Audit boss: Publishers won’t web audit because they’ve too much to lose giving up autorefresh

The outgoing boss of Australia’s audit body says that major online publishers have been resisting getting their web traffic audited because they do not want to give up auto refreshing pages.


Gordon Towell, CEO of the Audit Bureaux of Australia, was asked about autorefresh on this week’s edition of the Mumbrellacast. Towell, whose last day with the ABA is today, said:

“There has been resistance. I think it’s the major factor. I have no doubt about it. The problem is that for years, people have counted their page impressions in a certain way. The advertisers pay against the numbers. If we switched off autorefresh overnight it could have a profound impact on traffic.”

Although the issue of autorefresh has gradually moved up the agenda, the Interactive Advertising Bureau – which represents Australia’s major publishers – is yet to take any definitive action on autorefresh. Most recently the IAB put in a proposal to traffic tracking company Nielsen to offer an indication of autorefreshed numbers in its data.

Those who defend autorefresh argue that in some cases – such as a frequently updated news home page, autorefresh is justifiable. However, critics claim that where a page is open in the background on a browser, advertisers are being rorted because they not only pay for each ad impression but they also have to cover the costs of ad serving. They also point out that if the true purpose of autorefresh was to deliver new content to readers, there is no technical need to also serve a new ad each time.

The ABA’s web audit standards do not allow autorefreshed page impressions to be served.

(You can hear the full Mumbrellacast conversation with Gordon Towell here)

Meanwhile, The ABA’s online watchdog Alexx Cass has joined the debate over Mumbrella’s complaints that AdNews has been misleading the market in its own unaudited claims about its web traffic.

Cass wrote on Mumbrella this morning:

“Great to see a real change in the sentiment towards auditing (at least among Mumbrella readers). A year ago, ‘web’ and ‘audit’ were never used in the same sentence, but now the call for audited figures pops up more and more. But still a long way to go before it’s standard…”


  1. andy
    27 Aug 10
    1:18 pm

  2. “If we switched off autorefresh overnight it could have a profound impact on traffic.”

    Hahaha, i bet it would. A more accurate statement might be:

    “If we stopped artificially creating traffic, our advertisers might find out how many people actually see their ads.”

  3. Mike
    27 Aug 10
    1:54 pm

  4. Look at sites like facebook where only particular elements like status updates refresh. Surely this can be applied across news sites as well.

  5. Renai LeMay
    27 Aug 10
    2:08 pm

  6. To be honest I am kinda shocked this autorefresh issue is still hanging around … it should be incredibly obvious to advertisers that sites with massively high page impressions and only a small unique user base (and generally few comments or other reader engagement) are full of autorefresh crap.

  7. Mr. Dictionary
    27 Aug 10
    2:15 pm

  8. There – not “…their is no technical need to also serve…”
    Mr. Dictionary

  9. mumbrella
    27 Aug 10
    2:19 pm

  10. Thank you, Mr D.


    Tim – Mumbrella

  11. GreatStrategy
    27 Aug 10
    2:42 pm

  12. “profound impact on traffic” figures

  13. AB
    27 Aug 10
    2:47 pm

  14. This article comes across as a little dated, well a lot actually.

    Any reliable web site provider or owner would not give their page hits and page loads – and no credible advertising agency would request it.

    More appropriate are figures like Bounce Rates, Unique Views and Loyalty.

    I guess the author of this article is from the old rear guard of traditional based media taking yet another uneducated swipe at new media. At least websites have record-able stats – it’s more than we can say for “readership figures” of printed press and adverts in magazines.

  15. Mike
    27 Aug 10
    2:55 pm

  16. @ AB – Most agency RFPs include share of voice, which means the number of ad impressions against the capacity (or total), so impressions do count in that regard. And no one uses HITS anymore – Hits are what Idiots use to Track Sites!

  17. Zac
    27 Aug 10
    3:21 pm

  18. We’ve recently switched off auto-refresh (as it impacted our video embeds) and have seen no material reduction in page views. We used it on certain landing pages to provide additional content recommendations, not to inflate traffic.

    The only thing that has changed is our TOS has dropped a lot, from about 10mins to 3mins. But if a tab is open in someone’s browser and they jump through that tab throughout the day, then this SHOULD be counted as time on site (or some other metric) given the multiple views the page receives throughout the day which is now not being captured.

  19. Julian Time Out
    27 Aug 10
    3:25 pm

  20. I am pretty sure that some auto-refreshed sites would lose 70% of their page impressions.

    Unfortunately we cannot tell until they stop and audit.

  21. John Grono
    27 Aug 10
    3:30 pm

  22. Yes, AB websites do habe record-able stats. That does NOT make them useful stats. The Internet IS the most measurable medium for sure – it’s just that most of what is measured bears little relation to what advertisers want to know about … their audience. I prefer to call it the most measured medium.

    And from my unbiased perspective of working with all the mesured media, I would have to say that server-side site-centric metrics are the poorest of the lot. Sure readership has it problems, and yes we’re looking at ways to make the readership estimates more robust and stable. But when you look at Nielsen’s MI data using tags and find that the gross UBs for for Australia is 89m (could be more than that now) it just shows how laughable your claim is.

    Just because you can count things, doesn’t mean you’re measuring the audience with any degree of accuracy.

    And for the record, I believe that there are a handful of sites for which Auto-Refresh IS justifiable. Those publishers seriously need to look at ensuring such pages are written so that PIs and ad-counts are not inflated, but if they insist on sticking with AR for those pages then they must report the incidence and frequency whenever metrics for that site are used for sales purposes.

  23. Ben S
    27 Aug 10
    3:31 pm

  24. I couldn’t agree less Zac … a tab on in the background is not active so should not be counted as time on site.

    Digital needs an equal playing field for it to be taken seriously once the growth glow wears off. If you want to put your hand out and expect advertisers to offer up money for your site, do them the service of putting some rigour behind your numbers through Neilsen or ABA Audit.

    I guess right now there’s no reason to – agencies keep dumping money into unaudited/unmeasured sites. Why fix something if it’s not broken eh guys …

  25. Zac
    27 Aug 10
    4:09 pm

  26. Ben – the point I was making was that a site being opened in a tab will see it being viewed many times throughout a session (if my browsing habits are the same as others), but this won’t be counted if the session isn’t being refreshed.

    For example, I have my calendar open throughout the day in a background tab. To me, the TOS for this would be 9 or 10 hours, but if I’m just viewing the page from time to time it’s not counted. My guess is that many sites are also on tabs and people do view the pages throughout the day, but the limitation of TOS means after the first 5 mins, the session lapses.

  27. Alexx - ABA Watchdog
    27 Aug 10
    4:09 pm

  28. @Zac, I can see where you’re coming from but a core principle of audience measurement is that you should only count what you can be certain is user-requested content.

    An idle tab, longform media without interaction, an auto slideshow are all components that cannot be counted because there is no certainty that there is user activity beyond the start point.

    It’s not ideal, but it’s safer to slightly underrepresent figures than to vastly overrepresent figures by counting everything getting pushed out or sitting idly in unattended screens.

    This has come up a few times for review in our Digital Watchdog Committee so I understand the argument, it just doesn’t mesh with core principles.

  29. Alexx - ABA Watchdog
    27 Aug 10
    4:17 pm

  30. For your calendar example, if this metric is important it should be counted as what it actually represents – ‘Average Daily Screen Time’ not ‘Average User Session Duration’.

    We need to ensure that any metric is not counting different things among publishers and that media-buyers can easily interpret what is being counted.

  31. John Grono
    27 Aug 10
    4:33 pm

  32. Well put Ben and Alexx. Hey Alexx, I started typing about session durations in a previous reply but removed it thinking you’d probably end up commenting.

    Zac, there is a guideline in audience measurement that say you’re better to understate the audience than overstate it. For example Out-Of-Home discounts all the ‘traffic’ past their billboards, bus-shelters etc by a Likelihood-To-See factor because they know that not everyone sees every site. Also, the TV ratings only include viewing within a home that is done within a week of broadcast (to allow some catch-up time shifted viewing). So if a programme is viewed at the pub (footy GF), office (Melb. Cup), online etc it’s not counted – think of it as a bonus for the buyer. And regarding your calendar example, that would be the same as if we counted a Foxtel box that stays on all night even though the screen is off. Each of these media are diligent in weeding out “false audience”, rather than including it on the off chance that someone may have viewed.

  33. Rob
    27 Aug 10
    4:34 pm

  34. Issues surrounding measurement, accuracy, and consumer insight exist in every media channel in every market around the world. The reason why these issues feel worse in our market is because media owners are unable (or unwilling) to provide advertisers with proper visibility (with respect to the fate of their advertising dollar). Worse, toothless agency heads and trade bodies are powerless to rein in the behaviour of thes media cartels. Consequently, we are left with a ground swell of ill-will and cynacism but nobody to actually deliver chage.

    For instance, we get stuck in the symantics of what an ideal trading model would look like. Few people would argue that some form of user based survey/study would give a much more useful insgiht into the behaviours and attitudes of our consumers than ad impressions. However such a soltution has its own set of barriers and challenges – recruiting participants, ensuring the researched audience is truly representative of the whole audience, and how to accurately scale up the numbers to represent the total population. Right now, we are a long way from being able to trade our media against such ‘useful’ metrics so all we can do is continue trading against the basics (ad impressions) and ensuring some from of best practice is followed.

    the fact is, media owners have a responsibility to ensure every impression served is served to an active user (no auto-refresh). This speaks to the honesty and integrity of a publisher. However, it is far easier to take the short term win, ‘it is what it is and if you don’t buy it someone else will’ than to think about longer term objectives like transparency and accountability.

    Frankly, auto-refresh is just one line item on a list that could be accurately entitled ‘Media owners get away with murder because…’.

    Our failing as an industry has nothing to do with process and methodology. Our failing is that advertisers and their agencies are either too lazy or too weak to stand up to publishers and demand that our budgets be properly stewarded. How many times have advertisers had the IAB guidelines read to them chapter and verse at the end of a disappointing media campaign?

    10% under-delivery – acceptable
    10% discrepancy – acceptable
    Reconcile to publisher stats – acceptable
    The list goes on…

    The truth is, the only way out of this mess is for clients to vote with their wallets. They must demand their agencies end the pillow talk and deliver real trading solutions (NB: sitting in committess in the IAB and MFA does not equal a solution), and until such time as they do then all non-essential digital budget should be diverted away from the worst offending publishers and possibly even digital media in entirety.

  35. John Grono
    27 Aug 10
    5:00 pm

  36. Rob, if working in Committees to produce better metrics is not the way to a solution I would be very interested in your thoughts as to what is.

    The MFA can not direct any agency to not support a publisher who, for example, uses AR. That is illegal trade practices. Neither can the IAB direct any publisher to turn AR on or off. Again that is an illegal trade practice. Either or both would end up in ACCC action. The MFA takes the line that we can whinge and complain all we like, and threaten to move our ad-money elsewhere and that has achieved exactly zero. Or, we can roll up our sleeves and work on solutions that move the industry forward. I will admit progress has been painfully slow, much slower than the MFA wants or needs.

    Hence, what are your solutions? I would love to hear some concrete suggestions before we all decided to close up shop.

  37. Rob
    27 Aug 10
    5:18 pm

  38. Hi John!

    Apprecaite you’re trying to make the best of a bad situation so I hope you don’t take any offence at the above.

    I absolutely agree that it is not appropriate for trade bodies to influence trading strategy; I believe that in order to be a credible force in the market you must remain agnostic. It is the job of agencies, and the advertisers they represent to, define what is and isn’t acceptable when it comes to trading and accountabiity. Unfortunately this is the area where we fall down.

    I have given a great deal of feedback to my agency – who sit on your board – on this point. I have also offered to partcipate in the process if it is deemed helpful. I’m definitely not one to sit on the sidelines whinging, I’d rather roll my sleeves up and work towards a solution. That said, I really feel that we are past the point where an acceptable solution can be reached through diplomacy.

  39. John Grono
    27 Aug 10
    5:53 pm

  40. Rob, I may just have to call on your offer of help! If you want to take this discussion off line, you probably either know my email, or can get it from the MFA website.

    I would also argue that it could very well also be the job of the advertisers to include buying mandatories to their agency!

  41. Leon
    27 Aug 10
    6:21 pm

  42. @Ben S “Digital needs an equal playing field for it to be taken seriously..”

    Who isn’t taking digital seriously?

    There are many case studies of serious companies making serious money by
    harnessing the web’s measureability eg. dell, amazon, google, ebay, ing, amex to name a few.

    Also, don’t think the government would be looking to spend between $6BN – $43BN in infrastructure to deliver digital services if industry wasn’t serious about it.

  43. johnny
    27 Aug 10
    7:41 pm

  44. Why should we fix it when we all make more revenue with auto refreshed inventory.

    More ads mean cheaper CPM bookings, more adserving fees, more remnant inventory, more post view conversions and so on.

    I don’t know why we are even talking about it, this is what digital media is all about.

  45. Mike
    27 Aug 10
    9:24 pm

  46. It all assumes of course that companies like News are accurate with their traffic numbers. Which they aren’t.

  47. Mike
    27 Aug 10
    10:48 pm

  48. I’m the Mike who left the 2nd comment at the top of this thread. I can attest the Mike who left the last comment has no idea. I worked @ News for a number of years, and during that time not once was traffic overstated.

    Home page impressions were sold on a sponsorship basis, and at a considerably lower ecpm than ROS published on the ratecard – in line with the rate of refresh. Impression were based on Neilsen data, and on traffic over the preceeding quarter based on the average weekday figures.

    I don’t know what that ecpm rate is today, but I’ve seen no evidence to support what this guy is saying.

    @ Johnny, your view makes total sense based on the current reality..

    @ Tim, I understand this forum should allow posts of an anonymous nature for various reasons, but that also allows misinformants to post crap like this last guy did. I guess you can always check the IP to at least determine where the bullshit is coming from and moderate accordingly.

    I’ve had nothing to do with News for a couple of years, and am not posting in its defence. But I do want to take to task the validity of some views here. The technology is available to resolve this issue (refreshing content elements such as breaking news, while maintaining the original rendered ad image on the page), so why isn’t it being implemented?

    There are way too many agendas in this discussion for such a simple solution.

  49. Tech Focused
    28 Aug 10
    1:31 pm

  50. I think you will find that the Mobile channel is the most measurable medium, each mobile handset is attached to an individual and has a unique MISDN (globally). Mobiles measure everything down to the handset and firmware and when Apps are opened, coupons redeemed etc. They have a true CPA (Cost Per Action) model.

  51. Ben S
    28 Aug 10
    3:39 pm

  52. Printed coupons are pretty accurate too when you think about it.

  53. John Grono
    28 Aug 10
    4:15 pm

  54. Tech, I’ve pretty much thought that for a while as a mobile is probably the most personal device. There are some issues (solvable) such as people who own more than one phone (de-duplicating the contacts to get corrected reach and frequency).

    However, I have this nagging feeling that because it is the most personal medium then it is probably the medium that people least want to be interrupted with advertising. Further the small screen may reduce the impact.

    I also suspect that privacy issues side-stepped (or even ignored) in the computing world will loom larger on such a personal device. It should however be the most ‘monetisable’ medium as there is already a PAYG billing system which can be efficiently leveraged.

  55. mumbrella
    29 Aug 10
    1:46 pm

  56. Hi Mike (comment 24),

    I believe you that News Ltd does not deliberately overstate traffic. However, critics of the big networks would say that nobody needs to because they already have the artificial help of autorefresh. Arguably that in itself is the overstatement.

    And for as long as none of the big five are audited, the perception will remain that the numbers are manipulateable (is that a word…?).

    Personally, I think the first one of the big publsihers that ditches unjustified autorefresh and gets itself audited will give itself a commercial advantage in the market.

    After all the talk, the media agencies should feel duty bound to support the first mover. Private conversations I’ve had with one or two media agency people certainly indicate that they would.


    Tim – Mumbrella

  57. Julian Time Out
    29 Aug 10
    8:48 pm

  58. Latest ABC figures for UK newspapers – everyone is in there except News –

    When, not if, auto-refresh is removed from various sites in Aus – will advertisers claim back the proportion of what they paid that was not actually served to human viewers?

  59. Ben S
    30 Aug 10
    9:24 am

  60. The thing that is weird is this … People are are urging the large 5 to take action but at the same time not rewarding the smaller guys who have had it turned off since audit began.

    If the market doesn’t want to support those currently audited, why will they support the bigger guys any more (concentrating more spend in the large sites) if they choose to submit to audit?

    It’s all a bit confusing from the perspective of someone who is audited. We audit because we think advertisers deserve truly accurate numbers, problem is sometimes some of their suppliers don’t seem to worry as much about it.

  61. Alexx - ABA Watchdog
    30 Aug 10
    9:26 am

  62. Related to this story and the comments above:

    Maxus have just announced they are boycotting sites that are not ABA audited:

    This unprecedented stance, in addition to recent announcements that GroupM and OMD are giving preference to audited media demonstratres a real change.

  63. John Grono
    30 Aug 10
    9:38 am

  64. Hi Ben.

    I’ve been thinking about this for quite a while since we started talking these issues through. I can come up with just one conclusion (remembering I am an audience researcher and not a buyer). That answer is reach. That is the Big Five have mass (I know, AR overstates that mass – or at least over prices it), that in order to get bulk to a campaign a planner/buyer sort of has to include an element of them, or construct a complicated buy (who has that time), or use an aggregator (a less controlled tonnage buy).

    I’m not excusing that, or defending the policies of the Big Five, just stating that as a possible cause or rationale. Who knows, deals at that end of town might not be based on PIs and therefore AR is not really a consideration. I do know that among the smaller publishers, those that are audited ARE trusted. But still, the rules of the game should be the same across the board.

    Someone posted earlier that there could be an “initiator’ benefit among the Big Five for the first to recode their pages that justifiably utilise refreshing content and turning all other AR off. I 100% agree. To the best of my knowledge one of the large publishers is a long way down this track.

    My concern is that it is “not a good look” for the industry. With audience measurement very much in play, it is also vitally important that any ‘fake’ traffic is eradicated from the metrics, and that a solid bedrock is established upon which audience growth and trends can be accurately and reliably measured.

  65. Joel Pearson
    30 Aug 10
    10:42 am

  66. Big props to Maxus for taking such a firm stance on the issue.