I’m not sure about EMMA

Readership numbers make my head spin.

Not least because it often feels like the more I learn about how they work, the less I know.

It will be hard for you to go near the trade press over the next few days and remain unaware of EMMA (Enhanced Media Metrics Australia). Not least because it’s being launched with the first big above the line marketing campaign conducted by industry body Newspaper Works that I can recall in the organisation’s seven year history.

When I first starter writing about media, the metrics were hard to get my head around. First there was circulation – as in how many copies had been physically printed and distributed. Then there was readership, effectively a guesstimate of how many times each copy had been passed to somebody else.

Gradually, I began to learn there was such a thing as ‘good’ circulation –  copies of a newspaper or magazine bought at something close to the full price by the customer; as an advertiser, those are people you want to get near. And ‘bad’ circulation, perhaps made up of bulk deals and giveaways that might not even end up in human hands. The level of transparency about the difference between those two types in that audited circulation number will depend where you are in the world.

Readership is even tougher. Because, how on earth can you really tell how many people have read any given title?

I remember the boss of a metric in the UK for regional newspaper readerships telling me the ‘science’ behind their numbers over lunch. There may have been slightly more of a twist, but effectively it was to take each newspaper’s reported circulation and multiply it by a specific number to give a magic readership figure. I seem to remember it was about two-and-a-half.

In Australia, the readership number has long been the domain of Roy Morgan Research.

A common criticism from some parts of the industry has been that the Roy Morgan service lacks transparency. A media buyer once described it to me as “Roy’s black box”, meaning it is hard to tell exactly how they get to their numbers. On Friday, News Corp – a long term critic of Roy Morgan – said in a statement to Mumbrella “Roy Morgan Research did not respond satisfactorily to countless requests, over many years, to explain discrepancies and errors in its data”.

I have a fair bit of sympathy with that view. The reason we usually choose not to report the Ry Morgan readership numbers every quarter is because they lack the transparency of a joint industry currency with media owners, agencies and advertisers all at the table.

And it was undoubtedly a factor in the decision by Newspaper Works, funded by the major media owners including News Corp, to launch a rival service.

I must confess though that right from the beginning, I had doubts about whether the intentions behind the new survey – which has now been in in the works for an interminable four years and two months, by the way – went further than that.

At the time, I was sceptical. I wondered if this wasn’t an attempt by the newspaper industry to take control of the data for its own benefit. Not least because of the historically poor relationshiop between News Corp and Roy Morgan.

The comment thread under the piece I wrote in June 2009, includes a great discussion among researchers about what it takes to create a genuine joint industry survey with media owners, agencies and advertisers in equal partnership.

Along the way, the metric that was to be called EMMA had its ups and down. Not least when Fairfax Media withdrew and rejoined.

It wouldn’t surprise me if the abrupt departure of Newspaper Works CEO Tony Hale earlier this year might have been related to the failure to get anything launched during his four years running the project.

But finally, we’re here.

The first numbers are available.

So it now becomes a question of sanity checking the results by comparing the numbers to circulation. And there are some quirks, to say the least.

As it happens, when the numbers came in, I turned, first of all, to a title that ACP (now Bauer), withdrew from being audited a year ago. At the time, soft porn title People losing about a quarter of its sales a year. Being generous, if it was still being audited now, I’d expect its circulation to be down to about 20,000.

So how does EMMA assess the title’s readership? It reckons the readership of People is 222,000, or a pass-on rate of something like 11 readers per copy.

Now, even given the close-knit nature of your typical building site, that seems a lot. I’d hate to see what state the magazine would be in at the end of its service.

Another curious example is Zoo. Last week, its audited number dropped to 46,871. Yet is EMMA readership is 472,000 – again more than ten readers per copy.

Let’s go a tad more mainstream. Australian Women’s Weekly, Bauer’s flasgship, had a circulation of 459,175. And apparently a print readership of 2.29m. That’s 4.9 readers per copy.

How about Cosmopolitan (which did not have a great audit result)? The audited circulation was 98,294. The claimed readership? 490,000. Again that’s about 4.9 readers per copy.

Then there’s simply sanity checking them against the real world. Let’s take perhaps the most obvious newspaper test – the Sydney Morning Herald.

According to the Australian Bureau of Statistics, greater Sydney has a population of 4.6m. And according to EMMA, the SMH has a total monthly audience of 4.5m. Which is impressive.

Actually, of course, some of that can be explained by the fact that people still read the title elsewhere in Australia.

But how about print alone, which is more likely to be geographically limited? EMMA attributes the SMH with a print edition readership of 814,000. That’s compared to a Monday to Friday circulation of 141,699 – which implies that an impressive 5.7 people are reading each copy . No wonder I can never find mine if I leave it on my desk for a moment. If I leave mine at home, that means somebody else is going to have to share it with 9.4 other people that day.

Melbourne’s The Herald Sun does even better. 416,027 sales  equates to a print readership of 1.446m, according to EMMA. That’s 3.5 people per copy.

A further question mark comes, by the way, about section readership. Although EMMA has promised all along to make this available, and does (I understand) calculate the numbers of how many people read individual sections, at the time of writing this, they have not given us the data. We’ve requested they do so, but I’m not in a position to express a view here.

And at this stage, there is also no data available to us from EMMA on year-on-year comparisons. Again, we’ve requested that information.

For future data releases from EMMA, in the same way that I don’t routinely report Roy Morgan’s readership numbers as fact, we may well be similarly cautious.

Which isn’t to say that EMMA has no value.

Readership numbers is always a statistical exercise, not a census. So a new number gives another possible point of understanding audience behaviour.

And it will also help keep Roy Morgan on the straight and narrow. It’s no coincidence that they chose Friday to announce long overdue improvements to their own service including monthly rather than quarterly reporting and online data release.

But right now, EMMA is a long way from being an incontrovertible industry currency.

Tim Burrowes

Comments


  1. Anonymous
    19 Aug 13
    12:30 pm

  2. The two biggest lies in media:

    1) Pass-on rate.

    2) Jimmy Saville’s reputation.

    It is just not true that even 4+ people reach each copy of a magazine or a newspaper across the print run. It never has been.

    Even in a dentist’s waiting room (where people now read their phones) it wasn’t true – a quick flick did not mean a reader and certainly not a viewer of ads.

  3. Ricki
    19 Aug 13
    1:07 pm

  4. Ah so because you don’t see it , it doesn’t exist?

    Good thing you’re not a statistician then Tim. You’ve just denied the validity of the whole science.

  5. Dan
    19 Aug 13
    1:41 pm

  6. A dying industry still can’t justify it’s data.

    Time to go and….die.

  7. Jörn
    19 Aug 13
    3:54 pm

  8. The start of this article is what I’ve concluded. Even with Emma

  9. Nick
    19 Aug 13
    7:08 pm

  10. Finally Tim. Someone asking the right questions. Well Done.

  11. LW
    19 Aug 13
    7:43 pm

  12. #3 Dan, at least proof reading won’t be a dying industry thanks to people like you:-)

  13. Anon
    21 Aug 13
    12:45 pm

  14. At last some balance in the debate about readership. Thanks Tim for not just re-hashing press releases and asking some though questions that need to be answered.
    One thing I can’t get my head around is if the EMMA sample is 54,000…but they only ask newspapers read yesterday…does that mean there are only approx 7,700 people being asked about newspaper reading on any given day? How does that work?