 14+.

    In addition, Nielsen is a few years down a path of being heavily scrutinized and system-audited. That’s a lot of catching up to do for any other providers.

  7. Pingback: Last weeks highlights were reader comments – Comscore and Fairfax « talking digital – Ben Shepherd

  8. Hi Jason,

    The reason that I say “the Nielsen Online panel is the most accurate people-based data we have seen in Australia”, is that it is the only service that has undergone an independent audit and subsequent review by the IAB Australia’s Measurement Council (of which the MFA is a key part). That is, we have no basis on which to assess your claim.

    Regarding the issue of sample size, the MFA recognises that larger sample sizes allow you to drill down further – a key benefit, but only part of the story. Our PRIMARY concern is representativeness. All too often what is in a data base is taken as gospel – in some cases the sample size is looked at but rarely is the representativeness (i.e. what does that number really mean) taken into consideration.

    Therefore, in conjunction with the MFA’s research consultant John Grono, I’d like to ask some questions, which would assist in understanding ComScore’s service, and hence your claims to being the most accurate in the Australian market:
    – Can the MFA safely assume that you represent the entire market, from the very youngest (infants) Internet user to the oldest? (I don’t think so as you only report 14+)
    – Given that the official ABS statistics are very sparse (or current) regarding Internet penetration and actual Internet usage, how do you convert the “Australian population” into the “Australian online population” to draw your sample from and to project to?
    – More importantly, how do you ‘control’ your 30,000 panellists to ensure that they reflect this online population at a raw level before the data is weighted? As we all acknowledge, weighting will only magnify any biases and not correct for them.
    – We’re also interested to know more about how you recruit your panel. We understand it involves “affiliate partners” and that respondents get ‘free things’ like screensavers etc. We also appreciate that any affiliate site is not reported on as they are prime candidates for over-reporting. Can you please provide detailed information on how your 30,000 in Australia were recruited – for example, how many were locally recruited (i.e. using local sites), how many are affiliate, were there any quotas applied, etc.
    – And probably most importantly, the MFA is very interested to see the “health” and the composition of your panel. One key issue is that with many of our local major publishers having very specific geographic footprints (a legacy of being newspaper publishers) the geo-distribution of the panel is critical to us. Given this, it is critical that both the recruitment and the weighting of the panel reflects geographic distribution accurately. Any other composition guidelines you can provide (age, gender, educational, occupation etc) would be appreciated.

    I stand by my statement until Comscore can sufficiently answer the needs of the Australian market. We look forward to your arrival.

  9. Sorry to jump in late on this one.

    Paul. I’d be stunned if GA was only over by 20% when talking about “uniques” (browsers vs audience). When you aggregate all the tags in MI you get the Australian online population as 51m … that is 2.5 times the Australian population and at least treble the online population. UBs is a meaningless number – and increasingly so.

    Karl, the Nielsen Online panel is not “home based”. It is “people based”. That is, the sampling unit is people (as computers tend to be a single-person device) whereas televisions are a multi-person device so we sample the household (and all TVs in it). So, if we can assume that those 7,000 people live in different homes that is roughly one-in-a-thousands homes (not that homes are relevant as I pointed out earlier). For sampling, that is a good ratio. Somewhat better than the ratio used for a blood test! It is via very careful selection of respondents that we can represent all facets of the Australian population. There is one caveat though – Government & Defence, and Finance are EXTREMELY hard to measure – which is where the hybrid comes in to pick up the missing traffic. As Alex says, with ABA traffic audits merged in with Nielsen panel with the hybrid, we will be able to produce audience estimates based on the traffic to very low volume sites (aka “the long tail”).

    Ben, the biggest gain was not from the larger sample. It was that the work usage for people who have BOTH a work and home connection was being missed. A panel has been created to “plug this hole”. That is why the daytime usage is up.

    Jason, not surprisingly, I concur with Kerry.

  10. John: Now UBs are meaningless? Isn’t this a highly self-serving argument you’re pushing, given that you want publishers to pay you to manufacture this confection you call “audience”?

    It seems to me that an entire industry is being created around something that Google could do much more accurately via purely technological means. Or Hitwise, perhaps more pertinently.

    Take, for example, Google AdSense. Click fraud is a big issue, sure, but it is priced into the system. Google does what it can to combat the most egregious fraudsters, but it doesn’t pour huge amounts of money into validating every last click as that would not be economically sound, and that’s not their policy. The market only has a certain amount of supply (by publishers) and demand (by advertisers), and Google spending dollars on click fraud is just a tax on the market as a whole.

    As a publisher, I am yet to be convinced that a similar system would not work for Web site analytics. There are only so many dollars in the market. Sure, advertisers may not know as much about their audience and thus may get less revenue per impression, but they wouldn’t have Nielsen siphoning dollars out of the market, so with lowered expenses per impression they could buy more and would probably come out ahead.

    I may come off as disrespectful, but this is the Internet, this is the way Google has shaped our thinking out here. This is the attitude you will cop in the marketplace. I would be interested to hear your response.

  11. A very interesting POV Paul.

    Let’s be clear right up front – the only thing I am ‘pushing’ for is increasingly accurate AUDIENCE metrics, in the interests of the common link that publishers and media agencies have – the CLIENT.

    I do NOT “want publishers to pay you (i.e. me) to manufacture this confection you call “audience””. While I am a paid consultant to the MFA, I am not paid by any research company either in Australia or globally. Yes, I have worked in research companies in the past, but not for over 12 years.

    Paul, I also wasn’t aware that audiences were mere “confections”. What an amazing way for a publisher to look at the very thing that keeps them in business – lose your ‘confections’ and you’re in deep trouble. The objective of any communications plan is to connect with audiences in order to try to affect (positively) their behaviour and /or attitudes – i.e. to buy something.

    Let’s look at what a UB is. It is referred to as a “Unique Browser”. Most people take that to mean that the “browser” equates one-to-one with a person (i.e. the noun “browser”). Wrong. What it really relates to is the action of a “browse” having happened (i.e. the verb “browse”). We then use cookies to try and make them “unique”. What we need is UA (unique audience) – how many PEOPLE were hiding behind all those cookies. UBs overstate the real audience and sometimes by many orders of magnitude. You seem to have a very cavalier attitude to this overstatement of uniques, as well as to click fraud (well it will cost a lot of money to ‘get it right’ so we’ll just factor it into the pricing).

    Let’s also not forget that GA, Hitwise and the plethora of (in the main excellent) site analytics software are completely server-centric. They analyse the traffic to a site on the server side, and they measure that TRAFFIC very, very well. But, they have very limited knowledge of AUDIENCE because they rely 100% on cookies. If an advertiser is using traffic as their target they have a problem!

    Whilever there are people on the other end of your sites, we will need panels of people – i.e. your consumers, your audience – to establish WHO is using which sites. (Let’s park the issue of whether they are engaged or not – that is actually NOT the job of an industry audience currency as different ads engage at different rates with the impacted audience.) By all means use your site analytics to see how they move around and use your site – all great qualitative stuff that I love.

    Clearly, panels will never be as good as server-logs to quantify the traffic. Just like server-logs will never be as good as a panel for people-based metrics. What we need, and are heading towards, is a “hybrid” of the two. Using tagged traffic merge with the panel to understand the demographic composition of the traffic.

    I know everyone will scream BT at this point, but communications planning remains demographically focussed largely because marketers are always trying to break new ground with new products and brands that create new paradigms. You can’t BT when there is no existing behaviour! You can of course when you are running what is essentially a DM campaign for existing well-established brands.

    What I find extremely disappointing is that you see independent audience measurement such as Nielsen (or any other research company) siphoning dollars out of the market. Last time I looked the online market place is approaching $2b, surely there are a few dollars down the back of the lounge for independent market quantification.

    As a media planner and buyer we get all sorts of numbers thrust our way – and oddly, they ALL seem to have that publishers site as #1 in some way. What we need is and agreed independent currency. If you were a buyer would you see Google as independent?

    In order to keep online growing and to get he big advertisers involved, you need a CURRENCY. Currencies cost money. Most currencies cost around 0.5% – 1% of the gross billings. Online should be looking at $10m-$20m p.a. to ensure they get a currency. If you want to pay peanuts you will get monkeys (and I think you may have named some). I think it is time for ALL sectors of the online industry to show the maturity of a $2b industry and break-free from the cottage-industry mentality of some that holds it back.

    Paul, I didn’t see what you wrote as disrespectful. I just found it lacking in understanding of (i) how communications planning works (ii) that we need audience data and not traffic data (iii) that said audience data needs to be independent data (iv) that you need a currency that will cost all of you a few shekels – look at it as a sign of confidence in your product and your industry.


  12. I think it’s also important to note that the launch of Nielsen’s new NetView panel was the product of 2 years of work that was intitiated by a cross-industry group of 27 people representing 17 companies – publishers, agencies, advertisers, adnetworks, and the industry bodies representing 75 national advertisers, media agencies, creative agencies and the interactive media industry, totalling well over 800 member companies all up.
    An online audience measurement blueprint was created by that group clearly stating what the industry wanted and needed and invited responses.
    There were 4 responses to that blueprint, from Nielsen, Comscore, Roy Morgan, and Amethon.
    Each was then invited to submit their methodologies to an independent, IAB Measurement Council – appointed auditor.
    To date, only Nielsen has submitted to that audit.
    The audit made 16 recommendations and Nielsen has worked and continues to work closely with the IAB MC to address these 16 recommendations both in response to the stated industry needs and also to move towards full IAB accreditation.
    As John has written, the industry wants a currency with which they can plan and buy online advertising.
    The launch of the new Netview panel and data is a very strong and positive step in that direction and should be viewed by the online advertising industry as a significant milestone on that journey.

  13. Hi

    As a quick thing when you are talking about Google Analytics as a solution – the fact is that GA is a fundamentally flawed system in terms of actual analytics. It is based on 3rd party cookies (as ad systems etc are) and therefore misses out on correct figures by a reasonably large amount (about 20%).

    I know that many companies love GA because it is free but you get what you pay for in this instance and it would not be suitable for monitoring the traffic etc as described above.



  14. Thanks John, good to see we can have a civil discussion despite my provocative subject matter. 🙂

    I used the word “confection” after reading your sentence: “It is via very careful selection of respondents that we can represent all facets of the Australian population.” One man’s careful selection is another man’s biased stacking. I have read in the comments on this blog this week about how the industry has lobbied you to change your methodology to correct perceived unfairness to certain sectors. Is it worse that you had such weaknesses to begin with, or that you bowed to industry pressure to change them? It’s a confection, in my opinion, because in such a small industry, you’ve got pressures from people paying you the money to do the research to skew it this way and that, and the more I hear of you giving in to such lobbying saps my confidence that there is still enough pure science in your methodology to make it worth my money.

    And that’s my next point. You may think $10,000 (what I would have to pay to get full Nielsen service) is “a few dollars down the back of the lounge “, but to the medium-to-long tail that’s just way too much. The high price of your service means it will be denied to all but the largest of publishers – and will remain vulnerable to accusations of protecting the interests of those large publishers who pay for it.

    I brought up Google AdSense for a reason. Google has conditioned those of us out here on the Web to expect purely technological solutions for everything. It may not be greatly qualitative, but it works economically and we’re happy with it because it’s accessible to the long tail. (I have a lot of respect for Hitwise’s model, similarly, because it is also purely technological and covers the long tail.)

    I am a little surprised that Google hasn’t already attacked Nielsen’s business model, since advertising is Google’s main industry (not search). AdSense has already captured the low end of the Web advertising market, and it makes sense for Google to now aim for the high end, and Nielsen stands in the way.

    I am probably wrong in my assertions, John, in terms of quality of your product and the quality of service delivered back to the advertisers by using Nielsen. I’m only throwing out these points because I want to hear your defences to them, because I want to learn. This is the Internet, though, where quantity most often trumps quality.

  15. Great comments, very educational all, thanks.

    Very much looking forward to hearing back from Joe at ComScore to Kerry, if we ever do at all.

    Not knocking ComScore but I’m sure Joe has shared with us on a previous thread on this blog the supposed “Top 20 Australian Sites” or so according to ComScore and I can honestly say that from memory around half of them or more I have never heard of let alone visited! They were VERY US centric heavy and I find it difficult to believe that they so-called 30,000 “Australian” Panel is infact accurate.

    Just my 2c worth……

  16. Thanks for the bait, John.

    Kerry’s question and comments are very valid. We addressed many of these in the IAB MC meeting, and we have yet to hear any formal feedback from IAB. As we have said, we will undergo any audit needed but let’s make sure that we are on a level playing field, and not just being used.

    I would be very surprised if John has not heard of ALL of the top 20 sites I listed. The top 20 sites visited by Australians I mentioned before are (and associated monthly UV in millions):

    (drum roll please)
    1 – GOOGLE.COM.AU – 10,022
    2 – LIVE.COM – 7,822
    3 – NINEMSN.COM.AU – 6,876
    4 – FACEBOOK.COM – 6,102
    5 – YOUTUBE.COM – 5,782
    6 – GOOGLE.COM – 5,658
    7 – YAHOO.COM – 4,261
    8 – EBAY.COM.AU – 3,855
    9 – WIKIPEDIA.ORG – 3,833
    10 – MSN.COM – 3,556
    11 – MYSPACE.COM – 3,531
    12 – YAHOO.COM.AU – 3,269
    13 – MICROSOFT.COM – 3,096
    14 – BLOGGER.COM – 2,991
    15 – BING.COM – 2,724
    16 – EBAY.COM – 2,530
    17 – COMMBANK.COM.AU – 1,999
    18 – NEWS.COM.AU – 1,993
    19 – APPLE.COM – 1,866
    20 – WORDPRESS.COM – 1,748

  17. Hi Paul.

    I think you may have confused me with someone else. You say “the industry has lobbied you to change your methodology to correct perceived unfairness to certain sectors” … and … “you bowed to industry pressure to change them” … followed by … “you’ve got pressures from people paying you the money to do the research to skew it this way and that, and the more I hear of you giving in saps my confidence that there is still enough pure science in your methodology.”

    Paul, I don’t conduct or create the online audience measurement system (AMS). It is not “my methodology”, it is not “my research”. I don’t get paid one cent from any publisher. The methodologies and research belong to Nielsen, ComScore, HitWise etc.

    Where I fit in is research consultant to the MFA. I’ve been doing research for 32 years, of which the past 19 have been in audience measurement. I think I have learned a thing or two in that time. The MFA sits on the technical committees (not just online, but TV, radio, out-of-home, readership etc) to ensure that robust audience measurement is conducted. Research that has been tested and approved by people who know and understand such things.

    If anything Paul, you have the story completely around the wrong way. It was the MFA that originally lobbied the IAB to put an end to having a smorgasbord of disparate numbers in the marketplace that were being used and abused with an astonishing disregard for accuracy. This resulted in (i) the Blueprint (ii) the Nielsen panel audit (iii) Nielsen addressing the 16 issues raised (iv) the launch of the new NetView panel. At no point in time did myself or the MFA “bow to industry pressure”. If anything the MFA was putting the pressure on for change! Maybe you were at some industry meetings that I wasn’t at in which the MFA “gave in to such lobbying”, because having sat through two years of meetings, I didn’t see you at any of them. But then of course, you could be sprouting conspiracy theory hearsay rubbish.

    Running my own small businss I realise that $10,000 is a lot of money to a small publisher (again – it is not “my service” that charges this). It is $200 a week. A truism in life is that you get what you pay for. Clearly you’re the sort of bloke who buys the spanner set at the $2 shop instead of the hardware store, and then complains when it breaks the first time you use it. I suspect you’d also opt for the $500 website when launching a new brand rather than the $10,000 website.

    The MFA recognised that the “full kahuna” Nielsen service was out of the price range of the smaller publishers so was involved in and supportive of the Audit Bureaux service that was set up. It is “cheap as chips” I’m lead to believe (now I’m talking your language!) and I would urge you to tag your sites and get audited. While this system won’t provide detailed demographic data it will provide audited traffic data which is a great start. We’re hopeful that under a hybrid system, demographic profiles for smaller publishers could be modelled.

    I believe that the reason Google have not entered this space – apart from being the judge and jury (what click fraud I hear you say?) – is that they have purely technological solutions and not people-based solutions. Remember, we’re after unqiue audience and not unique cookies. Also, transparency of their algorithms is not Google’s strong suit, and transparency is something that all parties on the IAB Measurement Council demand.

    Finally, regarding HitWise (whom you say you have a lot of respect for), are you aware that they don’t include any Telstra traffic. What market share does Telstra have – maybe half? That is a LOT of the market to not represent at all. The issue is, if Telstra usage is like all other usage then it doesn’t matter, but one has to think that with “BigPond Unmetered” sites (e.g. AFL etc) that this makes a HUGE difference. This is why we insist on what you call “confection” – a carefully constructed and managed panel that represensts the online population, not just the bits it is expedient and cheap to do. By this I mean if 10% of businesses are “automotive” we need 10% of the business panel to work in the automotive industry. If 15% of the online ‘at home’ population is over 55 we need 15% of the home panel to be over 55. If 20% of the online population is in Sydney we need 20% of the panel in Sydney. (They are not the targets – just examples of the logic). If that is “bowing to pressure” as distinct from prudent research practice, then I am flummoxed!

    Sadly, you summarised the problem – “This is the Internet, though, where quantity most often trumps quality.” When it comes to an industry currency, that overly simplistic argument holds zero merit.

  18. I won’t add too much to this as I am enjoying John’s responses.

    Publishers need to feel comfortable with information about their site but the main audience for this type of data is media buyers / planners. They of course want to know how many people they can reach but often the most important factor is “who” they can reach. Apart from sites that have a high level of registration a robust panel is still the best way of supplying this information.

    The idea of a hybrid solution is very exciting and there will be a lot people from all sides of the industry helping to get the best solution into our market.

    Joe – I will be in touch this week.

  19. Thanks for the data Joe. Yes I have heard of all of the sites on your Top 20. Can you confirm what time period/month this is for?

    At first glance (it’s late here) I am extremely surprised that there is no Fairfax or Telstra site in the Top 20 – though this could be because of hierarchical/definitional issues. There are also some on the list that surprise me.

    What do other Aussies at the coalface think?

  20. The data is for the month of June 2009.

    This is our straight Web Domain report, rather than our Properties report. Bigpond came in at 22nd place and SMH much further down. The rolled up properties report looks like this for June 2009:

    === Properties UVs
    1 Google Sites 10,674
    2 Microsoft Sites 9,470
    3 FACEBOOK.COM 6,102
    4 Yahoo! Sites 5,563
    5 eBay 4,857
    6 Fox Interactive Media 4,271
    7 Apple Inc. 4,048
    8 Telstra Corporation Limited 3,977
    9 Wikimedia Foundation Sites 3,848
    10 AOL LLC 2,955
    11 News Interactive Pty Limited 2,939
    12 Fairfax Media 2,921
    13 Ask Network 2,764
    14 CBS Interactive 2,653
    15 Amazon Sites 2,170
    16 Commonwealth Bank of Australia 2,083
    17 Glam Media 1,861
    18 WordPress 1,766
    19 New York Times Digital 1,728
    20 Viacom Digital 1,654

  21. Thanks Joe.

    I will digest when time becomes available again.

  22. “Google doesn’t have an audience measurement service”
    “I believe that the reason Google have not entered this space …”
    “If Google published monthly country-specific Web traffic rankings, they could run Nielsen out of the Web business altogether”
    “I am a little surprised that Google hasn’t already attacked Nielsen’s business model”

    How can you work in this space and not know about the FREE Google Adplanner?

  23. Simple David.

    It is not audience measurement. It is traffic monitoring.

    How can you work in this space and not know the difference between audience and traffic. Scary shit that !

  24. Semantics.

    Google measure audience and traffic.
    They buy in panel data and add it to the information collected via search, Toolbars, Google Account etc. As such, I estimate they have access to a more rounded set of data than either Comscore or Nielsen.

    “If Google published monthly country-specific Web traffic rankings, they could run Nielsen out of the Web business altogether”. Adplanner allows you to do this.

    “It’s hoped that this will provide more accurate measurement in the area of reach, engagement and user information”
    Reach: adplanner does this
    Engagement: Adplanner does this
    User information: Adplanner does some of this

    I’ll admit that Adplanner doesn’t (yet) have all the functionality that Nielsen have with their breadth of products. But are you telling me that Google Adplanner in no way impinges on Nielsen business model? If you do then you are not using it right.

  25. talkingdigital

    The idea of Google being a legitimate audience measurement tool is ridiculous. They actively compete against the sites that they measure.

    The measurement currency needs to be impartial. What next, Fairfax Digital ad planner? Ninemsn Ad Planner?

  26. Thank you Talking Digital. At least this conversation now recognises Google is part of the mix.

    My work here is done.

    P.S. How many other media measurement systems are impartial? Let’s start with press circulation figures…

  27. talkingdigital

    i’m not acknowledging adplanner as even in the same space as nielsen and comscore, you have that wrong.

    adplanner in AU can’t be seriously considered as a tool.

    re impartiality. nielsen and comscore are both impartial. press circ goes through morgan which i believe is impartial too. Grono – get in here.

  28. Thanks for the opportunity Ben. (And as if I wouldn’t have leapt in anyway!)

    This is not a semantic argument. Saying the difference between audience and traffic is semantic is like saying the difference between black and white is semantic (just ask Daryl Somers … oops … sorry!).

    Just to clear the air as to where I fit into this jigsaw please allow me the indulgence of the following credentials and full disclosure. I run GAP Research which is an independent media research consultancy, and have two clients – ASTRA (i.e. subscription TV) and the MFA. In my role in the MFA I sit on the technical committees for OzTAM, RegTAM, CRA (radio), MOVE (out-of-home) and The Newspaper Works (readership tender). I have been in research for over 30 years (owww!), which included time at Nielsen which I left 12 years ago to move agency side.

    [One small thing Ben, readership is done by Morgan, while circulation is done by the Audit Bureaux of Australia. And yes, I sat on the committee that redrafted the audit rules a few years ago.]

    David, Google may well “buy in” panel data (OMG, don’t tell me there is something Google can’t do), but what is the data that they are actually getting. There are three primary needs for panel data (i) demographic composition (ii) duplication – over time and between sites (iii) cookie deletion over time. An “off-the-shelf” panel will (with varying degrees of accuracy) address the demographic issue, so a tick of some description there. An “off-the-shelf” panel may or may not address issues of duplication – I’d need to see and understand the panel to make judgement, so maybe a tick or maybe a cross. An “off-the-shelf” panel will not address the most important issue of all which is cookie deletion. In order to track/understand cookie deletion this must be done at an OS level, not at a browser level or at an ‘opt-in’ panel level … so a cross.

    Impartiality, transparency and robustness are essentials in media currencies. That is why the MFA sits on these research committees – it certainly isn’t for the witty repartee over a beverage or two let me assure you. The key to these currencies is that they are a collaborative effort. The seller (the media companies) meet with the buyers (the media agencies) and the clients (the advertisers) and thrash out what needs to be measured, what we would like to be measured, how we measure and report it, and finally how it is all going to be paid for. In all of these committee meetings (all held under NDA to protect IP) nothing (apart from financial issues of course) is ‘out-of-bounds’ for discussion, questioning and investigation. I’d like to share more of those sorts of things, but I simply can’t – but they DO happen.

    Similarly, I would also love Google to approach the market with the same transparency, and would love to “look under the hood” to see what makes the engine tick, but Google (understandably) jealously guard their algorithms. However, the price for industry currency is the sharing of those ‘secrets’ with the industry representatives on such committees. This is not what has happened. We have a “free service”, with its workings known to no-one outside that organisation that has been launched into the marketplace. Google are quite at liberty to launch such products – but they can’t become currency until they “pass muster”. Why risk $2b of ad-spend on an unknown system, no matter who owns and runs it, just because it is free?

    In all research (if not life) you get way you pay for. I think it was David Ogilvy (though am not sure – please feel free to correct me anyone who knows) who said, with research everyone wants high accuracy, quick turnaround and low cost … you may choose two of the three. This applies equally in the analogue and digital worlds – it always has and always will. A “free” service is the ‘ultimate’ in low cost, so it either isn’t really free (i.e. has hidden costs to claw back expenses), or is very slow to produce, or is not of sufficient quality. I’ll let you make up your own mind.

    Finally, I would like to raise what I think could be another interesting issue.

    David, you wrote “Thank you Talking Digital. At least this conversation now recognises Google is part of the mix. My work here is done.”

    This seems to imply that you had an agenda – something along the lines of “get GA talked about”. David, may I ask who you work for? Why your ‘evangelism’ for GA? Is there anything in this for you? Or do you just like free-stuff? If it’s the latter I can send you some links for “free cool screensavers and ringtones”.

  29. Now whether or not Google’s method and methodology passes muster is an entirely different matter. Indeed, after the litany of acronyms you reeled off, I’m quite happy to believe you.

    My point was simply that quotes like the below;
    “Google doesn’t have an audience measurement service”
    “I believe that the reason Google have not entered this space …”
    “If Google published monthly country-specific Web traffic rankings, they could run Nielsen out of the Web business altogether”
    “I am a little surprised that Google hasn’t already attacked Nielsen’s business model”

    …and, knowledge that Google Adplanner uses panel data (however poorly) showed a lack of knowledge about developments in the space from some of the participants of this conversation. You may argue that flaws in Google’s approach mean that you can ignore it; I would argue that Google’s entry into this market is the biggest development that has occured for years and therefore, you need to know about it.

    I don’t work for Google and have no particular bias except from perhaps that the accuracy of some of these tools (and therefore to some extent the value) is overstated. For this reason, I probably fall into your “just like free-stuff” group. “Can I send you some links for “free cool screensavers and ringtones” – throw in a monthly prize draw and I’ll be on your panel!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s