Upcoming Industry Conferences
Streaming Media West [19-20 November 2019]
Live Streaming Summit [19 November 2019]
Past Conferences
Streaming Media East 2019 [7-8 May 2019]
Live Streaming Summit [7-8 May 2019]
Content Delivery Summit [6 May 2019]
Streaming Forum [26 February 2019]

Streaming Media
Magazine

Winter 2019
Subscribe


Digital Editions

Current Issue:
Spring 2019 (Sourcebook) 

 

 

 


Broadcasters and Advertisers Still Drinking the Audience Measurement Kool-Aid
What's better: "Millions" of viewers as defined by sketchy science and unreliable data, or thousands of viewers as defined by accurate streaming logs? Old media continues to bank on the "millions," and that will eventually lead to their demise.

Sitting on the train on the way back from the Westminster Media Forum meeting on TV and the Second Screen last week, I found myself full of raging emotions provoked by the discussions I witnessed.

The event was very well attended and organized, and I look forward to attending more of these discussions, which aim to bring together key stakeholders in the UK and European media space.

However, as was the case with the recent Radio & Internet News RAIN Summit, I am not quite sure what to make of my emotions.

The day was divided into two parts, with the first part focusing on social media and its role as a part of the TV broadcast ecosystem, and it appeared as if the majority of attendees were focusing on opportunities for social media. While there was little doubt that social media has significant uptake in the context of TV broadcasting, frankly even the Twitter "Guest of Honour" (Dan Biddle, head of broadcast partnerships) was clutching at straws about the real commercial value proposition. The repeated use of the new currency in this sector—“engagement”—was underpinned by extremely vague data, beyond the repeatedly highlighting of volumes of tweets sent during particular broadcast programmes.

The BBC’s head of social media for iPlayer, Alister Morgan,  could afford to be more pragmatic in his comments, since there was no pressure on him to show more than "brand value;" as a public service broadcaster the BBC is not tied to monetising social media in the way that others are. And while there was an enourmous amount of data about how much TV broadcast viewers were using a second screen while the TV was on, it was only the BBC speaker who highlighted the extremely obvious: that there was little or no measurement that what these users were doing on the second screen was even related to what was on the TV.

Rather than assuming that they were "augmenting and enriching" the broadcast experience, broadcasters need to acknowledge the lack of quality and engagement of the broadcast audiences.  It's just as likely that audiences were simply distracted by the availability of their own interactive choice of content consumption as it is that they were engaged with something relative to the content or advertising on the TV.

Still, attendees seemed more than willing to accept all the statistics relating to audience measurement around TV broadcast as, where any facts and figures about "new media" were viewed with suspsicion and doubt. I asked Lindsey Clay, managing director of Thinkbox, about this, highlighting the fact that it is in the interest of the established media and advertising industry to stick to the "science" of BARB, RAJAR and other "trusted" audience measurement which underpins the entire economy of the advert-funded broadcast media industry, rather than to reflect on the fact that if something online has an audience of 200 people it is unlikely to be getting a broadcast audience of 2 million.

The lack of trust in real usage logging from second screen systems was pervasive, while the absolute trust in the measurement of these extremely speculative "science"-based extrapolations produced by BARB, RAJAR, and the like was incredible. This is where I start to get quite emotional: We in the streaming media community can practically tell you the colour of each audience member's eyes. We have to, because if we couldn’t then our CDNs would not be able to bill us for each unicast delivery. Note that I am not talking about the fudgery of web advertising, where, for instance, banners and "measurement pixels" can be manipulated to fix page impressions.  I am talking about streaming sessions—something that simply cannot be faked without incurring huge infrastructure costs which negate the motive to fake such audience levels. I believe that CDNs know exactly how much video is consumed. To the byte.

In contrast, BARB has a few thousand sampling boxes placed around the UK broadcast audiences’ households, and myriad qualitative modifiers to extrapolate that data from a few thousand use cases into the "millions of viewers" statistics that underpin the TV advertising industry. And that industry is a very powerful lobby, one that drinks its own Kool-Aid and merrily spends millions on reaching supposed millions of viewers—and then equally measures the effectiveness of its campaigns on equally weakly extrapolated models, ensuring that the marketing directors who commission the TV ads don’t get fired.

So when I pointed out the problem of marrying the second screen’s more accountable measurement with the first screen's more established, but far less accurate data, there was a nervous laugh across the room. This is the very entrenchment that I feel also leads someone like Lindsey Clay to open her presentations with a naive attack on Netflix and YouTube’s public positioning that they were going to eclipse and replace broadcast TV. "This is simply not going to happen," she said, and made some scathing comments about Reed Hastings' bullishness.

She is obviously deluded by her own data—a common problem for someone who sells that data to advertisers.

YouTube has hundreds of millions of users consuming billions of programming hours. The BBC has tenss of millions of users consuming tens of millions of programming hours. And that is by anyone’s statistics.

Likewise Netflix has some 53 million subscribers, against BSkyB’s 10.6 million.

So while Clay would probably point out the global vs UK nature of those figures and begin a triage of why her comments can be extrapolated to define subcategory after differentiation, etc., the fact is that 53 million people are engaged commercially with Netflix, and only 10 million are engaged commercially with Sky.

If that isn’t at least something close to an eclipse, I honestly have no idea what is.

Given that in the context of TV broadcasters and advertising there are billions of pounds/dollars at stake, it is amazing how much people want to believe the conjecture of organisations like BARB, and yet how little real dialog occurs around the accountable data presented from the CDNs that deliver online content when the BARB fans compare the two.

Returning to the RAIN Summit I attended some weeks back, I saw the same thing: a room full of digital ad agencies working out how to descend like vultures on music streaming and convert each and every audience touch point (apart from the contact with the CDN itself) into something from which they could extrapolate data. Most of these idiots are still relying on banner adverts displayed on the same pages as the streams, and using page impressions (or the like) to convey audience measurement to their potential advertisers. And these advertisers want to hear "millions" before they part with their money, so the measurement systems find a way to measure millions—when in fact if you aggregate all the statistics from all these bizarre measurements systems together a country of 80 million such as the UK probably has 200 or 300 million people actively listening to streams at any one time. (And yes, I made up that number to emphasise my point..

These people need a dose of reality. They need to respect that the biggest resistance to unicast streaming media happens when you get an audience of 180 on a webcast that the client was anticipating 40,000 to connect to. But we have real data, and with that to fall back on it becomes hard to lie. The "finger in the air" measurement, that the advertisers use for TV advertising has become so so so entrenched that they simply cannot face making the accounting adjustment that streaming has made undeniable—and so streaming media has to, for now, be seen as "second screen" lest the traditional economics will crumble.

Pretty soon though our kids' generation will grow up, and they simply won't engage with TV ad brands. They wont be cord-cutters, they will be cord-nevers—and at that point the products being sold through TV ad campaigns simply wont… whats that magic word… "engage" any customers at all. And all those companies will either embrace streaming as the first screen—whatever the audience sizes—or they will die.

Related Articles
The industry is challenged to find a single trusted trading currency that can look at both linear and nonlinear viewing in a unified way.
Online access could look greatly different in the near future. No industry is too big to fail, and that includes today's biggest telecommunications companies.