What does Fairfax’s Media’s data dump actually mean? And what’s going on at ACP Magazines?
Although I rather like stats, there are a few days a year where they become a little overwhelming.
Radio ratings releases offer eight such days annually. Over the space of a couple of hours, the data drops for the five main metro markets. Generally the phone starts ringing within 10 minutes, from station bosses aiming to give their interpretation of those numbers. It becomes a game of keeping them on the line long enough to sift through the data to try to discover the real story you need to ask them about. Within minutes a blizzard of press releases follow too.
In truth, the press releases mostly get ignored in the race to write the story. Then they’re mostly ignored because the story is already written.
And twice a year, a similar exercise surrounds the release of the monthly magazine sales figures, albeit under embargo a day or so before. Four times if you include the weekly mags and newspapers too.
Friday was such a day.
And in the blizzard of topline headlines, I think most people missed chewing over what it all means.
So 48 hours on, I want to draw breath and do so. Because there’s a lot to think about.
Not least because Fairfax chose that day to release a heap of new data from outside of the audit process on how it is travelling digitally.
Clearly it was designed to counter negative headlines around its poor print numbers. From that point of view, it wasn’t much of a success – not least because the print numbers are released under embargo, so the stories were already written and published.
But more importantly, it would have been almost inconceivable much more than a year ago that Fairfax Media would be the company setting the agenda on data transparency. I suspect that there will soon be pressure from the market for News Limited to respond. And then for the Audit Bureau of Circulations to provide a wider digital data verification service, although that is already in train.
The digital numbers are important, but let’s start with the print side of things.
Men’s magazines
The biggest shock for me was just how badly FHM – and indeed ACP Magazines generally – fared. I’ve been writing about media on three different continents for a decade now, and I’ve never seen a major media title lose half of its sales in a single year before.
One possibility I’d anticipated was that the monthly FHM had been giving away cheap copies and had perhaps stopped. But a look at the ABC certificate for the last year negates that. There never were any of those declared. Instead what we see is an accelerating decline.
A year ago the title had just over 50,000 sales. It lose 10,000 of them in the first six months of last year, dropping to 40,000. Then the decline accelerated, dropping a further 14,000 in the last six months.
There are, I suspect, a few factors at work. First, ACP’s debt-laden parent company – now known as Nine Entertainment Co, previously PBL – was being prepared for a float. I wonder if marketing costs were being cut from the budget to try to make the numbers look better. The float didn’t happen, and of course, ACP’s numbers are now worse than ever.
So today’s news of the talks to sell to PacMags make sense even if it would create a mega mag group.
Another factor is simply that of the threat of the internet to the men’s market. When a big element of the product is semi-naked bodies, it’s hard to see a bright future for the sector. Which would explain fellow ACP title Zoo Weekly’s big fall too. And indeed the drop for ACP’s soft porn titles Picture and People.
At least Zoo Weekly’s decline – about 15,000 in the first half of last year and 8,000 in the second half – appears to be slowing down.
Another bellwether would be NewsLifeMedia’s GQ, by the way – but it isn’t audited. Based on its last available Roy Morgan readership numbers (96,000 for GQ vs FHM’s somewhat doubtful 144,000), I wouldn’t be surprised if its sales are even worse than FHM.
It feels to me like all the publishers may simply be squeezing the last drips of profit from a sector they view as beyond saving.
Then too is the cyclical nature of magazines – even in normal times, some titles decline and die as new ones are launched. It tends to be a swifter cycle than other media channels.
That may explain why the younger (and in my view excellent) Men’s Health, from Pacific Magazines is holding relatively steady. (Its non reliance on naked women is of course another reason.)
However, assuming it is a category specific move away from print, ACP’s claim to be investing in digital does look thin, for male readers at least. As I write, much of the content on the FHM homepage is weeks or even months old, and there’s no app for either FHM or Zoo.
Newspapers
Meanwhile, Fairfax got the worst of the headlines. The SMH’s Monday to Friday sale was down by nearly 12% year-on-year to well below 200,000, a historic low. It was down nearly 8% on a Saturday with sister Sunday title the Sun-Herald down by just over 8%.
Fairfax has given a few indications that its strategy is to stop so many cheap giveaway copies to concentrate on genuine readers. One hurdle to direct comparisons is that the way these are declared changed after questions were raised about potentially dodgy deals (not that the new system is much better at identifying genuine full price sales).
So is that why The SMH fell so badly?
Take out the different types of bulk giveaways and the latest sales number was just over 168,000 (although I stress again, that doesn’t mean all those copies were sold for $1.50 apiece.)
A year ago the same maths would deliver a number of nearly 192,000.
That gives a year-on-year drop of about 24,000 – so still just over 12%.
So the idea that the fall is about stripping out “bad” copies doesn’t really stand. The top line sales are down by slightly more.
Which is why the digital numbers start to matter. At some point the newspaper publishers need to start putting on digital readers at a faster rate than they drop out as print readers. Which is why the new Fairfax data comes at an important moment.
Let’s take it a line at a time.
Monthly unique audience
Audience matters more as a reach argument in sales negotiations, but is also relevant when sales conversations are based on sponsorships rather than CPM.
Last year, Nielsen puts the SMH’s unique browser audience at about 2.8m and The Age 1.7m. because the IAB pushed Nielsen into a hybrid panel and cookies system, those numbers aren’t comparable to previous years. But hopefully there is less double or triple counting of people across multiple devices than previously.
Average time on site per person per month
About three-quarters of an hour for both – down slightly on 2010, but again the methodology had changed. I honestly don’t know that this number is that helpful, except as the first point on a trend graph. Compare it to TV or radio and it seems low.
Page views
We’re still with publicly available data at this stage. Insofar as you need to deliver a number of page views to charge your CPM, it matters. About 150m for the smh.com.au and just over 100m for theage.com.au. On a CPM guesstimate of $5, that equates to revenue of $1.25m. I hope that’s per month, not per year, although the data is not clear on this point (I’m waiting to hear back from Fairfax).
Video streams
Video is where it gets seriously interesting. Across smh.com.au and theage.com.au there were more than 10m video plays per month.
A few points occur around that. First, that’s up about 20% year-on-year, according to Fairfax’s numbers. It’s solid growth, but it’s not the kind of exponential growth I half expected to see.
And let’s remember that it was driven by autoplay. You can see why Fairfax is reluctant to turn it off. Call it a CPM of $50, and that’s a half a million dollar a month revenue stream.
Although, to put it in the context of TV, something like Seven’s My Kitchen Rules would deliver that number of advertising impacts in a single ad break, with room to spare.
Mobile sites
Across the two mastheads, mobile usage and page views have both doubled – although mobile traffic still makes up less than 10% of the total.
Phone apps
Again, the apps numbers are fascinating reading. You can bet your life the staff of News Limited were scrutinising this data.
However, this number is not as transparent as it first appears. Based on the fact that tablets are broken out later, we have to assume these are smartphone apps only and we don’t know the breakdown between paid and free. (Edit: Fairfax now confirms this)
There are five SMH branded iPhone apps in the Apple store – SMH news (free), SMH Good Food Guide (free), SMH Good Cafe Guide ($2.99), SMH Good Food Shopping Guide (99c)and SMH Everyday Eats ($2.99). There’s just one for The Age news offering which is also free. (Edit: Fairfax says that the only apps included in its data are the news app for The Age and smh.com.au, not the other branded apps. It is choosing not to release the value of in-app subscriptions)
The number of downloads to date sounds kind of impressive. Until you think about it. Around 75,000 of the SMH app and 62,000 of The Age.
Fair to assume that most were for the free news app. Let’s be generous and assume that there were 20,000 downloads of the $2.99 apps. That’s only $60,000.
But the number that shocked (and depressed) me was average daily unique browsers. Just 3576 for the Herald and 2924 for The Age. So about 5% of those who downloaded the apps use them every day.
Tablet apps
This is slightly more encouraging. More than 200,000 downloads of the SMH app and just under 200,000 for The Age. The apps were free, but the numbers were still decent.
But, but… the average daily unique browsers number is disappointing – not much more than 10% of those who have downloaded the app use it every day, by the looks of it.
I’m also suspicious of the claimed monthly page views via tablet of 33m for the SMH and 29m for The Age. I suspect (and am waiting for Fairfax to clarify) that Fairfax has counted both in-app views and also those reading the website and happening to use their tablet to do so. Clearly this is key information for in-app advertisers that could do with being broken out separately. (Edit: Fairfax says this figure is for page views within the apps only. I remain a little doubtful that 26,000 daily browsers would really generate 33m monthly page views although it’s not entirely impossible.)
Digital editions
I’m less surprised at how low this number is. Three or four thousand people across each of the mastheads choose to read what is effectively a PDF of the paper. It feels like the worst of both worlds – the lack of interactivity of a newspaper, combined with the lack of tactility of online. it never seemed a great strategy.
Taken as a whole, these numbers give a snapshot of a company in transition from print to digital – and finding it tough, but no tougher than any other major publisher.
It’s also worth noticing that these are only the numbers from Fairfax’s Metro media division. Also lost in the noise was a separate announcement from the Australian Financial Review that it’s total online subscriptions now sit at 10,987. Which based on a sub of $680 would be $7m (although there are probably a fair few bulk deals in there).
However, by putting the numbers out there, Fairfax has given a better picture of what everyone is dealing with. You may recall that to coincide with the launch of The Australian’s paywall, News Limited launched The Future Of Jouralism website, to foster informed debate about the digital transition. A good contribution would be a similar level of transparency.
What these numbers make very clear though is that the transition is going to be hard.
Tim Burrowes
I’m one of the few thousand ‘Digital Edition’ readers. I may be a luddite but I still find the other App and website seem to have a different editorial tone. I hope the small numbers don’t mean they are going to ditch it.
User ID not verified.
@Tim re: Time on site being useful or not
In 2007 I made the mistake and I’m sure other publishers did too of making our site (my247.com.au) a lot less useful by spreading content on as many pages as possible to increase the number of page-impressions per user. Ultimately equaling a crap user-experience.
Today, we’re experimenting with a new technology that allows navigation of our entire site and user-experience via one page. So our new stats will register as one unique visitor and one page impression. Already we’re seeing the results of a much higher time per page better .
So for us going forward it’ll be about how much time we can show that ad on screen for – not clocking up unnecessary page-impressions as our metric.
Cheers
CK
User ID not verified.
Phew! Thanks for drilling down into these numbers. It was a useful read…. Definitely an industry in transition.
I think you might have mixed up your words (?) slightly on this bit btw: “the lack of interactivity of a newspaper, combined with the lack of tactility of online.”
(smiley face)
User ID not verified.
Hey Tim,
great article.
One comment I would add re FHM. Having been around a little bit longer than you (and only a little, I hasten to add) this phenomenon is not necessarily new. I can remember when in the late 80s, Penthouse had an ABC Circ of 150k+ and Playboy Australia was 50k+, and they both fell off a cliff in the early 90s, heavily affected by the increasing penetration of the Internet, and the fact that they didn’t adapt to the changing times.
I can also remember working at Australian Personal Computer in the mid 90s, when paid Circ was well over 100k a month (on a $9+ cover price), and it declined not because of the Internet, but because the quality of the product declined, and they went away from their traditional audience and tried to go very consumer.
I think you’re seeing some of that now – marketing spend is definitely down at ACP, but so is the level of spend on editorial staff, with an inevitable drop in quality. I think FHM is definitely a victim of that.
I also think that the 2-35 y.o guys that used to read FHM and Ralph have grown up a little bit, and have been replaced by a generation who don’t buy magazines, as they access content via smartphones and tablets.
cheers
matt
User ID not verified.
Also, fantastic online men’s magazines like the massive AskMen are really putting archaic print mags to the sword – their audience is up roughly 35% YOY. Pretty impressive.
User ID not verified.
I suspect the print numbers are being held up by the number of papers being given away at events, Fitness First, the airport et al. I suspect that accounts for quite a chunk of the circulation. The real drop in numbers would look a lot worse without the free giveaway. Both Fairfax and News are doing it.
User ID not verified.
@Chris,
I guess you will see higher per page times if you only have one page.
How do you work your SEO?
User ID not verified.
Big job Tim. But you need to take another step I think. If an advertiser is estimating reach they’re generally thinking of demographics. Some want all people, which is fine. But even they want people within their brand reach. Which is where the multichannel data idea gets messy.
How many smh.com.au page impressions actually come from people in Sydney? Nothing like the zilions they report, I’d guess. And how many page impressions are generated by auto-refresh etc? Finally, how many are there to look at boobs or whatever (and so are hardly what we might call ABs)?
The print brand has a physical footprint. Because its paid for, generally, it has a client. And because it’s identifiable it has a demographic.
On the current evidence it looks like Fairfax has not only given up on print, it has given up on Sydney and Melbourne and its AB demographic?
User ID not verified.
@ Simon van Wyk, the circulation issue you raise was discussed in the following thread: https://mumbrella.com.au/abcs-smh-drops-12-rival-accuses-fairfax-of-abandoning-print-74083#comments
User ID not verified.
We’re quite proud of our online numbers for the men’s lifestyle category – 420k+ uniques for January (Nielsen) and consistently growing year on year.
User ID not verified.
Simon, those ‘paid promotional’ copies (they are dropped-off under a paid – discounted – arrangement which anyone can take) are capped at 2% of the final paid circulation figure. This used to be 1% and is now 2%, so I fail to see how they ‘prop up’ the headline figure. Plus they are clearly reported and have been for at least five years. While there maybe shit-loads more dropped off, all returns must be accounted for (zero returns is a BIG red flag to an auditor), and if they exceed 2% then all those extra copies are simply ignored.
Tim, you refer to the various ‘buckets’ of distribution that are not traditional newsagent or subscription sales as “bad” copies. Why is putting your product in the hands of business flyers, opera goers etc “bad”. Sure, the recipient doesn’t pay cash at the time, but there is money exchanged so it by definition is “paid”. Also, as I have explained before focussing on “fully paid” copies is the tail wagging the dog. It would mean that the circulation metric is dictating the consumer pricing policy of the publishers. It would be like NOT counting products sold by Coles or Woolies because prices are “down, down and staying down” – which would be plain stupid.
It would also be the equivalent of forcing all digital publications to have the equivalent of a ‘cover price’ and enforcing that in order to audit the PIs they must be paid for. That would be antiquated I am sure you agree.
What I think we need to focus on is not the price paid (impossible to audit – how do we know that each agent charge the cover price for every copy sold) and focus on the distribution channels. So we could have newsagent and sub-agents, delivered subscriptions (home and office), school copies, university copies (I know if I was a poor student I’d appreciate the discount), airport copies, accomodation copies, event copies, digital copies (PDF and richer format) … and whatever else is dreamt up as a distribution channel.
Does that make sense – look at the channel and look less at the cover price and cover revenue?
If so, that is basically what the rules already are, or for the bits that are missing, probably soon will be.
User ID not verified.
Simon Van Wyk – why are you making comments on something you clearly know nothing about. The comments above are incorrect, not sourced and lazy. I hope the advice you give clients is a little less flimsy.
User ID not verified.
@ Chis K and AJAX users
Be careful with only creating one page impression per session because tagging-based tracking generally requires two page impressions with timestamps to calculate the time.
Many sessions may be counted as zero duration if there’s no second timestamp, so this technique could backfire if the goal is to increase duration metrics.
Most AJAX (or similar) dynamic sites count major content changes on the page as extra page impressions so it’s not a common problem.
Alexx
User ID not verified.
Tim
A good analysis of the digital numbers. The monthly page views on tablets of 33m for the SMH and 29m for The Age divided by the number of average daily unique browsers gives a figure of about 40 pages per browser a day. This seems a lot. On the other hand I agree that the digital edition sales of about 3000 seems very low.
It would be great to have the same analysis for the Financial Review where the online subs are still a miserable 11,000.
Hey Tim, I read this while on hold to the ATO. Amazing to have a better understanding of print now thanks to you delving into the underbelly of this part of the business. Very enlightening for someone like me who is an avid consumer of newsprint and magazines but not au fait with the details. I found myself quoting your piece to my local newsagent. Word of mouth – still the best way to spread the love….J BTW we subscribe to the SMH, read it online too and hate the autoplay and video in general. We also subscribe to Vogue, Vanity Fair and GQ – (US). Wonder what the figures are for the AU edition of Rolling Stone…?
User ID not verified.
Good read,
Everyone in Media Print has had their head under the covers for years.
It had to happen,the lack of vision from the top only made things worse, what is it with media companies, why are they so stale, there are certainly alot of opportunities still out there for them.
User ID not verified.
There’s a sense from these numbers that the engaged weekday daily audience for the newspapers is around 60,000 with a very long tail. The long tail consumes the papers like a freeish OG1 metrowide paper with the engaged audience split across devices and paper segmented by device ownership demographics.
Be useful to see the saturday numbers split across distribution media – likely to be more paper and less device share.
Whats an App pageview – pageview counts from Apps depend on the architecture of the app as well as activity so they are not usually a directly comparable metric.
User ID not verified.
Jerrys, every audited publication (e.g. the Saturdays and the Sundays) has the distribution channel volumes reported separetly – and has for around the past five years.
It feels as though a lot of pundits look only at the headline number (as carried in the press) then freely comment with their interpreation while not doing a deeper dive into the data.
Of course it could be that the more granular data is only released to ABC members and therefore is not widely released,. Yes, the publishers and media agencies PAY to be able to do such analyses in order to make informed analyses, comments and client recommendations. It is folly to expect data that members pay for to be immediately made available for free in the public domain, as clearly there would then be no revenue stream to pay for any more auditing. Paul Dovas has elsewhere indicated that if serious pundits want to contact him in order to write informed stories then they are free to contact him.
But Jerrys, I wanted to ask what you meant when you were referring to App pageviews. If you are meaning comparing App pageviews to paid circulation then you are comparing apples and oranges.
Circulation data is based on paid copies. Pageviews is closer to a readership metric than to a sales metric. I suppose you could take paid print copies, multiple that by the average pages per issue, and then discount that by a factor that represents the proportion of the printed copy that people read to come up with something vaguely comparable.
User ID not verified.
John
The Fairfax numbers are a bit of this and a bit of that – perhaps deliberately.
All the numbers Fairfax presented are in essence daily except for the web. The web pageview/user numbers would be better presented as Monday-Friday/Saturday like the print numbers. Not really any reason to keep the historical legacy of monthly numbers for digital particularly for properties like media.
Whenever a new access technology comes along the what’s a pageview issue returns – both whats being counted and what it means – client side and server side.
The online advertising opportunity to see is increasingly being detached from the pageview for media properties. There isn’t a pageview number attached to the paper readership and it’s increasingly redundant as a topline digital metric for many online properties except for providing publishers with a very large number.
Adding a time spent number for the digital devices would be useful.
What would be your estimate if the print numbers included an average time per reader.
User ID not verified.
First, I agree that monthly Page Views (and Page Impressions) are among the less meaningful metrics. PIs do come in handy when looking at the total ‘weight’ behind a campaign, particularly monitoring the weight of yesterday’s buy etc. That is, as you point out daily PIs do have some utility as opposed to the less meaningful monthly PIs.
The ABA circulation data is “average issue” circulation and mirrors the daily PIs (with the caveat that not all publications have to report DOW circulation)
I think that a ‘page view’ number attached to readership would be an exercise in arithmetical gymnastics that we could do without. Of interest to a media buyer would be what the (estimated) average proportion of the newspaper was read – how many claimed to ‘read all’, read half’ and so on as a guide to depth of reading or engagement (i.e better chance the ad was seen). To do this at the page level would be folly.
My gut feel – zero empirical evidence but some ethnographic experience – is that the time spent per page by a reader of a newspaper is higher than that of an online reader. The reason I say this is that the eye tends to skim the page of the paper (a broader real-estate) whereas online the eye skims the screen (a smaller real-estate) and then the hand has to scroll the mouse in order to keep reading below-the-fold. For many people if the brain hasn’t said … gees that looks interesting … then the hand doesn’t scroll. Order of magnitude would be … damn who knows … something between 2 and 5 times longer per page with the physical paper. Anyone have a better gut feel?
A word of caution though, while I am a big fan of “time-spent” I do NOT think that this allows for direct comparability across the different media. For example, we know that TV averages around 3 1/4 hours a day (and has since the early ’90s) – but would you seriously expect someone to look at a billboard for that long? We also have to consider what proportion of time spent with each medium should be considered valuable ‘advertising time’. We know that around a quarter of TV is, probably the same for radio, more for print/press … and who knows what proportion of the total online time is ‘advertising time’ (I know that most of mine is spent with email and I suspect that the same applies to most of the heavy online users). In a nutshell, I think time-spent is poor for comparability between media, but good for comparability of media properties within media. I also suspect that as you point out, for online device type would be important for time-spent (and yes I think tablets may rank #1).
User ID not verified.
A great resource, thanks for the analysis.
Good to see others struggling with Digital Editions, which i believe are a fairly pointless excercise having offered these for free for several years as an alternative medium for our hardcopy.
The insight that publication apps might also be less favoured when compared to good old fashioned news websites with a mobile device version is also valuable.
I note that the plethora of options are rapidly driving prices for off the shelf magazine app creation down…if apple enters this market as they have done with book creation, the gig will surely be up for those charging custom development rates for what are essentially SAAS offerings.
@Dave #16
If you don’t mind I’ll exclude myself from the ‘Everyone in print media’ characterisation. Print publications with intelligent revenue models and intelligent cost controls are thriving….there might not be many of them getting around but it’s a mistake to see digital as the final inevitable destination for all print titles…the good ones will learn to walk both sides of the street and i suspect the titles that havent rushed to undermine their print products by chasing the digital dream will be well placed.
User ID not verified.
Given that the audience is so obviously moving away from print, where the money is and the pseudo audience online is delivering dimes, we should be looking instead at the profit figures. This industry is on a cliff and the bunge is fully extended.
User ID not verified.