Looks like some progress is being made on the ABA’s web audit … maybe …


According to reports, last Friday the ABA, IAB and MFA all sat down and talked about the increasingly heated topic of auto-refresh and the new ABA guidelines – http://www.auditbureau.org.au/web_audit_services.php

The MFA had come out in December of 2009 urging MFA members to demand ABA compliance from their suppliers.

Reports are, the IAB and its board members weren’t happy with this approach and the issue became a delicate one. Hence, the meeting last Friday.

Mumbrella is reporting that all groups are close to a resolution – http://mumbrella.com.au/peace-talks-held-over-autorefresh-rates-for-websites-16289#comments

I covered the issue a week ago – https://talkingdigital.wordpress.com/2010/01/14/what-is-happening-with-the-abas-web-audit-for-au/ – and IAB CEO Paul Fisher made some comments.

Reading back all of these articles the issue still seems cloudy. I’ve heard numerous reasons for why the process hasn’t been adopted by the larger players – everything from cost (which cannot be true given the costs to audit) to labour required to the wider issue of measurement of all elements of digital media.

What is clear is that there’s some frustration out there around the issue – particularly from advertisers and media buyers. Another thing that appears to be clear, is that there is a perception that many suppliers inactivity around the issue is due to the fact it’s not in their interests to stop the practice in terms of inventory and engagement metrics.

The excuse used for auto-refresh is that it provides the user with the most up to date news – which is debatable. Why? Well – loads of the sites practicing it are practicing it on pages that don’t update with new content, and pages that update so infrequently the page would need to auto-refresh 100 times to get anything new.

Plus sites like Twitter and Facebook auto-refresh without generating a new page impression – so the technology argument appears a limp one.

The main question that needs to be answered is – what would happen to the non compliant sites user numbers if they turned off auto-refresh tomorrow?

What impact would it have on
– page impressions
– session duration
– pages per unique

So … if the ABA, MFA and IAB are all sitting around chewing the fat around transparency, may I suggest some other issues that could be brought up at the table?

– stop auto-play on in article video. Fairfax are a particularly enthusiastic adopter of this practice and it’s increasingly annoying. Would be more respectful to the user if these in-article videos were user initiated. It’d also give advertisers much more honest, clear metrics around usage.

split out traffic numbers across sections. It would be handy to know traffic figures around index pages, article pages, slideshows etc. For TV we can generate these sorts of numbers.

– discussion around stronger metrics around die-hard, regular and passer by users. This is something available in the US and helps advertisers get an idea of how loyal an audience is. I’d imagine if Quantcast can do it, a Neilsen or equivalent should be able to do it. Given SEO and SEM is still vital for many sites to generate traffic (often shallow engagement traffic) it would get an idea of how many people enter via a front page or visit regularly, and how many are entering via search – reading and then leaving straight away

some guidelines around clutter. DO we really need 10 ads on a page? With pressure on yield increasing the issue of clutter is relevant as throwing another ad on a page is often an answer to the question of generating more yield. Problem is – how many ads can a user feasibly take in? And doesn’t placing another ad on the page just devalue the other ones?

IP/location reporting through Nielsen. Impossible to split out where users are coming from through any of Nielsen’s tools for agencies … so we never really know how many of Brisbane Times users live in QLD … or how many The Age readers are in Melbourne … which makes things difficult as we know this for all TV shows, radio stations and magazines. Sometimes feels like this data is something the sites often don’t want out in the public (as most of them have it through Site Census and other anayltics tools)

Anyway, the debate so far is encouraging. Let’s hope it moves from discussion into action and change.

Advertisements

12 responses to “Looks like some progress is being made on the ABA’s web audit … maybe …

  1. Hi Ben,

    The impression I got speaking (mainly on background) to some of the key players on Friday is that they recognise from both sides of the table that they now have to tackle this.

    It’s also something of the testament to the influence of blogs, at least in this particular part of the industry, that everyone I talked to had read your original piece – and many seemed to view it as a major factor in finally getting this on the agenda.

    And I thought the publication of the case study from the ABA was ballsy behaviour from an organisation that hasn’t always been famous for it.

    I’m sure there will be delaying tactics from the major players, but it does feel like this issue has finally got some momentum.

    Cheers,

    Tim – Mumbrella

  2. Hey Ben. A great post and great ideas.

    I could go on about the history behind the “Great Aut-refresh War of 2009-2010” … but I won’t. Suffice it to say that the discussions have resulted in a common desire to resolve this issue amicably to everyone’s satisfaction as soon as possible. That is, we all agree it needs to be fixed pronto … we just have to nail what “fixed” really means so that we can develop a set of rules that can’t be ‘gamed’.

    Taking one of the issues you raised – “auto-play” (which also drives me up-the-wall as a user) – this is to my way of thinking a variant on “auto-refresh” – both are non-user generated file requests, and our objective is to quantify consumer behaviour and not server behaviour.

    As you know Ben, the IAB has a Measurement Council, the MFA has its Digital Committee and the ABA had it’s Watchdog Committee. Given (i) you’ll be relocating to Sydney soonish (congrats on the new gig btw!) (ii) have great knowledge, insight and ideas on these issues … I can see a very active role for you on some of these committees … wanna play?

  3. Hi Ben,

    Great suggestions! If anyone has others – add them in and we will start working through them.

    On your last point re: State-based reporting. That exists in Nielsen MI already and should already be free and visible for all MI subscribers.

    It is located under the Geo-target tab under the “MI – Australia Domestic” market domain (where all categories are shown).

    Just used it for the first time – it’s good, it shows all the metrics for the state segment not just session %. Interesting numbers if you run your examples above …

    I have to applaud the publishers and Nielsen for showing this transparent data. Mark that one as done!

  4. Ben,
    The issue of auto-refresh still seems cloudy because it hasn’t yet been sufficiently tackled by all the relevant industry bodies.
    We’ve both been around this industry for a number of years now and the auto-refresh issue is one of a number that keep cropping up over many years – you highlight some others in this post.
    This is the first time in many years (ever?) the various industry bodies and in particular the individuals who contribute to their functioning are committed to resolving this issue and any others in the interests of continually growing the industry.
    I’m really keen now that we can get all the major players around the appropriate tables – the IAB Measurmeent Council, the MFA Digital Sub-Committee and the ABA Digital Watchdog Committee to tackle this and any future issues to come up with a workable and meaningful, sustainable set of standards, guidelines and best practice.
    We do need to bear in mind that these industry bodies represent literally hundreds of members and members’ interests, and meet monthly. Whilst I am confident we will reach some agreed standards etc, it won’t happen overnight.
    Realistically, we need another 8 – 12 weeks I think to resolve the auto-refresh issue given the number of players and the (in)frequency our respective committees meet.
    Given it’s taken years to get this far, another 3 months shouldn’t be too onerous a wait!
    As John stated, if you and any others are keen to contribute, the forums exist.

  5. Ben

    Just a couple more points from an IAB pov to add to Paul’s comments above.

    This is a tricky issue and one that the IAB Measurement Committee wants to get right. Defining when it is valid to refresh content (& ad) has not successfully (in my opinion) defined anywhere in the world. At the moment in this market we have Netview data which excludes refreshed pages and MI which includes them.

    The IAB want to make sure that we do not separate the issue of what is being served from what is being counted otherwise we will end up with a meaningless number. So the debate will now focus on when is it valid to refresh (type of content, duration) and then the counting/auditing will follow.

    One option is to follow the UK and track page impressions, auto refresh pages and ad impressions as 3 different audited numbers for total transparency.

    Compared to many markets we really do have a high level of transparency due to the level of site centric data available in MI (daily data, section data for many large sites etc).

    The IAB are committed to working with the MFA to provide them with the information that they need to make informed decisions when planning and buying.

  6. I agree with Gai – no-one has nailed the definition yet, but that is not going to stop us from trying!

    I think that one thing is clear – that it will NOT be a ‘black-and-white’ rule. To have a rule that says “all auto-refresh is allowed” istoo easily gamed by publishers. Conversely, to have a rule that “all auto-refresh is ignored” is too harsh for certain sites (news, sports, ASX etc).

    I think we can take a leaf out of the ABA print rules regarding things like ‘airline copies’ and ‘hotel copies’. To some advertisers, these have real value when targeting affluent travellers, whereas to other advertisers they have limited value. Therefore, we separated them along with other ‘classes’ of circulation.

    I can see the same approach working online. PIs could be grouped into ‘user requested’ and ‘server initated’ with a Total PIs also being generated. Of course the publisher will want to talk Total PIs for every site but that becomes part of the negotiation during the buying process. For a site which is not ‘breaking news’ as a buyer I would want only ‘user requested’. but for a site that is ‘breaking news’ then ‘user requested’ + ‘server initiated’ would be fair.

    As Paul has alluded to, we need to get out heads together as seller, buyer and umpire and work out what the rules need to be and how we’re going to implement them.

    So far the comments posted have been IAB, ABA and MFA. Surely there are other pundits out there sufficiently interested to post their thoughts … are we heading in the right direction or have we missed something? Speak up while you have the chance!

  7. Be very interested to know a little more from the buying community how impression counts for each publisher or network impact the buying decision making process.

    Accuracy is very important and frustratingly challenging in digital to execute. What seems easy to measure is often extremely complex to achieve to execute as the CMS, tracking, serving and optimsiation systems all vary and integrating across platforms throws up technology challenges.

    It would be great to understand where in the heiracrchy of buying decision making that auto refreshing sits.

    That would be super useful to understand.

  8. Just for clarity, from my own experience I have not heard from any customers expressing auto-refreshing as a problem.

    Mostly the dialogue is about reach, audience composition, price, price again, ad size, placement etc.

    I don’t quite understand why anybody is spending time on this.

  9. talkingdigital

    because it impacts on what advertisers are buying – or what they think they’re buying.

    it affects the accuracy of data around the areas many publishers sell off – impressions, engagement etc.

    whats the point of buying an impression if no one sees it? what’s the point of counting hidden browser tabs as an engaged user?

    surely you can see some rationale behind this issue?? 😉

  10. Liam, I suspect that clients do not know about auto-refresh and the potential (and it is a potential as not all sites use auto-refresh) to be paying for Page Impressions (and not all deals are on PIs) that are not requested or are either ‘duplicate’ or even unseen. I also suspect that those clients would be far from pleased if they were affected – hence the commitment from the IAB, MFA and ABA to tighten this up.

  11. Pingback: A Sydneysider dressed as a Brisbaneite: a tale of audience deception? « talking digital – Ben Shepherd

  12. Edit: Comment has been removed due to the company the allegations are being made against believing the comments are defamatory.

    ‘Anonymous’, unfortunately Talking Digital can’t run this comment without a name and/or ID disclosure.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s