Questions raised over news sites’ local stats
Australia’s major online news providers have suffered a new blow to their credibility with the publication of analysis that suggests that their local mastheads are receiving a large volume of out-of-state- traffic.
This is an issue because advertisers who choose to go with local mastheads are often paying to target a local audience.
The figures have emerged at the time while controversy is still raging over the use of autorefresh to allegedly massage traffic numbers.
Shepherd gathered the data by examining the little known state-based-reporting function on the Audit Bureau of Australia figures.
It revealed the proportion of traffic that was identified of coming from within each state:
-
Brisbane Times 32%
-
WA Today 34%
-
The Courier Mail 42%
-
Adelaide Now 43%
-
Perth 47%
-
Herald Sun 54%
-
The Age 54%
-
The Daily Telegraph 58%
-
West Australian 65%
-
SMH 65%
In his posting, Shepherd asked: “Are the networks pushing people from site to site to try and build topline unique browser numbers?”
The issue arose last year when Mumbrella reported that Sydney based readers of the smh.com.au were being pushed onto sister title theage.com.au.
Tim Burrowes founded Mumbrella in 2008. He worked on Mumbrella until 2021. At various points he was editor, content director and editor-at-large.
Linkedin
This is getting ridiculous. Name one advertiser on the SMH.com.au website solely dependent on a market defined as metro Sydney?
User ID not verified.
Hi Terry T,
The SMH is probably a bad example for you to choose, not least because it has the biggest local audience as a proportion, at 65%
Cheers,
Tim – Mumbrella
“Are the networks pushing people from site to site to try and build topline unique browser numbers?”
I believe so. I’ve lost count of the number of times I’ve clicked a story on SMH and found myself reading it on WA Today or Brisbane Times, even when it’s not a WA or Qld-based story. The only reason I can see for that would be to use the SMH’s brand muscle to bolster the numbers for its smaller siblings.
User ID not verified.
Sorry Tim, are we talking degrees here or the principal of not reporting interstate traffic? Given that the SMH is listed here as part of this ‘analysis’, my question stands.
User ID not verified.
I’m pretty sure all RSS feeds for Fairfax websites now direct you to the WA Today version of a story when you click through to read it online. No matter which site you subscribed to the feed on. Definitely the case across most of the blog feeds.
It’s coded very specifically, too. A link that appears to deliver a blog feed from blogs.smh.com.au is hard coded as blogs.watoday.com.au. I can only imagine ‘tweaks’ like that help prop up less populated sections of the Fairfax Network.
User ID not verified.
I understand the issue of audiences arriving from ‘out of town’ but this information has always been available from the publishers themselves, all you had to do was ask. Often enough their figures would be very close to what Nielsen is now reporting.
That said, I completely agree with Ben when he asks if they are tunneling traffic to help build overall audience figures.
It shouldn’t be new, but if you’re buying a state targeted campaign, use IP targeting and if you’re buying Australia only, ask for it.
User ID not verified.
Hi TerryT …
And it’s not about metro Sydney … it’s about NSW. I’m not saying the SMH should only be Sydney … where do I say that? I included NSW and VIC to show the difference between these 2 states (where all titles have over 50% domestic traffic from the state they’re based in) and the other 3.
You ask a question around an advertiser looking to use the SMH for NSW only … erm, the NSW govt and its various departments? Bing Lee would be another I imagine. I wonder if their buys on the SMH and Tele are NSW IP only.
I wonder if the WA Govt are aware that if they’re not IP targeting their ads on WA Today and Perth Now are reaching more people who don’t live in WA than do?
My question is – why are so many NSW and VIC based people visiting sites in QLD, SA and WA? If someone can answer that it’d be great because I am stumped!! It can’t just be search and expats.
Just a small correction – the figures aren’t ABA – none of the sites have taken part in the audit. They’re taken from the geolocation tab in Nielsen Market Intelligence.
Umm, do you realise that the IP ranges on which the geo-location data depends changes on a regular basis? There is no consistent list of IP ranges. Over time IPs drift, causing fundamental issues for services like Nielsen’s and any IP targeting systems.
User ID not verified.
yes, but isn’t that a side issue to your original point. (ie what advertiser is limited to one market)
my question still stands – why are so many vic and nsw users visiting website in QLD, SA and VIC? and why do some sites have much larger %’s than others.
sorry, correction – :visiting websites in QLD, SA and WA.”
Was Google perhaps punishing duplicated content in search results, leaving the sites to link to the original article rather than reposting it under localised sections? Or perhaps we’re not as tribal as we used to be and we’re sharing/clicking links outside of our territories?
I’d be interested to know what the entry pages for traffic on these sites are now, are readers still heading to their bookmarked front page and browsing or are we more likely to be clicking on links forwarded by others or coming up in news/keyword searches?
User ID not verified.
mandi i agree it’d be interesting to know what the entry pages are and what % search is pushing to these sites.
my feeling is if search is pushing a large % over the brand of the search engine for news has become stronger than the traditional news brands. what impact does this have on the audience value to advertisers if such a large % are coming in via a search engine …
Ben, you’re using Nielsen data which itself is reliant on out-of-date IP ranges. ISPs change IP ranges all the time and can change IP ranges from one state to the next. Get it? Your analysis is using inaccurate data. The % figures indicating local versus interstate traffic is wrong.
User ID not verified.
You’re presuming that the ads on each site are blanket and not targeted to begin with. I know with the likes of The West they’re not targeted, but the Fairfax and News site mostly are. Example: if I’m in the United States, I’ll get primarily US ads based on the IP address when visiting local newspaper sites. I’ll usually get Melbourne ads on the Sydney Morning Herald, least on the display ads (and where there’s local ads in their system.)
To answer Ben and the overall point about non-state based visitors: you need to consider that content is universal and a good story on The West or Courier Mail could be linked to from across the world. Every now and then I see a link to The Advertiser on Fark or Reddit for example.
If the papers are doing very basic IP targeting of ads (and I’d argue most do) the statement “This is an issue because advertisers who choose to go with local mastheads are often paying to target a local audience” isn’t really an issue because the company will be offering a local audience for those ads. IP targeting isn’t hard, in fact it’s a basic feature on all ad platforms.
User ID not verified.
TerryT I am using the exact same data the publishers use to demonstrate their audience.
I’m unsure what your issue here is – first off you had a go at me saying no advertiser would do a one territory campaign … now you’re flaming me for using the industry standard data which is a completely different thing.
If I’m reading correctly you’re saying any geo attribution system or even basic IP targeting is way off which I can’t say I agree with you on.
Of course sites receive traffic from other states – and countries too. You cant lock out people just because they are in another state…
I frequent the BBC and Guardian websites daily from Australia but see Aussie adverts. I was thinking about this the other day. I am planning a trip to Europe mid year and when I look at UK sites I get Aussie ad’s served up. (I have the strong Aussie dollar and it would be beneficial for UK companies to be advvertising to me, because i am heading over there…)
Geo targeting can narrow ad’s down to very localised area’s – I suppose it can be one or the other:
You get ad’s served up based on where you are or ad’s served up based on where the product is. perhaps the user could choose what ad’s they wish to see…?
“View ad’s where you are” “View ad’s where we are”…
It will only get better and better for the user and the advertiser.
User ID not verified.
I understand that advertisers would want to focus on their local niche market – but in the case of theage.com.au for example, a luxury car manufacturer would not mind cashed up Sydney-siders viewing the advertisement as well.
User ID not verified.
to clarify – i am not saying you lock people out if they’re not from the state … nor am I saying that these numbers are a surprise … I am more questioning the volume of users from VIC and NSW on specific properties and not on others.
You could accept the reasons given (sharing sites, SEO, news aggregators, reddit/digg etc) if the numbers were reasonably similar across titles.
They’re not … WA and QLD in particular have loads of users from NSW and VIC. SA not so much. Both WA and QLD have a well established local ad industry with loads of agencies servicing local only clients. SA not so much (yes still to an extent but not the volume of WA and QLD). I’d say these WA and QLD titles have local only sales reps selling to local only titles.
The issue isn’t at all with traffic coming from interstate … it’s about how this traffic is packaged up and presented to advertisers. That’s it.
Where is the suprise? ANy media planner with a calulator should have known this. The dynamics a pretty simlple and hardly new.
If specific local market coverage is an objective then try using geo-targeted adserving.
A more interesting issue would have been publishers farming impressions to affiliate mastheads to bolster site numbers.
User ID not verified.
that is an interesting issue stephen – would you want to write something on it?