Why auto-refresh is a massive issue – a case study

So, I admit it … I’ve been banging on about auto-refresh most of this year.

It’s an important issue. Why? Well … there is huge interest around the area … my posts around the issue have been some of the most read things I’ve written in the past 3 years, and the emails I’m getting from people with their stories of auto-refresh and the rationale behind it have been some of the more eye opening I have encountered.

Some people have the viewpoint ‘auto refresh, so what … what’s the big deal, consumers don’t know, don’t care or advertisers don’t care.’

Not true. What auto-refresh does to site numbers is astounding.

Last week a website took auto-refresh off some of their sites. Great move and they should be applauded if they continue to keep it off. (however if they initiate it again I am sure a load of people will be happy to point out to the market what they’re doing and how it inflates their numbers.

What it did to their numbers was amazing. Just amazing.

– Their average daily UBs across their site remained the same

– Their page impressions dropped 40%

– Page duration dropped by 75%

– Session duration dropped by 85%

If this is happening on 1 site … imagine what impact it’s having on all the other sites – hundreds of sites – that practice auto-refresh.

Makes you wonder what is really being sold? How many of these impressions are being seen by human eyes.

And how much of the ‘engagement’ is just fudging figures through technology?


17 responses to “Why auto-refresh is a massive issue – a case study

  1. You sound surprised. I’m not even slightly.

  2. The appetite for impressions is what drives this. Advertisers want more so the publishers give it to them. Whether the advertising actually generates business seems to be beside the point.

    The perception is that more impressions = more business. Historically this may have been true but web users aren’t interacting with ads the same way they used to. Responses are down so the instant solution for the advertiser and media buyer…get me more impressions. Publisher doesn’t have any more inventory so what are they gonna do?

    There are tons of neat little ideas out there that advertisers can buy into to promote their product but because they don’t operate on a CPM model they just don’t get it.

    What would you rather pay for $30 for 1000 imps that might get you 2 clicks. Or $5 each for 6 deliveries of your message to qualified leads? It is possible. Advertisers need to start broadening their minds into the different channels that are available in digital. If there was less demand for impressions then publishers wouldn’t be forced to engage in this kind of rubbish.

  3. I know exactly which site this is, as it’s a competitor, and I watch their numbers every day. And I know its a tough decision for a publisher to make, as we made this same choice a few years back. I agree with Ben in applauding them, and getting honest about their approach.

    Based on the PI drop that’s reported above, and other case studies done recently, I’d say you’re probably wasting 40% of your media buy, as almost all of the auto-refreshed pages aren’t being seen by human eyeballs.

    If I was an advertiser or media buyer, and wanted to buy on a site that still inflates their numbers through auto-refresh, I would be stipulating that my ads only appeared on user-initated pages, not auto refreshed pages. If the site can’t do that for you, then go somewhere else that can, or go to one that doesn’t auto refresh at all.

    And don’t forget your ad serving costs – you are probably paying 40% more than you need to in costs, as 40% of your ads are never seen.

    With all the information that’s now out around this issue, would media buyers now have a tough time explaining to their clients why (up to) 40% of their budget is never seen by human eyeballs?

  4. talkingdigital

    brad – good point. numerous parties have contributed to the situation we have now with auto-refresh etc … it’s not just the publishers but everyone involved.

    too much emphasis on the cost, not enough on what the cost is actually buying.

  5. Although far from scientific, an easy way of ballparking the $ impact of Auto-refresh would be to calculate a % of total ad spend of FFX, NDM & ninemsn (I believe Yahoo is the only Tier 1 pub not using Auto Refresh – why they are not beating their chest about this clear USP on their main competitors is beyond me??).

    Usage of it varies, so 40% page increase across the board is probably extreme. But even if it was 10%, the $ disappearing into the the ether would surely be in the millions…

    Which is why getting auto-refresh turned off is going to be a significant challenge as it will ultimately impact bottom line in a big way, not to mention reputation, positioning etc.

  6. talkingdigital

    It would be great to get an understanding of the IAB’s position on auto-refresh

  7. An announcement from IAB Australia re autorefreh – standards and guidelines – will be out within a couple of days.

  8. This subject is a teacup in search of a storm.

  9. talkingdigital

    i don’t agree one bit. What would happen to some of the networks inventory if auto-refresh was taken off? đŸ˜‰

  10. talkingdigital

    Liam – can you explain why it’s such a teacup looking for a storm (or whatever it was)

    do you disagree that

    – autorefresh artificially boosts impressions (ie what advertisers buy)
    – inflates session duration (ie the engagement many sites trade off)
    – means that loads of impressions are never seen by humans (ie loads of wasted advertiser investment)

    how can this be a non issue? is it perfectly okay for advertisers to be mislead that all their ads are being seen?

  11. Hello cobber

    On point one, have no idea. Not much I imagine?

    On the do I disagree pieces.

    I dont know what artificial means in this instance?

    Session duration? I know what this means but I can count on one finger the amount of times I have had customers tell me they bought based on session duration.

    Loads of ads go to non-humans? Probably. Loads of ads got non-humans all the time. Its called the cinema.

    I dont see any advertisers mislead. I am sure all media buyers know what auto-refresh is and what page impressions are.

  12. Liam,

    Hypothetically, if the sites using auto refresh removed it and saw a similar result to the sites in Ben’s case study (eg. 40% drop in PI’s). Wouldn’t that significantly limit the available inventory available for ad networks to sell?

  13. Hi Neil

    Nah it wouldnt.

  14. Unfortunatley I can only guess that the site(s) in question did see a major drop in inventory this week, and may have had trouble delivering booked campaigns, as it looks like they have put the auto refresh back into the site.

    And according to Market Intelligence figures from Wednesday, the PI’s increase signifciantly and session times are back up again – largest in the category. Its a shame, because I thought they were taking a big step in the right direction by removing it, but it shows that they need to show delivery of the ads (even if 40% of them are never seen).

    Liam – I think there’s no doubt that auto refresh boosts ad impressions (inventory) artificially. I disagree with you that “all media buyers know what auto-refresh is” – in my experience most have no idea what it does to their media buys, or which sites do it. Hopefully these discussions will help bring this issue to the fore.

    What I find crazy is that media buyers negotiate hard to get the best placements and rates for their clients, and then place the ads on pages that auto-refresh, where (up to 40%) of the ad spend is wasted!

  15. Neil Ackland: you highlight the issue rather well. The problem is you’ve taken the most measurable advertising medium in history and taken away the ability to measure it in a meaningful way. It’s important that we just don’t know how many of those impressions are initiated by humans. For my money, that means the impressions are worthless.

  16. Statement from IAB Australia:

    IAB Australia Measurement Council moves on auto refresh guidelines

    IAB Australia today announced that its Measurement Council is drafting guidelines and best practice for the application of auto-refresh following a review of the practice, measurement and reporting of auto-refresh of web content in the Australian market. The guidelines, which are being drafted in collaboration with the MFA and AANA, are expected to be published for industry wide comment by the end of March 2010.

    IAB Australia’s initial position is that, in line with global industry guidelines, site-set refresh rates are assessed for reasonableness based on the refresh rate and the nature of the content being refreshed. Consumers expect and want updated information on sites where content is being regularly updated, such as news, sport and weather.

    The review acknowledges the importance of providing agencies and clients who are buying advertising on these sites more detailed information to help them make informed buying decisions.

    The IAB is working with measurement providers, primarily Nielsen Online in the Australian market, to provide transparent data and to split out auto-refreshed page details from other page impression data. Nielsen will work with publishers to implement code that makes this data separation possible. IAB Australia will also encourage publishers using other site centric measurement solutions to provide this information to clients.

    IAB Australia is committed to the growth of the Australian interactive advertising industry and through its various Councils; and in collaboration with other key industry groups such as the MFA, is working to provide advertisers, agencies, and its members with global best practice and local guidelines to make advertising online more effective and more efficient.

  17. talkingdigital

    yes rhys it is unfortunate they have decided to put auto refresh back on.

    i guess the numbers are out there now – ie, numbers with and without AR so they’re in the public domain and can be accessed and used.

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