Yesterday two major TV networks breached the rules on how to use the new OzTAM catch-up TV data. Nic Christensen breaks down what the numbers mean and how they should be interpreted.
It’s in part a sign of how eager the TV networks were for accurate online audience figures that both Seven and Ten yesterday decided that “close enough is good enough” and unfairly inflated their total TV audience figures.
For those not across the intricacies of the new metric (which I’m about to jump headlong into) Seven and Ten’s sin was to combine the new VPM ratings, which measure the average number of devices watching a program nationally, with the TV ratings, which measure the average television audience in the five major metropolitan cities.
Put simply, they were comparing apples and oranges. Well, to be more accurate it’s more like apples and orange crates that definitely contain fruit and, it’s thought, an unspecified number of oranges, but more on that later.
What is the OzTAM VPM Report?
Put simply the OzTAM Video Player Measurement (VPM) report (it’s difficult to think of a less captivating name for what is a new and interesting metric), referred to by some as, essentially, ‘the digital ratings’, is Australia’s first official metric for viewing internet-delivered TV content.
At the heart of the report is something called the ‘VPM Rating’ which is the average number of devices that played the content throughout a program’s duration.
From the outset, it’s important to note that this number is very different to some of the ‘live stream’ numbers that were being floated by the likes of Seven over the summer with the Australian Open.
Throughout much of January, Seven was highlighting live stream numbers that ranged from 100,000-500,000 a day, but it’s important to recognise that there was no ‘time spent’ or ‘duration’ attached to any of these numbers, meaning that someone who watched a two-hour match and someone who jumped in for 10 seconds counted as same.
More concerning, every time someone reentered the stream or their connection cut out and a new stream started (likely when we’re talking about mobile connections) a new stream was counted.
Seven, in one of its more brazen sales pitches to market, deliberately put these numbers alongside total TV audiences while declining to release any figures on the number of devices that had downloaded the app or the number of users registered – which would provide a far more accurate metric.
This new metric will hopefully avoid this type of mischievous side-by-side comparison of two things that are not the same but which appear to a lay reader without knowledge of the space, as if they are.
A VPM Rating is the total minutes of a program played across all devices divided by the content duration and rounded to the nearest thousand.
How OzTAM displays the VPM ratings.
As OzTAM noted in its guidelines last week: “You can also think of this as the average number of devices playing the content across the program’s duration, which is similar to the way television program ‘average audiences’ are calculated – albeit for devices rather than people in the VPM Report.”
What the VPM Report is not.
This is brings us to what VPM is not.
It is not yet an accurate measure of audience (ergo yesterday’s snafu with Seven and Ten). It measures the number of devices connecting (via a media ID) to the catch-up players of all major TV networks – ABC, Seven Network, Nine Network, Network Ten, SBS and Foxtel – and will, in the coming months, also have numbers for live streaming (more on that later).
OzTAM infographic on how the metric works. Click to enlarge.
The problem is that right now OzTAM doesn’t know how many people are behind the screen. Sure, there is one person who is connecting to watch that content but they might have a smart TV and be catching up on the latest episode of Molly with their family of five.
The even bigger problem is that these are national figures.OzTAM measures only the five major metropolitan cities but by default of how catch-up TV and live streaming work these figures are national.
To understand why this is important just look at the lawsuit billionaire media mogul Bruce Gordon launched last week against Nine, claiming it was undermining his business model by broadcasting in his area.
Related content: ‘The great geoblock of Wollongong’: Bruce Gordon tells Nine to turn off live streaming, but to what end?
At this stage it’s arguably important for these two figures to remain separate. While both factors may seem small they are important – particularly to media buyers, who aren’t happy with the TV networks taking very different sets of numbers and deciding, off their own bat, that it’s okay to lump them all together.
As one senior media agency executive, who declined to be named, told me: “The additional data on today’s audience behaviour is of course welcomed, however, we are looking at oranges and apples and the two can’t simply be added together.
“I see this more as an additional lens we use to quantify behaviour we know is happening.”
So what do the VPM numbers show?
In many ways, the first VPM numbers confirm what we already knew, but didn’t really have hard numbers on.
By giving the industry an average number of devices that watch catch-up TV and telling us which programs these devices watch we have a new perspective on consumer viewing habits.
Perhaps, unsurprisingly, the likes of Seven’s Home and Away dominates the Top 10 with around 50,000 devices watching each episode of the TV soap.
In fact, across the various networks the shows that appear to do well are those that, generally, skew to a younger audience (again a relatively unsurprisingly revelation).
On Seven it’s Home and Away and MKR dominating; on Nine it’s Here Come the Habibs, The Farmer Wants a Wife and new episodes of The Big Bang Theory; on Ten it’s The X-Files, I’m A CelebrityGet Me Out Of Here! and Neighbours.
All of this is positive, particularly as OzTAM is claiming that, in total minutes, the new metric has already added about 2% to the total TV viewing pool.
And much of it will be very attractive to advertisers and media agencies, if and whenOzTAM and the TV networks can put demographic data behind it.
What are the next stages for the metric?
Yesterday’s data release was the first step in a process that has some way to go.
Dickens: the minute pool of TV is in great shape.
The next stage of VPM Ratings is the live streaming component, which will likely further boost the total size of the online audience.
How much of a boost is unclear at this stage but Clive Dickens, chief digital officer of Seven, who spoke with Mumbrella last week, claims it could be as much as an additional 2%.
“The next stage is we will drop live streaming into that and we’d estimate going into this year that live streaming minutes could be another 2%,” Dickens said.
“If you put that in the context (of declines) and see that two plus two you start to see the minute pool of TV is in great shape.”
Personally, I’d be surprised if the size of the audience for live streaming, which only really kicked off last year, is already as large as the catch-up TV audience. Hopefully the industry won’t have to wait too long to find out.
Dickens’ reference to the decline of the TV audience is interesting and significant.
Much has, rightly, been made of the fall in TV audiences in the past year or two. While the TV industry often tries to point to the fact that the reductions, overall, are not that large, there are clear and arguably precipitous declines in prime-time audience (the important one) and in key demographics – particularly the younger ones.
Media buyers who are focused on the all-important TARPS, which media buyers trade, show very clear falls when you look year-on-year over the past decade, as audiences moved online.
Source: Nielsen data. Click to enlarge.
Yesterday’s announcement is important (as will be the live streaming numbers) but the real focus will be on getting demographic and panel data behind the VPM figures.
As Dickens, who was part of the oversight panel on VPM, notes: “The next stage is to hybridise the data with people. The most important data our clients will want is ‘who is this person? is it a 20-something female or a 50-something male?’. That is something we will need the panels for.”
Indeed. And by 2017OzTAM hopes to be able to turn what is currently national device data into real metropolitan audience data that can then be merged to gain a total audience figure.
Until then, however, some of the TV networks may need to cool their heels.
Nic Christensen is the media and technology editor of Mumbrella.