Adobe reports that while TV Everywhere viewing -- authenticated viewing of premium TV content on digital devices -- was up 63% to 159 billion online video starts in the second quarter of 2015 over the same time period a year ago, overall TV Everywhere growth is slowing.
By way of comparison, TV Everywhere usage growth was up nearly 250% from the first quarter of 2014 over the first quarter of 2013.
Adobe lays blame on setting up TV Everywhere through a not-so-smooth sign-in process. TV Everywhere efforts, backed by pay TV providers -- cable, satellite, and telco companies -- ask consumers to input account information in the sign in process to access video content.
The report says online video advertising viewing (non-authenticated content) showed strong growth across every genre of content –- up 30% -- with reality TV viewers watching the most advertising per video view, at 3.9 ads per view. Drama TV shows had the largest increase year-over-year, when it comes to ads per view -- an 83% rise to 2.72 ads per view.
advertisement
advertisement
Premium TV Web sites saw tablet viewing growing by 29%, smartphones by 16%, and desktop viewing by 15%.
The report says share of viewing via connected TV devices is shrinking quarter to quarter. But Apple TV has gained 10% during the period, to a 12.8% share. Other digital TV streaming devices, such as Roku, showed no gain or declines.
Abode’s data comes from aggregated and anonymous data from sites on consumer video viewing in 2014 and 2015. Some of that data includes 1.49 billion TV Everywhere authentications and 300-plus different sites and apps.
If I view something on my ipad/iphone, but I use airplay to send it to my TV, is that counted as a "tablet/smartphone" view? Or as a "Connected TV" view?
I suspect it is the former, and that 'Connected TV' counts only views that originate in a connected TV app...If that is the case, then this data is pretty much useless...does anyone know?
It's always interesting to translate these big sounding numbers---in the billions---with their actual meaning in terms that most of us are familiar with. For example 159 billion "views" in a quarter represents only 12 billion per week. Divide that by the U.S. population and you get about 38GRPs, cut that down to a daily figure and we're talking about 5 GRPs --- that doesn't take into account the length of each "view" or whether anything was actually "viewed.
Luke, it depends on the type of metering used in the panel.
Here in Australia we use audio matching. That is, we send the audio from the TV set (not the iPad) to a central processing area. Simultaneously, all broadcast TV (FTA and cable) is also recorded so that we have reference signal.
When the panel data arrives each morning, eash TV set's audio is then 'matched' to the reference signals. If they match we know which channel was being 'viewed' (well at least tuned). If there is no sound on the TV but there is a picture there is no match. We also analyse the 'change-line' data - if we see continuous tuning with no activity (channel switching, volume changes, pauses, remote control activity) then that activity is rejected (e.g. they could be asleep).
If only part of a programme matches the audio (e.g. the FFWD through the ads) then that proportion of the programme is not 'credited'. For example, if someone watches 45 minutes of a 60 minute programme they only count as three-quarters of a viewer (but they are deemed to be reached).
So, in the instance you cite, yes your viewing would be matched and your viewing would be credited. If you watched part of the programme on the iPad without airplay then that proportion would not count to the TV ratings. When you couple this with 'away from home' viewing (e.g. sport in a bar), TV ratings are a 'low ball' estimate.
I'm sunsure of the metering used in the US, but I hope this helps.