It’s not just eyeballs, darling, it’s how much the eyeballs like it
nzherald.co.nz has signalled a strategic change in direction, moving away from page impressions as a predominant measure of a site’s success and towards metrics that provide greater transparency to advertisers.
Spencer Bailey, general manager at nzherald.co.nz, says the way page impressions are measured no longer effectively represents advertising impressions.
“This lack of transparency makes it very difficult for advertisers and their agencies to place a real value on the media they are buying. We aim to fix that by focusing on the quality of user engagement, not the quantity of page impressions reported.”
In 2009, he says it became possible for publishers to report a new page impression if half or more of the viewing screen was updated, without necessarily delivering a new advertising impression. As these in-page interactions became more popular, publishers have been able to report higher page impression numbers, without delivering any additional value to advertisers or users (StopPress has already waded into the at-times confusing realm of online advertising and particularly whether foreign eyeballs were a curse or a blessing).
Instead, nzherald.co.nz is focusing on improving truly important advertising metrics such as audience reach and delivering transparency to agencies and advertisers without hampering user design.
One of the measures nzherald.co.nz will be looking at is how readers can be encouraged to stay longer on each page. Improving the average page duration creates greater value for advertisers as users have more exposure to their advertisements. Domestic Average Page Duration at nzherald.co.nz was up to 71 seconds on the first full day of results since the change was made last Friday and, while a few weeks of data will be needed to see what the trends are, Sarah Marshall, marketing manager at nzherald.co.nz, says “common sense would suggest this will only continue to go up as we launch more features”.
Nigel Tutt, acting group head of digital at Fairfax Media says the company ultimately takes its direction on metrics from clients in an attempt to meet their needs. And the metrics for stuff.co.nz make for fairly good reading (if you’re from Fairfax).
“Since overtaking [nzherald.co.nz] in July 2009 as New Zealand’s most popular news and entertainment site, Stuff.co.nz has extended its lead to 321,812 domestic users to reach a massive 2.26 million New Zealand browsers.”
Over the last year, he says average daily domestic unique browsers has grown 43 percent, monthly domestic unique browsers have grown by 73 percent, monthly domestic page impressions have grown by 138 percent and average session duration has grown by 17 percent. 
Back at nzherald.co.nz, Bailey says the decision to change the model was simple.
“We could have competed in the muddy and often confusing battle for page impressions, or make a clear stand in the market with a major focus on delivering a better experience for users, and a more engaged audience for our advertisers. For us the choice was obvious.”
In other news, nzherald.co.nz has also been named as one of only five mobile news sites in the world to be awarded Official Honoree status at the Webby Awards (check out a few other Kiwi Webby nominees/honorees here)
The Official Honoree distinction is awarded to the top 15 percent of work entered that exhibits remarkable achievement, with Webby Award organisers saying the mobile Webby Awards Honorees “have shown the foresight and ingenuity to excel within a quickly changing and increasingly fragmented Mobile Web universe”. And with nearly 10,000 entries received from over 60 countries, Bailey says the award is a massive achievement for the site.
“To be recognised as an honoree by the leading digital awards in the world, alongside organisations such as The New York Times and CBS, is something we’re hugely proud of. To receive that recognition for our mobile site truly highlights the work we’ve done to provide innovative ways for people to access their news – whenever they want and wherever they are.”
Somer
April 20, 2010
Sounds like the old 'we're getting whipped, so we're going to change how we measure success' game to me.
Looking at audience obviously makes sense (though has always been the main metric used by advertisers), but the 'Page Duration' metric is non-sense since it's way to easily gamed.
ie: downgrade your servers, or reduce your page caching and you'll see Page Duration go through the roof instantly. Having terrible usability is also an outstanding way to increase Page Duration.
Point is – metrics with ambiguous causes shouldn't ever be used as a basis for allocating spend and the argument that people are spending more time on our pages means that they are more engaged just doesn't hold water (ask Trade Me).
Preston
April 20, 2010
Great point. Spoken like a guru. Talking of which, has anyone heard of Nigel Tutt? A quick search online suggests he ain't from the media – here or overseas. Which makes it an interesting strategic appointment from Fairfax, even if he is only acting.
Somer
April 20, 2010
I wondered that to – who is this mysterious Nigel Tutt? Sounds like a made up name to me.
Least he didn't talk about 'hits', I guess.
Digi
April 20, 2010
Last time I looked in Comscore the Herald were still ahead of stuff on domestic audience, so not sure one can really say they are 'getting whipped'.
CS
April 20, 2010
@Digi, when was the last time you looked?
I'm looking at ComScore's March numbers now and Stuff is ahead of Herald.
I'm comparing panel numbers only of course, since the ComScore unified numbers has not been widely adopted here yet.
Somer
April 20, 2010
Comscore? Really? – That's not even close to being real data yet.
Nielsen has it's issues of course, but making decisions based on Comscore is no better than using Alexa.
Course the ability to feed tagging data into Comscore will address that, but most publishers aren't doing it at this point.
Matt Harman
April 20, 2010
….and not to go on about it, but it would be hard to believe that when the tagging data was incorporated into Comscore, it would show a different picture than the comprehensive 'whipping' indicated by Nielsen.
Jono
April 20, 2010
isn't there a few big sites missing from the Nielsen report? Youtube comes to mind
Somer
April 20, 2010
@jono – thought we were comparing NZH v Stuff.
Preston
April 20, 2010
Is that the sound of the herald's Gold Sponsor renewal forms hitting the round file?
Margie
April 20, 2010
Nielson? Really?
Counting 10-15 page impressions for a quiz or photo gallery that people spend 3 sec looking at (not that the page duration is counted anyway) isn't a very useful measurement. And doesn't tell me how many ad impressions I have to plan my campaign with either.
Georgio
April 20, 2010
Sounds like this guy has been reading pages from Steve Jobs’ – “How we do it: A guide to user experience”. Finally!!! For too many years the on-line media cronies have pandered too muchto the ‘lowest common denominator’ not realising that more tech savvy, educated demographics don’t even bother clicking on those “Ads to NowheresVille”.
With iAd arriving in this realestate soon the game is about to change considerably and take this platform to A Whole New Level!
CS
April 20, 2010
@ Margie: If you're planning ad impressions off page impressions then that's already a problem there…
Margie
April 20, 2010
@ CS. Re read my post: Ad impressions for planning.
If sites are being transparent then page impressions should match ad impressions anyway.
CS
April 20, 2010
In any case, impressions have never been the single definitive measure of a website’s success. It is just one of many performance metrics.
So suggesting that one is moving away from “page impressions being the predominant measure of the site’s success” is pretty silly. What about the fact that a site may have more or less visitors/browsers (whatever) than another? That counts. Duration? That counts. Time of day on site? That counts. Demographics? That counts.
I'm pretty sure that any half-decent media planner will be asking their publishers for such relevant information when building their media plans
CS
April 20, 2010
@ Margie: A page impression is generated by the user. An ad impression is created by the site/ad-server serving an ad.
A single page on a site can host any number of ad impressions, and most sites don’t only host 1 ad spot per page. That’s ridiculous.
Saying that page impressions = ad impressions is like saying that the number of newspaper ads in a paper should be the same as newspaper pages.
Patrick
April 20, 2010
I'm not a marketing guy – but when I flick through the m.nzherald site – I hated clicking next page next page. in the end – just stopped using it as it was a hassle. m.stuff is much easier to use.
Somer
April 20, 2010
The NZH seems absolutely committed to "moving away" from any and all metrics that show they are behind the game and quickly losing ground.
You don't have to go back to far to remember how they went on and on about being the number one news site in terms of audience size….until they got beat.
Then it was engagement as measured by pages per visit….until they got beat.
Now Page Duration. Sounds to me like an excuse to load up the pages with even more huge, slow to load ads.
Tim P
April 20, 2010
TVNZ and TV3 have already shown leadership on this issue, don't know what all the fuss is about ! I'm sure the rest of the market will follow soon enough.
Toni
April 20, 2010
@Patrick Luckily you weren't a judge at the world renowned Webby awards then ! Pretty good achievement to sit alongside CBS, WSJ, Politico and TMZ, as voted the world's best mobile sites. but hey these things are subjective right LOL
Penner
April 20, 2010
Page Impressions cannot be equal to Ad Impressions for a number of obvious reasons – websites don't need to display the same ad for every page impression they can rotate and do rotate Ads on pages.
The Average Page Duration metric like any other metric is not a definitive measure of how successful or unsuccessful a website is. All other metrics also need to be taken into consideration to come to a rational conclusion. Its like I want a really fuel efficient car that gives me 50 kms/litre but, I don't really care how fast it can go so will you settle for a car that has a top speed of 30 kms/hr that would not even be road legal would it?
Finally if websites were to artificially increase their page load times how many people do you think will have the patience to wait for the whole page to load… If any page took more than a few seconds to load I know that personally I wouldn't stay on that page.
Somer
April 20, 2010
It wouldn't make a lot of sense to artificially drive up page load, you would think – but the fact that a metric can be caused by both positive and negative factors does suggest pretty strongly that Page Duration has low relevance when planning campaigns.
Also – does the fact that someone's more engaged with a page really indicate they are more receptive to advertising?
Surely the reverse would make more sense? (ie: when the content isn't engaging, i'm much more likely to look for advertising to interact with).
I've found we often get much better results (click through and conversion) from advertising on low engagement sites – which makes intuitive sense to me.
2 Cents
April 21, 2010
Buying, selling, measuring 'impressions' (ad, page or otherwise) is about as useless as tits on a bull – What advertisers want are 'audiences' and the more relevant the audience the more valuable the opportunity for and advertiser to reach those audiences should be. The notion of measuring "consumption" as opposed to "impressions" is a good one. We have been trying to convince publishers and advertisers alike for the better part of 10 years of exactly this so it not so novel an idea. Consumption should however measure “ad” consumption and not page or site consumption. Tell me how long a user actually views an in-page banner or display ad (if viewed at all) and now you have something valuable!
Publishers need to stop flogging 'numbers' and instead should be focused on providing good content and user experiences. The online space is about “Sociographic” affinity – i.e. a golfing website is not surprisingly going to attract ‘golfers’ and golfers are more likely to have ‘x’ characteristics and therefore more likely to have more affinity to products ‘y’ and/or ‘z’. For large portals this gets difficult because you can’t be all things to all people.
Instead of fighting over who’s site is bigger, who’s site attracts bigger numbers, how much time is spent on each site, Herald vs Stuff, TVNZ vs TV3, etc, etc, publishers should instead focus on providing good ‘clean’ attractive sites that offer great content, easy access to information, segmentation of content into relevant sociographic areas of affinity and apply a ‘less is more’ approach to advertising. In turn users will be more satisfied with their browsing experience and not be inundated with countless non-relevant ads that add to the clutter. Offer one or maybe two uncluttered ad executions (i.e. per hour) that provide maximum opportunity and value to advertisers to reach their targeted audience and charge a premium – after all every online user on every website across the entire www is a consumer of some product or service and is of value to some advertiser somewhere. My bet is advertisers will be willing to pay a premium if they can be certain that the audience is more likely to relate to the advert and their not having to compete with other advertisers all fighting for the same “eyeball.”
Give advertisers what they want, be transparent, stay focused, charge a premium – 1 ‘reach’ (user) at $0.10 versus 1 ‘frequency’ (impression) at $0.004
Everyone seems to think that the novel technology of the day will solve this business problem – instead we should first actually try to understand what exactly is the business problem we are trying to actually solve is and then use available technology to build novel tools to offer solutions – not the other way around.
Toni
April 21, 2010
@ 2 Cents Well said, wise words at last !
CS
April 21, 2010
@ 2 Cents: That's a lot of talk for someone who obviously has no clue.
"Ad interactions" are already tracked and is provided for by the likes of Eyeblaster. Usually, publishers have no view of this as the relationship is between Eyeblaster and the agency and/or client.
It's just so sad how some of you cannot differentiate between DIFFERENT measurement tools, DIFFERENT measurement metrics and DIFFERENT measurement purposes.
Above the line metrics like browsers, impressions, durations etc are presented to provide an overview of an asset's performance. Any decent publisher will provide their clients further information on specific content verticals, audience demographics, traffic drivers etc depending on the client.
The fact that you don't know this is already happening or are assuming that we don't have the technology for it only says 3 possible things:
- You're not in media
- You're in media but are too small that publishers don't talk to you
- You don't understand online media
- All of the above (oops sorry that's 4 possibilities)
Preston
April 21, 2010
We need to get this back on track. Who is Nigel Tutt??
2 Cents
April 21, 2010
@CS: I am very aware of the DIFFERENCE between 'above the line' and 'below the line metrics' and am fully aware of what metrics Eyeblaster, DoubleClick, and others provide.
Just so I have this clear… You're saying that publishers sell you one 'bill of goods' something like the total minutes of a TV program (i.e. $10 per program min.) and the client then receives a whole other perfomance metric post-campaign (i.e. if only 1 viewer saw the program – at $10/min thats the equivalent of $300 for that viewer) sounds like a DEAL – where do I sign up?! – NOT!
Publishers need to focus on doing what they do best – provide good content to attract AUDIENCES. Metrics (publisher or 3rd party iniciated) need to measure those AUDIENCES (and there are tools out there that have been doing this for years as its not so novel) and sold to agencies based on "who is viewing or likely to be vieweing" NOT how many pages are being 'flipped' or how long it takes them to read the content itself. And finally like most other media that I am aware of – there needs to be a measure of near accurate predictability – so an agency has a good idea of what audience they are trying to buy on any given publisher otherwise its kinda like fishing in the middle of the ocean with only a line and hook.
Toni
April 21, 2010
A marketing lesson in brand exposure (BXD) and why page duration is so important to advertsiers?
There are a number of marketing levers that impact brand awareness,
message association, purchase intent, and, ultimately, sales. The classic
levers of any advertising campaign have traditionally been reach—how
many people are exposed to ads—and frequency—how many times
those people are exposed. However, neither reach nor frequency speaks
to the quality of the ad exposures. The element of time is missing from
the basic reach and frequency equation.
To illustrate this, imagine two campaigns, both of which have a reach of
1,000,000 and an average frequency of 3.1. They look exactly the same
until we add the third element of time, measured as Brand Exposure Duration (BXD)—the length of
time an individual is exposed to a given advertisement. One campaign may have an average BXD of
18 seconds, while the other may have an average BXD of only two seconds, thus revealing a dramatic
difference in media performance.
Why is BXD Important?
Not all impressions are created equal. It doesn’t take a big leap of faith to believe that the longer a
person is exposed to a brand, the greater the branding impact is on that individual. The economics of
television pricing demonstrate this point. Longer spots cost more. 30-second spots cost about 20%-
40% more than 15-second TV ads.
Toni
April 21, 2010
Brand exposure duration is a relevant metric for any medium. When a person is exposed to ads in TV,
print, online, radio, or even outdoor, they are exposed to the ad for some duration of time. A
magazine ad placed next to the feature article is going to contribute more brand exposure than the
same ad buried in the classified section. Someone stuck in rush-hour traffic will have a much greater
opportunity to pay attention to a billboard than someone driving 80 miles an hour along the same
stretch of road. The fact is, BXD is an important element to advertising effectiveness. and a little more on BXD……….
The challenge
lies in the ability to measure and monetize it from one medium to another. In TV and radio, BXD is
how the media is bought and sold. For online BXD has hardly existed to date but as a medium matures…….!
2 Cents
April 21, 2010
@ Toni: Well said!
BDX has in fact existed since early 2003 (at least here in NZ) – I know of one particular company that has been promoting this for some time now and especially with duration of exposure to video and rich media ads. I suspect however that the 'penny has yet to drop' and publishers and agencies are still struggling with how this is suppose to work. BDX is not about click-troughs, conversion or even impressions – these are metrics that we all know have been used by the early adopters of online ads – that doesn't mean however that they are the best metrics for pre or post-campaign analysis. Campaign metrics needs to be measured against campaign expectations i.e. if its a brand exposure strategy then duration is key – if its an acquisition strategy then click-through/conversation should be the key metric – of course there is always going to be some aspects of both.
The one flaw in your comment Toni is that 'page-duration' does not necessarily correlate with duration to an online ad unless a publisher can guarantee that the ad is and will remain in view of the audience. This may work in TV or print in that all of the available content is in full view – in online however pages (as we all know) extend beyond the users screen so if an ad is 'out of view' then page duration or impression metrics have zero value as user never actually see the ad – BUT the agency is still going to be charged for the impression or duration (I'm sure Eyeblaster or the likes forget to mention this to their clients in their 'below the line' metrics)
There is no one specific answer or solution to this ongoing dilemma – but some fundamental basic ground principles should be applied to 'level the playing field'.
1. Content is not 'free' and users need to understand that advertising is the currency for consuming content online (Good luck to Murdock is he thinks subscription based models are going to work – NOT!)
2. Ads need to be interruptive if they are to have some effect – doesn't mean it has to be in your face i.e. take-over’s, interstitials, etc. but at least they should be in you peripheral line of sight even if the content extends beyond the visible part of your screen (maybe the answer lies in a new way of presenting content through better website design like one area that stays fixed to display ads and another area that scrolls to deliver content – at least then the concept of page duration goes hand in hand to exposure to ads being displayed)
3. Finally, the real power will come when online and offline campaigns can converge in support of each other so TV ads coupled with Print coupled with online are assessed collectively in support of the overall media strategy. For now it seems that online is still in a world of its own – maybe agencies should consider online as part of the overall campaign strategy i.e. design a cross platform strategy that uses TV, print, and online to promote a message but uses the strengths of each media platform to support the same message. Maybe the sale cycle should start from the 'creatives' rather than the 'media buyers'
Build a solution that gives easy access to content while being sensitive to the user experience by giving them a simple but powerful and measurable brand experience and I'll bet users will easily begin to accept that some interruption online is acceptable and advertisers will be willing to pay a premium if they can be assured of exposure, relative audience, predictability and performance.
Sorry about the long post.
MG
April 21, 2010
Stoppress – you're really making us look like the geeks of the adworld!
Anyway, in an attempt to deflect the focus off Nigel …. Somer – you're right in the very first comment. Here's some stats comparing nzherald, Stuff and Trade Me in 3-2-1 order.
ASD for nzherald 6.28 mins, Stuff 6.32 mins, Trade Me 17.16 mins (Nielsen MI March 2010). The point of this stat is that on Trade Me, our Skyscraper and Banner ads run across the site so Average Site Duration (ASD) is a good measure of exposure for these ad placements.
Average daily domestic UB's: nzherald 197,739 dom UBs, Stuff.co.nz 220,504 dom UBs, Trade Me 572,210 dom UBs (again from Nielsen MI March 2010). This is a useful measure too.
Monthly domestic UBs: nzherald 1,914,776 dom UBs, Stuff.co.nz 2,263,588, Trade Me 4,913,603. Ouch.
Page Impressions: nzherald 66,394,649, Stuff.co.nz 97,758,710, Trade Me 1,385,244,307 (Nielsen MI March 2010). I can see their point that not all page impressions are created equal, but cpm rates, of course, take this into account. For example, Trade Me Sky/Banner Combo generally runs at around $0.56 – $0.85 eCPM for the standard tenancy spot, depending on the impressions actually served that week.
Total Time, Total Sessions … a quick glance at the Nielsen stats and I can see why they're keen to pitch up a brand new measure. Stuff are smoking them. The numbers tell the story.
2 Cents
April 21, 2010
@MG : Wow! So your saying that Trade Me has more unique browsers per month than total population!! Would appear to me that NZH and Stuffs monthly UBs might be more in line with the 'real' world if for example 1/2 the NZ population are internet users. Your numbers appears to claim that every man, woman, child and a few hundred thousand sheep visit at least once thoughout the month.
Also, doesn't Trade Me offer an entirely different proposition that HZH or STUFF – are you guys providing news now??
I'm not trying to offend – just saying that the current metrics are flawed and publisher's need to stop trying to 'one up' each other – other than toughting who's 'bigger' I'm at a loss to understand what exactly this tells me when I'm trying to locate and target a specific audience to 'flog my wears'
Can you tell me of those 5 million unique browsers (users) what the audience is interested in? Let's just say I'm looking to sell baby diapers to expecting mothers… can I book my ads to narrow areas of Trade Me that are likely to be visited by expecting mothers or are you suggesting that I buy across the whole site and hope like hell a few might relate to my goods??
As for Nigel Tutt, I suspect he'll be somone on the 'inside' at Fairfax that no one has heard of… I believe they like to promote from within the organization… anyone's guess really.
MG
April 21, 2010
Yup, we can target categories, sub-cats, even cats.
UBs are browsers not people. Two computers = two UBs. Old argument. Just reporting the stats from the industry standard – completely across your very valid point.
Question Please?
April 21, 2010
@MG I think this article is more talking about Average Page Duration (APD), not Session Duration (ASD) – how do the numbers look for each of the sites for APD?
JB
April 21, 2010
@MG Its a real shame the Chairman of the IAB in NZ, hasn't got the statesmanship to put the industry first and not wade in with ANOTHER TradeMe sales pitch. Shame on you…..
MG
April 21, 2010
@anonymous Question Please? person, you're right. As I said in my comment, I reckon ASD is a better model for ROS advertising than the APD suggestion that Spencer is seeking to implement. The Trade Me ratecard takes into account its huge page volume with low page duration – normal for an auction site. That's exactly why APD doesn't work as 'the better industry measure' that Spencer's looking for.
MG
April 21, 2010
JB – got two jobs, fella. One pays for the other. Shame on me.
Question Please?
April 21, 2010
@MG I see your point, but in this instance I would have thought that due to the way both NZH & Stuff sell their site (not just ROS ads) that APD is a very valid measure. TM is a very different site so therefore probably doesn't have much relevance to this debate – I'm not actually sure why it's being talked about in this discussion? Also, I have had an answer to my original question… NZH 1.07, Stuff 0.54, TM 0.26 (Nielsen MI March 2010)
2 Cents
April 21, 2010
@MG: I understand the notion the UBs are actually unique browsers and that if I have Explorer and Firefox on my current machine plus a laptop, home PC, work PC, iPhone, Blackberry, etc I would be counted as 7 UBs OR 1 reach x 7 frequency to an ad exposure (assuming of course I don't keep deleting my cookies and keep getting recounted) – BUT WAIT – your selling on impressions so if I visit 10 pages on each UB that's 70 impressions and an advertiser is buying impression based ad metrics -doesn't this mean that the real cost to reach a unique 'user' is actually a heck of alot more and don't advertisers want to buy audience rather than impressions or frequency?
My point is that reporting Neilson numbers or Session Duration is also about as usless a metrics when buying online media – the holy grail is in the ability to measure, predict and buy based on Audience Reach and Frequency (or as close as one can get).
Somer
April 21, 2010
Isn't it weird that no one from the Herald has (in a transparent way) attempted to defend this announcement in the comments?
They've clearly rushed out an press release without doing the strategic thinking behind it and now quite rightly being rubbished for it.
Equally odd that the mysterious Mr Tutt doesn't appear to know enough about online advertising to respond to the NZH announcement (when asked by Stoppress) with some of the excellent points that have been made in this thread (re: the obvious limitations of the Page Duration metric and need for advertisers to take a more holistic view of site environments).
2 Cents
April 21, 2010
In fact if you couple R&F with campaign performance metrics like BDX and Clicks and Conversion Rate then you have a model that is truly powerful and in fact one can even go so far as predicting expected results based on past perfomance and benchmark analysis – In turn, publishers sell less ads so less page clutter – users are happy – Advertisers pay more but get more targeted audiences and predicable results – clients are happy.
Everybody wins! and isn't this what we all want?
2 Cents
April 21, 2010
@ Somer: I wouln't bag Spencer or anyone who aims to consider, or propose 'change'. This is exactly the kind of discussion the industry needs and I am surprised that the NZ IAB does not promote opportunities to have open forum discussions amongst all its members.
Instead, my experiences with some of the members that sit on the board are more about promoting themselves, their own businesses and 'look at me' rather than taking a 'let's work together' approach to better the industry as a whole.
Ask yourself what exactly has the NZ IAB done for you except a few social events and a quarterly/annual report that by the way is only complied by PWC and not necessrily audited. I wonder where exactly the numbers come from and who actually validates them?? My mother always said… "believe half of what you see and nothing of what you hear"
Talk is good and I commend guys like Spencer for thinking outside the box – we don't always get it right but that's ok too – maybe instead you should be asking the 'body' that is suppose to represent the 'industry' to consider stepping up – or maybe its just suppose to be a social club and another excuse to get together a few times a year for drinks.
Somer
April 21, 2010
@2 Cents – for sure. Spencer and the gang at APN have led the industry on a variety of issues over the last few years.
I just think, relating to this particular announcement – while the basis for encouraging change makes sense, the conclusions they reached are questionable.
The debate it has generated is fantastic and should be encouraged.
Toni
April 22, 2010
Great to see industry support from TVNZ on the NZ Herald’s position. Their response is at iab.org.nz
JB
April 22, 2010
At least we are getting some consensus on this issue. So are Fairfax the only news site to generate pi’s without generating ad impressions?? If that’s the case why aren’t the IAB getting more involved in trying to create a level playing field here.
Vincent Heeringa
April 22, 2010
@MG Hey, geeks of the adword, keep it flowing. This discussion is very instructive for everyone. Not every reader will comment so don't think that it's just guys talking. We're listening and learning too.
2 Cents
April 22, 2010
Maybe I've missed something here but apart from the obvious, how does reporting 'ad' impressions as opposed to 'page' impressions improve things for the industry?
This article makes reference to NZH claiming that they believe increased 'page duration' will in turn create more value for advertisers through better exposure (this sounds like a website design issue and no doubt a better presented site will create more page 'stickiness' – not a novel idea but certainly a valid one and I agree – tomorrows websites will be about clarity not clutter which has been a past trend).
Let's assume that this can easily be achieved so now we have viewers spending more time 'per-page' – great we now have problem 1 solved – a more captive audience.
The next issue should be how do we report that audience's affinity and predictability i.e. what is the viewer's likely characteristics and what or where is that viewer likely to do next i.e. read more, leave site, go to section 'x' etc. – this one has yet to be solved but problem 2 would be to "know your audience"
Of course problem 3 is to count your audience and by 'count' I mean 'targeted reach' as the primary metric not 'page frequency'
On the basis that 1, 2 & 3 are addressed a publisher now has a valuable proposition to sell to advertisers – relevant, sticky audience with some understanding as to how many and how frequent.
At this point I would assume it starts to become a no-brainer for advertisers… match audience 'quality' and affinity to relevant goods and services and your odds go up in the likelihood that your message will connect with your intended consumer.
Of course, execution is also key and provided that a given publishers is able to offer various formats then the agency has various ways to get creative to deliver their message. Clearly we have lots of formats so this one it pretty much solved.
Wow, we could be onto something here!
Isn't it now a simple matter of campaign performance? – provided of course that the publisher is able to deliver the audience/frequency promised. Here is where ad exposure, clicks, conversion, etc come into play. Its up to the agency to determine which sites/campaigns work and which ones didn't – this gets a bit tricky because a bad ad is a bad ad and agencies have to accept responsibility rather than blame a publisher or the 'format'.
I believe all other media works basically the same way… or am I from another planet.
Everything else is gravy – optimization tools, analysis tools, planning tools, etc are all in support of 'bettering' the basic performance of the media format. And yes – of course the internet is a more powerful medium because of its measurability, immediacy and interactivity – hence the reason that publishers should be charging a premium for your audiences.
I wonder why the online world still hasn’t got this one yet?? Maybe its because (as it is being demonstrated yet again) that the "cart seems to always keep coming before the horse"
Exciting times ahead and hoping the sleeping giant is finally awake- breathe deep… hold breath…
Matt
April 22, 2010
Haha wow this is really firing people up.
stoppress.co.nz most be stoked their APD is rising with all this talk!
2 Cents
April 22, 2010
Do see any ads through :-)
Wait, I think there was something displayed way up at the top of the page – or not. Page impression = 1 / Ad impression = 1 / Page duration = … / viewed ad = 0
CS
April 22, 2010
@2 Cents. None of what you say is new. Publishers can already do this with their "internal" site analytics.
Why do you think they know what to change in order to increase 'durations', in the first place?
The reason why you're all getting worked up over simple "reporting" is because you see what everyone else sees… basic, topline information.
No publisher will report ALL metrics and ALL possible data in the public domain. A lot of that data is sensitive.
But, if you're a smart advertiser and if you had a capable publisher, then they will already be providing you with key insights in the form of which you mentioned. You just need to ask.
CS
April 22, 2010
@2 Cents: Just read your initial response to me previously.
When you're buying TV, you get told what audience a programme is reaching and how many people watch it. The "inventory" is the 15/30 second ad spot.
Does TV tell you how many viewers per second of your ad? Do they tell you if pupils dilate in the 3rd or 30th second that indicates connection with your message?
No. And you wouldn't expect it to. It's a different form of media… passive consumption, no interaction needed by the viewer.
But since you feel the need to standardise digital/online with other, traditional media… I'll humour you.
When you buy online, you go after the site's which have the audiences you want (by demos, which is comprehensively covered by Nielsen, for eg). The "ad impression" is the measure of inventory you buy.
Online post-campaign analysis, is different from online audience metrics. PCA's tell you how the campaign performed based on your goals, be they click-throughs, conversions, interaction time etc. However, most of these metrics are measured from the point of interaction with the ad-display unit, ie the AD IMPRESSION (technically, these are all click-stream data – the root that makes online measurement possible – but who's being technical right? Not like we really know what we're talking about anyway).
So lets recap.
Buying TV, papers, radio or online… you look at the 'audience' or the closest measure of 'persons'.
Money is then exchanged for inventory; impressions, tv spots, columncent's etc…
Performance is then measured; How many people claimed the vouchers? How many walk-ins on sale day? How many text messages? How many call-ups? How many website visits? How many click-throughs? How many conversions?
How many people "saw the ad"… on TV… on paper or online?
Do publishers sell based on 1 metric and report success in another?
They sure as hell do. Because you're not getting 1 million odd people seeing your ABC ad and buying your ABC product in the middle ad segment of the evening news… that's for sure.
And if you're measuring your success based on that, then you're doing it wrong. Predictability my ***
Just browsing
April 22, 2010
Didn't the Herald just change their website to split articles up over a number of pages to increase the number of impressions?
Obviously that isn't working too well for them so they want to change the game (maybe they'll make the page longer to load to increase time on site).
If they put all their articles on one page (which is the normal convention), impression numbers would be down at least 50%.
Somer
April 22, 2010
What Stuff's got right that the Herald has got (and continues to get) badly wrong, is that Stuff focus on the user experience first.
That's why Stuff are smoking the Herald on audience, PI, ad performance (in my experience) and virtually every other metric you can think of.
Not doing a full page refresh when a user initiates an action such as clicking through a photo gallery or doing a quiz is simply better for the user — AND — it's better for the advertiser, since their ad stays in front of the user longer.
Now clearly that isn't captured/reflected by the Page Duration metric – but it doesn't change the reality, the Stuff approach works better for users (no waiting for full page refreshes) and works better for advertisers (longer ad exposure duration).
NZH is the worst offender when it comes to tactics that inflate page impressions to the detriment of the user experience (story pagination being the most obvious example) and to start whining like little girls now just because they find themselves in the unfamiliar territory of being WAY WAY BEHIND there nearest competitor is just sad.
Least the Stuff page impressions are being generated by tactics that are to the benefit of users and, by extension, advertisers.
If the Herald focused on their users from the beginning they wouldn't be in this position.
JB
April 22, 2010
@Somer You repeated the same thing about 3 times yes we get it. Move on and get some work done !
We appear to be going round in circles. Lets get together as an Industry and try to move more money into Interactive, thats surely what we all want. When we actually start pulling in the same direction and stop this petty squabbling we might actually get somewhere. Over it and bored now !
2 Cents
April 22, 2010
@CS: For the record, no one is getting 'worked up' – we're just exchanging views and maybe, just maybe, initiate some positive change or at the very least recognize the areas that need change.
1. Just because the internet is interactive it doesn't mean that it doesn't provide opportunities for passive engagement and therefore should not be measured strictly on the basis of clicks – the emerging brand engagement metrics is a testament to the fact that unlike TV you can measure how many seconds of your ad is viewed by x number of viewers. If agencies learn how to use this BDX metrics they in turn can gain better insight in support of TV campaigns. The various media platforms should work together in support of each other and not in isolation.
2. Just how exactly do you buy 'demos' online? Dad (male 35-45) views NZH at 5pm but then his 18 year old daughter jumps on at 6 and does some NZH surfing of her own – so when exactly is my targeted 'demo' ad being viewed and to which 'demo'?? The PC is recognized as an anonymous machine – that is of course if some new browser comes out which requires unique user to provide demo profile and then login every time you surf – not sure if the privacy police would like this one too much though…
3. Yes, the industry does need to attempt to standardize online with offline – that doesn't mean make it the same – what it does mean is make it so you can make some fundamental and sound decisions that demonstrate how online works to support offline. i.e. see TV ad during 6pm news, hear supporting campaign on radio while driving to work, notice campaign again when flipping through morning paper AND see/interact with the online execution while surfing the net. Sounds to me like a good strategy…
4. Yes again, I agree that post campaign analysis is between agency and client – no one is stating that sensitive information needs to be shared with the masses. I didn't realize that disclosing your website audience numbers was suppose to be top secret – and for publishers who elect to provide more data in support of that I suspect will gain more credibility when competing for ad spends.
5. I don't believe that this discussion is about arguing the 'technical merits' of the various media options or the solutions that support them – I have been in the online media game for a very long time and have watched the NZ market grow from its infantile stage so I am very aware on what's out there, who doing what, what works and what doesn't. Cudos to all who get involved – no one is 'more' better – we're just 'different' aiming for the same carrot.
6. Money is exchanged for the expected audience (in TV), subscription (print), listening audience (radio) – not the 'spot' itself but what the 'spot' is likely to achieve in reaching an audience at a given time or on a given program – some spots are more valuable than others because… obvious answer which I trust you know the answer
In online we seem to work backwards – buy the frequency and then post-campaign we'll tell you (if at all) the audience – and even that is flawed because campaigns run across multiple sites and no one knows if its the same viewer across both sites adding to even more frequency or two different viewers
7. Your last comment… so the issue raised in this article is about better transparency but you're saying that publishers sell you one thing and report on another – seems to be like something doesn't add up here… If I buy a print ad in a magazine that claims to have 1,000,000 subscriptions I expect to get 1,000,000 unique subscribers (at a minimum) – I am not naive to think I'm going to get 1,000,000 buyers – and I'm also not going to pay for 10,000,000 pages or page turns.
How then do you suggest success should be measured – or should I just trust all those transparent reporting metrics?
2 Cents
April 22, 2010
@JB: here, here! well said. Who really cares who has the bigger d***
adpro
April 22, 2010
Great discussion here folks. Thanks for sharing.
Matt Harman
April 22, 2010
Having read through all the comments, i'm wondering whether this is less about a metric issue, than it is about outdated terminology.
ie: pre dynamic page updates we all happily accepted Page Impressions (divided by Visits) as a proxy for site engagement (along with Session Duration).
Page Impressions were never commonly treated as a proxy for Ad Impressions, since there's a raft of reasons why the two metrics shouldn't be expected to be the same (as 'CS' explained brilliantly with the newspaper analogy above).
The change Neilsen made to its methodology makes a lot of sense, because it allows publishers to take advantage of dynamic page update technology (for embedded Photo Galleries etc) without 'costing' the site Page Impressions (and perceived engagement).
It incentivises sites to use technology in smart ways to create the best possible user experience – which benefits everyone.
Thing is, while the new methodology makes sense, it means that 'Page Impressions' aren't really 'Page Impressions' any more – at least in the way we used to think about them.
They have morphed into 'User Interactions' (a quiz question completed, a request to see the next picture in the gallery, or whatever). Perhaps not exactly Page Impressions anymore – but probably actually a more useful measure of site engagement than any metric we've had before.
Certainly much more useful than Page Duration, which as discussed by various people earlier in the thread, can be influenced by factors that are either positive or negative, making the metric an inadequate proxy for engagement and certainly one that is inferior to Session Duration or Page Impressions divided by Visits.
Toni
April 22, 2010
I think we’re about done here. Good luck everyone, my hope is by the end of this year, we will all be singing from the same hyme sheet. That’s why IAB UK has been so successfull 25% of the media pie ! IE publishers getting behind a common goal and really going for it. Debate is healthy slagging off the competition is unprofessional ie so I’ve always thought !
Yup!
April 22, 2010
Well put Toni. Preaching to the choir is how it sounds to me and, in this case, everyone wants to be the Soprano! and yes the bullets will start firing at me (predictably) but…damn…someone had to say it.
Chris
April 23, 2010
Well said Toni. Overall great debate and good for the industry but when Publishers slag each other off they imediately lose credibility and I have to say NZH didn’t do that in their statement. Let’s start showing some maturity and move this fantastic industry forward.
Brent
April 23, 2010
If NZ media providing online advertising were REALLY confident about the effectiveness of their offering (as opposed focusing just on selling space) they would be prepared to sell on click-thru.
Now what I would REALLY like to know is what the CTR of nzherald.co.nz (and other) sites are, on average.
NZH and others may well have that info – but I don't see anyone sharing it.
I wonder why not…
CS
April 23, 2010
@2Cents:
As far as I my experience goes with online media buying/selling the process is always:
1. Identify your client's target demo
2. Profile that demo to the site. If you don't have access to the tools, ask the publisher (Eg. Can you give me sections or content verticals which have X% Males, aged 20+ who live in Auckland and are interested in sports)
3. Publisher supplies the relevant information
4. You buy the ads, relative to the publisher's inventory (ad impressions) to ensure you maintain a good level of visibility.
5. The campaign runs. Post-campaign you'll find out how well it worked, look at your ROIs and performance against other sites etc.
So I don't really get where this "publishers aren't transparent about their audience", "we're only finding our audience after the campaign" etc stuff is coming from?
Duplication of audience is also easily solved. Run some cross-duplication reports between your prospective sites and find out how much of a cross over there is.
Principally, it's the same as buying other media. You know the audience for a programme, or time slot, or publication. You buy with the expectation to hit that number of audience, yes, but the inventory measurement is different from the audience measurement.
On TV you're still buying by time. On print you're buying by space (not page turns). And online you're buying by impressions served.
It's pretty straightforward and it's not this huge conspiracy which you seem to think it is.
Yes, improvements can definitely be made, but 'switching' from 'impressions to duration' isn't game-changing at all. It is still the same methodology of measurement, using 1 of the metrics in question (total duration/total impressions) and it's still coming out of the same horses mouth…
2 Cents
April 23, 2010
@Brent: You are mistaken in your assessment and expectations of 'online media'. Your questions illustrate exactly what is wrong with online advertising and how it has come to be perceived.
What is unique about online is that it has characteristics of all media combined (can be passive, can be interactive, can be textual, visual, interactive, etc, etc) so trying to measure its success using only one aspect is all wrong. Using ‘CTRs’ as a means of 'success' is a metric used to gauge campaign performance – BUT only to the extent that the campaign’s strategy was acquisition driven. If the campaign strategy was brand building/awareness then measuring brand-engagement (BDX) should be the key metric (i.e. how many/how long as opposed to who clicked/who bought).
Better creative’s, better strategies and better execution of online digital campaigns will certainly go a long way to improve performance metric (acquisition or otherwise) – but that is another discussion in itself.
Where you are mistaken is that you suggest that a publisher needs to be accountable for click-performance (or any other performance for that matter) ?! What I’m saying here is that a publisher is a ‘media owner’ (i.e. like a TV station) – it’s web site is kind of like a program (i.e. 6pm news) and like all media there is an audience who will watch that program. TVNZ will NOT be responsible for how many people actually view your TV ad (only the program itself) or how many people call the 1-888 number to ‘order now’ – so why do you feel web publishers should. A publisher (like print, tv, radio) is selling you an opportunity to ‘pitch your wears’ to their viewing audience – their only obligation is to provide you with an sound audience measurement and the opportunity to show/deliver your ad. The success and performance of the campaign itself is between client/agency/media buyer i.e. bad ad/wrong placement, versus good ad/right placement, etc has nothing to do with NZH or any other publisher.
So yes, NZH, Stuff, or any other site should be doing just that – selling you the space and opportunity to promote your offering to their viewer audience. The issue being discussed here is what standard metric should be adopted to assist advertisers when trying to buying online media and I for one think that neither ‘page impressions' or 'page duration' is the answer.
2 Cents
April 23, 2010
@CS: Thanks for illustrating exactly my point!! Your 5 steps to online buying and post commentary is deeply FLAWED on so many levels and just because this has been the standard practice since the mid-80's it doesn't mean it is correct or that it works.
Also you might want to know that you DON'T buy time on TV you buy RATINGS (audience) and combine this with a time spot (15s/30s/60s) amongst the program stream – you DON'T buy SPACE in print you buy SUBSCRIPTION or CIRCULATION (audience) and combine this with 1/4page/1/2page/full-page spot amongst the content print. In online we’re still missing the first part of the process – as you noted… kind of like target practice with a machine gun! Hey with enough fire power you’re bound to hit something!!
My objective in raising my point of view in this, or other discussions, has always been to the betterment of the online ad industry in NZ and abroad – there is no conspiracy mate.
Michael
April 24, 2010
Interesting debate.. and I don't understand much of it at all. I'm a creative so I don't get involved in counting eyeballs very much – but I am involved in trying to convert those eyeballs by doing my best with the fluffy bit.
Sometimes our job is to get people to like us.. so we try to make a engaging ads that people might notice (and that's the big that should be counted) And sometimes the goal is to get people to click through and either do something or buy something (at that point click-thru is key).
Now, like I say, I know nothing about this stuff… but surely media people would choose the most relevant places to put stuff, based on the type of job it is.
i.e. (to attempt media language and probably fail) wouldn't we place ads with publishes who have a hot BDX measurement when we're looking for brand exposure.. and buy space with with people who have impressive click-thru rates when we're looking for click-thru?
It may be they're the best websites for both are the same ones… but it doesn't feel like they should be.
Question.. just out of interest, cos there's lots of smart media people on this post – how do you count the effectiveness or otherwise of branded content? Is it as simple as "youtube hits"… and how does that work in other channels? Or did the very fact that I asked the question show how little I know about what media people do?
Michael
April 24, 2010
Sorry, several typos in that. I think the maths and measurements debate fried my brain a bit. Heading off to be an "audience" now. :)
2 Cents
April 25, 2010
@Michael: Look at your question this way… say you ran 2 campaigns online – say they were both online video or rich-media ads (say c1 is a 15s ad and c2 is a 10s ad) – say you used impression based buying model and bought 1,000 impressions for each ad. Using the example of 'shoot first ask questions later' – you buy your impressions and you get your unique browser (or even better 'user') results after the campaign has finished and assess that you reached 250 unique browsers who would have seen your ad 4 times each (250 uniques x 4 impressions = 1000 total impressions per campaign).
You should get 2 calculations – campaign 1 played 1000 times for a total 10,000 seconds of the available 15,000 (total brand consumption) versus campaign 2 which totalled say only 7,500 seconds of the total 10,000. Clearly campaign 1 had a higher brand 'consumption' value – users took in more of the c1 brand message – even if they didn't 'click'
The other measurement is average 'engagement' – if both campaigns delivered to 250 Ubs then c1 had an average engagement per user of 40sec of the max 60sec and c2 had an average of 30sec of the max 40sec. This is probably too simple an example but I assume you'll get the idea – campaign 2 was more 'engaging' because the user stayed with the ad longer – 75% of the ad in c1 as opposed to 66% of c2
Take this to the next level and if you monitor average engagement during the campaign and start to notice a drop then maybe users are either getting it and the frequency is to high – now you have some sound data to make adjustments like swap to a second ad to re-engage users or slow down the frequency buy spreading out the delivery times (don't burn them out so fast) – I think they call this optimization :-)
The catch here is that the ad itself needs be in 'full view' otherwise the ad is off page and the numbers are not accurate. In-stream and over-page ads are more measurable than in-page video or banner ads.
I think the concept of 'page duration' as presented in this article is attempting to provide you with an aspect of this i.e. higher duration means more consumption means more engagement means more value — BUT page duration (and how its measured) is not likely the right solution as there are too many flaws with the current methodology.
Michael
April 26, 2010
Umm… I think I'm with you. Lots of maths in Media, eh?
Here's another stupid question. It might well be covered above. Does page duration measure my time with the ad.. or just my time on the page. Eg. I've spent a bunch of time on this page reading and replying to people's opinions… but I couldn't tell you anything about any of the ads?
Do I secretly read them by magic in the same way that my 2 year old knows what Coke is even though we've never had it in the house?
Michael
April 26, 2010
So I've been thinking. And it might be that this conversation started and ended ages ago. But here's what I thinked:
Coming from the creative department, I only really get simple stuff. So in the spirit of simple (on the understanding that I have very little understanding, and I’m opinionating wildly) here’s what I reckon.
If we go back to why we’re counting stuff, it’s all about getting the right message to the right person for the best value. And with that in mind, surely it’s less about accuracy and more about relativity.
At the end of the day we’re judging cost vs numbers. So if the numbers are consistent across all publishers, it’s much easier to make a call on the best place to be.
Take a look at radio. Everyone knows that the way they count “earballs” is deeply flawed. But it’s a consistent way to count. So you always know (within reason) that one station is better value for money than that another. Sort of.
The challenge (as someone said above) is when a publisher tickles the publishing to make the numbers better (like the way radio stations only really self-promote during survey season). But once again, if we assume everyone does it, then everything’s relative and the answer is relatively useful (pun intended).
I still think there are different types of ads (eg. direct response and brand) so it’s handy to have different metrics based on desired outcome (eg. CTR vs BDX). But in the spirit of the debate I tend to side with the cynics who feel that the Herald’s recount of numbers sounds like it’s more about them losing market share than us getting better value.
As a non-advertising example of the same thing, my Dad is a scientist who counts greenhouse gas emissions. It’s all pretty random and difficult to count. So he has a look at specific indicators and then does some educated guesswork and a bit of maths.. but it’s never going to be 100% accurate. However, if the goal is to show year on year improvement (or otherwise) accuracy is less important that relativity. As long as you count the stuff in the same way every time, you get a pretty good idea where you stand.
So yeah, that’s what I reckon. But I might have completely missed the point.
Oh… and while we’re here. Did anyone nail a great media sales argument for creating branded content? All we’ve really got in creative is “ain’t it cool, you’re customers will love it”… and the usual response is “gee that’s expensive and the logo is too small”.
Any thoughts?
Open Mind
April 26, 2010
One of my 'traditional media' know-it-all mates sent me this link. Not a bad place to source open ended articles and debate on media in general.
http://www.digitaltonto.com/2010/5-problems-digital-media-needs-to-solve/
'Joe Public'
April 26, 2010
"Users viewed 33.2 billion videos during the month of December 2009, according to comScore. "
Online video:
http://searchenginewatch.com/3640021
Users both engage and follow video content!
Reading text and looking at photo stills of second hand items on websites is NOT a user experience or engaging for advertisers brands..
Whatever cpm, apd, yawwwnn( is there a video on this?) etc etc etc metric you throw at it!
Is it also possible we are not speaking the language of 'Joe Public', so they can understand this scary online space, nobody buys what they don't understand!
Josh Borthwick
April 27, 2010
2 Cents you're on a roll here. Before you get too carried away – please make sure you stop and check your facts. PWC do a fantastic job of collating ad spend and auditing both the numbers and the process on behalf of IAB NZ. I've personally been involved in two different, routine (just to quash any obvious banter about being dodgy), audits in two different companies and I’m sure there are plenty of publishers who can attest to the robustness of this process. It’s easy to have a crack at the efforts of others from behind an alias. Having said that, I think you’ve raised some great points in this forum.
2 Cents
April 27, 2010
@Josh: Apologies if my comment re PWC was mistakenly understood. I am aware that PWC compiles industry reported numbers from the various publisher participants and 'compiles' those numbers into their quarterly/annual report. PWC 'stamps' the report on the assumption that the numbers are sound. To my knowledge (and I'd even be willing to bet my left one) the numbers themselves are not audited as this would clearly be time consuming and very costly.
I wasn't trying to undermine the report or that there is anything 'dodgy' going on – I was merely stating that not everything you read on the internet is true (or entirely accurate)
Look at the disclaimer in the report itself… "PwC does not audit the information and provides no opinion or other form of assurance with respect to the information" – I'm guessing no one really looks at this huh?
Almost every major player i.e. Microsoft, Fairfax, INM, (except Google) have consistently posted losses in online advertising for as long as I can remember – so no one is making any money. Revenues are growing, yes, but if the big boys can't figure a way to make a buck then we're still a long way from establishing a credible new advertising medium despite all the cool techno mumbo-jumbo. So maybe the problem isn't how we're 'counting it' – maybe its how we're 'selling it'
Don't get me wrong, I'm pro IAB, pro industry, pro online and will certainly stand up to promote or endorse our industry sector. I do feel however that everyone needs to take a good hard look at where we're headed and aim to all benefit from the growth in years to come. 3 or 4 big boys (NZH, Stuff, TradeMe, MSN) doesn't make a market or an industry and clearly they don't have the answers otherwise i suspect they all be reporting online ad 'profitability' – everyone has to be able to win in some way, shape or form and yes the little guys matter just as much if not more.
Consider this… you jump online and your NZH/Stuff home page starts up – you read the headlines, browse a bit to catch up on current events and then enter your 'google' search phrase – you get 1M search results and you start drilling down to what you really wanted to find in the first place. Well its no surprise that its on those sites (the smaller guys albeit millions of them worldwide) where the real value for advertisers exist as these sites are designed to target specific interest groups and are more easily matched to brand affinity. The unfortunate thing is that the little guys never gets a say, or ever gets heard, yet they offer the most 'advertising' value.
When you have guys like the IAB Chair boasting how great their numbers are and how they're 'smoking' the competition in UBs, Impressions, etc, etc, then you have to wonder if the industry players (big and small / publishers and agencies) are being fairly represented and if it will ever have a platform to debate issues, grow business and evolve the industry. There is a lot that the NZ IAB can do in helping to solve the various issues we all face including a standardized ‘metrics’ methodology – self promotion or ‘chest beating’ however is probably not what I would classify as a good choice.
Paying Client
April 28, 2010
@Josh: I have an open question I was hoping you can answer?
I had a look at your posted rate card on AdHub and something doesn't ad up.
Example; Network Buy/ Big Banner / CPM 15 / Weekly Impressions 5.1 million / Weekly Uniques 390,000
Assuming I buy 1 week worth…
- 5.1m impressions / 360,000 uniques equals a frequency of 13 impressions for every 1 unique ?!
- at 15$ CPM the cost to me would be $76,680 for 5.1m impression BUT this also equates to an astounding $196 per thousand Uniques!!?? That's nuts!
How come this is SO MUCH higher than TV, Radio or Print?
How do I convice my clients to shift more budgets to online when the actual cost per audience reach is HUGE!? and frequency is way over the expected norms?
Have I missed something or is this how ad networks attempt to pull the wool over our eyes so I don't really know what I'm buying? Why would I want to buy an audience for almost $200 per unique??
Also, how is measuring page duration going to improve things?
Sorry about all the questions but I don't see any posts here that gives us, the paying "clients" some real answers.
Josh Borthwick
April 29, 2010
Firstly to 2 cents – I’d be interested in your opinion on how other media audit their annual advertising revenue numbers. My understanding is they’re certainly no more robust than the IAB NZ figures. It’s a bit hard to gauge your experience or contribution to the figures, IAB NZ fees or the industry, given you’re not disclosing who you are.
The board has a mix of larger publishers, who all pay considerable fees to fund things like the audited spend figures, and smaller publishers like, Matt from eBoss, Craig from Metservice, David from NZRU and myself – representing more than a dozen small to medium publishers via adhub. Nominations and entries are open and they’ll be coming soon in August, so I look forward to seeing you represented…whomever you are. Perhaps you could come in a mask like Zoro :D
Paying client: I’m a little embarrassed to say those UB figures are wrong. They’re currently sitting at just over 600,000 according to Nielsen – if you look at domestic traffic for the adhub network across a week. Having said that, it certainly isn’t adhub’s expectation that you would use 100% share of voice in one ad unit to manage frequency. I would recommend frequency capping your campaign to get the desired result – i.e. 390,000 uniques (using the old figures) x3 gives you 1.2-million impressions for $17,550. Like other media, we’d also look at the volume you’re buying, how many campaigns you intend to run over the year and negotiate the rate card on that basis. How many media buyers / marketers buy TV at rate-card I wonder? How many can frequency cap their newspaper advertisements?
So taking it at face value, your campaign of 1..2million impressions over one week would reach 390,000 unique browsers at a cost of $45 per thousand uniques. Measuring page duration or better yet session duration, gives you heightened comfort that the 390,000 users have actually seen your ad or had a significant opportunity to engage with it. A user bouncing through one page in under a minute has far less opportunity to see your ad than someone chewing through multiple pages over 5minutes.
I like the debate, although I think it would be far more constructive face to face or in a real working group. If you are indeed a paying client or an agent for paying clients, then please get in touch directly and tell me what you want. I’ll do my best to accommodate it.
iChild
April 29, 2010
Really interesting to see all the comments on this. I have been reading with great interest.
@Josh – I think you're right that this would be more constructive if it was put to an actual working group. It would be nice to get publishers and agencies in a room together and actually discuss what agencies are looking for for their clients, how we can work with the publishers to deliver what they need and whether there is any point trying to compare online to the 'traditional' mediums.
I wonder what StopPress' page duration on this page is?
2 Cents
April 29, 2010
@ Josh: My earlier post was not intended to discredit you, the NZ IAB or any of its members. Everyone is an equal and valuable contributor in this arena. Instead I am simply sheding some light on a few areas that I for one believe need to be considered by all when debating the issues as presented in this article.
As for reporting numbers – yup you are correct in that its not a perfect science and all media have their short-comings. Yup, I know TV is panel based, so is Radio, Print is circulation based etc etc. The problem with online however is that the folks out there know it IS measurable and as a result expect more so the 'standard for accounatbility' and 'transparency' for online is much higher. Its both a pro and a con of the medium and I suspect there is still alot more to come before we reach an online metric standard that satisfies the needs of both buyer and seller.
Your response to paying client raises an interesting perspective… my question is why don't you simply sell on a unique browser basis i.e. 15$ per thousand UBs and let the buyer control frequency so if they wanted 3x exposure they would buy 3 x 360,000 UB campaigns? Wouldn't this make more sence? I suspect then that post-campaign analysis would tell the client if and how long the ads were viewed (duration).
Sounds to me like this could make alot of sence and create alot more transparency – or am I missing something.
Josh Borthwick
April 29, 2010
Hey 2 Cents, I don't think UB's are a bad metric at all. That gives publishers and agencies much more healthy debate around quality of audience too. I'm sure there is a valid metric beyond impressions out there and there are plenty of smart people in this industry to work it out.
I'll have a chat to the other board members next week and see what appetite there is for a working group on new inventory metrics. I mean what I say about valuing your contribution too, so please get in touch direct if you'd like to get involved.