From InteractiveTV Today
THE DATE: June 26, 1974– 40 years ago this month – the first Universal Product Code (UPC) bar code was scanned at a checkout register in Troy, Ohio. The item was a 10 pack of Wrigley’s Juicy Fruit chewing gum (the unopened pack now hangs in the Smithsonian). Today, bar codes reliably track the identity and price for retail goods in more than 90 countries, millions of times per day.
It’s hard to imagine a world without product bar codes, but try to imagine this: a bizarre world where every retail company had blindly developed their own (sometimes competing) bar code standards and then petulantly expected them to work with different types of readers.
Madness you say? Well, welcome to the current (dysfunctional) world of video and advertisement identification.
That kind of madness has us talking about tables and money. Big tables. Lots of money. Hmmm, forgetting something? That’s the itchy feeling washing over the TV industry today. From an operational and business perspective, there is money being left on the table today – and the root cause is asset IDs.
“IF YOU CAN’T IDENTIFY IT, YOU CAN’T OPERATIONALIZE OR MEASURE IT; AND IF YOU CAN’T MEASURE IT – YOU CAN’T MONETIZE IT.” – Clyde Smith – SVP New Technology – Fox
In some cases – it’s a lack of audience measurement (or inaccurate measurement). In others, it’s cumbersome digital asset management (DAM) workarounds. Not to mention increasing online video ad fraud and piracy. These lost revenue (and increased expense) sources are due to an industry clinging to historical, proprietary and bespoke solutions for asset identification, tagging and tracking. It is also due to the lack of agreement on common metrics for cross-screen / cross-platform measurement of viewership and engagement. And as TV/video delivery screens proliferate (from wearables to airline seatbacks to connected cars), so does the need to validate and measure content and ad views across these devices.
But when you flip the script of a problem, you sometimes find an amazing opportunity. In 2013, CIMM (the Coalition for Innovative Media Measurement) completed a study called , “TAXI” (Trackable Asset Cross-Platform Identification). Led by Ernst & Young, 28 top US media companies participated in a 2 1⁄2 year feasibility study. The result? If adopted, TAXI would enable $2.5 billion in year-over-year benefits across the M&E (media and entertainment) supply-chain – a clear economic case for adoption relative to less than $500 million in one-time investments distributed amongst the same companies. At the core of TAXI are two global standards, not yet completely adopted – EIDR (the Entertainment IDentification Registry, led by dozens of movie, TV and ecosystem players) and Ad-ID (led by the 4A’s and ANA).
Sounds great on paper, right? But it turns out the promise of a couple billion dollars annually is just table stakes on the long road to the goal of standardized ‘file-based workflow’ for the content and advertising ecosystem. Just a seat at the table, if you will – in the high-stakes, glacially-paced poker game of supply-chain automation across TV content and ad delivery platforms. And the agendas of the various poker players dealt a hand in the game (Nielsen/Arbitron, comScore, Rentrak, MVPDs, broadcasters, movie studios, cable programmers, Rovi, Gracenote [formerly TMS], advertisers, agencies, next-gen metadata providers, etc.) are not always aligned. (You’re no doubt “shocked, shocked” by this revelation – but read on).
IT MAY BE THE GOLDEN AGE OF TV CONTENT, BUT THE MONETIZATION OF THAT CONTENT ACROSS PLATFORMS HAS BEEN STUCK IN THE BRONZE. –WK
So, TAXI made its point, but where the rubber really met the road was with the next step: Project Blueprint. Because the world still runs on a combination carrots and sticks (when it comes to industry standards adoption) – Project Blueprint was tested in 2013 to prove the feasibility – and value – of cross-platform audience measurement, without having to wait – and hope – that EIDR and Ad-ID would become ubiquitous overnight.
“Project Blueprint’s work has opened the door,” said Jane Clarke, managing director of CIMM. “We got comScore, Arbitron and ESPN (who funded the study) to step up to the plate. They proposed ‘small-scale solutions’ – getting accurate, non-duplicated viewer measurement across TV, computers and smartphones.” But how?
Turns out that Arbitron (pre-merger with Nielsen) offered comScore their PPM (Portable People Meters) – which measured both TV and radio usage (via audio). “They had the data on the TV side, they just weren’t reporting or selling it to anyone,” said Clarke. “They had all the encoding equipment still in place at the networks from their old test of Project Apollo (2006-2008).” It would seem that the goals of “single-source data” were still alive.
“So, there were only a few networks that had to install or re-install new equipment from Arbitron,” Clarke continued. “We did a small-scale study with Arbitron – it was only 500 in-tab samples – single source, same person, small scale. comScore did it with AT&T U-Verse data (from set-top boxes), matching it against their digital panel and they brought in all the digital tagging for mobile, then picked-up the census-measurement from the server. It’s called hybrid-measurement – you get a quantitative measurement from servers and set-top data (machine-based) for the volumetric – and then using their panels, you get the demographic and inter-media relationships. We got to the end and looking at the two panels of 10,000 and 500, we said ‘Wow, you can get all this cross-platform data on usage – from the same person. But, how would we scale it for the industry?”
The good news is that CIMM members are some of the biggest names in the industry — with the most to gain from finding a better cross-platform measurement solution. Disney, CBS, NBCU, Discovery, P&G, Omnicom, Unilever, Time Warner, Viacom, Publicis, Microsoft – to name a few.
Eventually, both comScore and Arbitron realized there wasn’t enough business in it go it alone – and that if they worked together, they had complementary solutions (Arbitron could have a calibration panel would work across all the platforms, including radio and tablets and comScore could bring in the Big Data approach from digital (servers and set-top data). Together, they came up with a number on how much it would cost. CIMM couldn’t afford it, but Disney’s ESPN could.
ESPN funded the first development year, and CIMM will push forward with the second. “Artie Bulgrin and Glenn Enoch have been the leaders at ESPN for this cross-platform effort,” said Clarke. “ESPN’s first year of hybrid measurement has gone well, and the other CIMM members are just starting to take the look at the data. ESPN didn’t care much about time-shifted viewing (since so much of their sports programming is about live broadcast), so now in phase 2 – we’re bringing in time-shifted viewing [VOD, DVR], as well as under-18 / children’s viewing, cross-platform ad campaign measurement, and mobile video (via the comScore duration-based streaming tag).”
“What the industry needs is the ability to continuously monitor the migration of consumption [of TV] across platforms and understand the value of one platform versus another” – Lyle Schwartz, managing partner and director of research and marketplace analysis at GroupM Worldwide. – Adweek, March 23, 2014
The TV advertising business may annually generate in excess of $70b value in the US, but it’s important to consider the wheezing, creaky infrastructure that makes it possible – and the mission-critical sense of urgency for its improvement. “The model [for tracking digital video assets] that’s existed up until now was in-house solutions, up until broadcast,” says Clarke – “when they [broadcasters] put the [proprietary] Arbitron watermark on the content. They’ve relied upon Nielsen and Arbitron to do the audience measurement tracking after it leaves their shop. The ridiculous part is that to calculate a Nielsen C-3 rating (the 3 day window post-broadcast) it takes 15 days – because they have to go back to match against their program and commercial logs, to sync them all up and verify that the as-scheduled logs are the same as the as-run logs. It’s a ridiculous, non-automated legacy system that hasn’t kept up – they still have people making phone calls and sending faxes to local stations. To top it off, people in local markets don’t even want to respond because they’re not Nielsen clients – so they get a lousy response rate. And syndicated and local content broadcasts are another weak link in the chain.”
“But,” Clarke continued – “if you could make it an open standard (both the watermarking and the detection) – it would make that data available to the entire ecosystem. The issue the industry is facing is that it has relied on Nielsen to do this kind of work for 60 years, but as new technologies and new options emerge, people are peeling back the onion layers and saying “hey, maybe this isn’t rocket science?”
“When you have this explosion of content, where media companies are putting their content on multiple platforms, it doesn’t all go through Nielsen’s watermarking step (especially mobile). That’s part of what’s taken so long to get multiplatform measurement. So, I think what the industry has to decide – esp. the media companies – is to make a commitment to and investment in using EIDR and Ad-ID to support them, and whether they’ll ask Nielsen to do it for them (which is possible), or whether we’ll have parallel open standard that sits alongside to the Nielsen watermark.”
That said, if (most of) the industry is behind the adoption of EIDR and Ad-ID metadata tags on a business level, why hasn’t it been done yet? The answer lies in the necessity of a mighty durable, open-standard content watermark that can survive all the various possible transformations and permutations a video file goes through on its way to your TV, tablet, phone, screen-of-choice. CIMM members (and the engineers behind EIDR and Ad-ID) realized they needed to make that process – that ‘data binding’ to the content – a standard in itself.
The logical place to turn was SMPTE – the Society of Motion Picture and Television Engineers. SMPTE put out an RFI with use cases and requirements in Dec. 2013, and by the end of January – the responses poured in. “1½ weeks later – they came to NYC to present the case,” says Chris Lennon, of Media Answers (a SMPTE standards veteran). “We learned a lot from the vendor community – six formal responses plus three to four informal responses. I was under no illusions this would be simple – and we knew that the requirements bar we were setting was high. The diversity of responses was great; the creativity of approaches to solve it, as well. We need this not just for audience measurement, but for targeted advertising, multiscreen apps, and new innovations.”
“When SMPTE came on board, it gave the [TAXI] effort a new life,” said David Kohl of Ernst & Young. “We needed a mechanism to keep track of the IDs that didn’t create yet another process – that’s ID Binding.” Kohl, who led E&Y’s consulting engagement on TAXI, remains deeply interested in the success of CIMM’s efforts.
Clarke agrees: “Our work with SMPTE takes the position that this should be an open standard, that the Ad-ID and EIDR data should be bound to the content in a sustainable/durable way – so they’re not lost. We need to track both ads and content. Nielsen does not watermark ads today (they can, but they do not), separate from the content (including time code, station, network, distributor code).”
The [TV INDUSTRY] is a supply chain that doesn’t see itself as a supply chain, but rather as a bunch of disconnected silos. – Harold Geller, Ad-ID
Building a solution to work for the entire TV ecosystem is, at this point – the logical next step – a “Go Big or Go Home” approach that can usher in a rising tide of revenue growth for all boats. “You do it for scale,” says Dale Rochon, Managing Director, Advanced Advertising Forum. “In micro-targeting ads and content across platforms – traffic systems and ad decision engines matter – the requirements have gotten more rigid, demanding shorter translation and processing times. MVPDs are moving toward the agility and speed of digital/online content/ad companies for data mediation – by competitive necessity. File-based workflows are imperative. The explosion of new video viewing and time-shifting devices defines the playing field.”
Not surprisingly, comScore and Nielsen are approaching this opportunity differently. Nielsen will introduce mobile measurement in the fall, by turning their watermark into an ID3 tag for digital. But it’s not clear how they’ll add the online viewing back into the cumulative linear, on-demand and mobile number. Clarke remains very “Missouri” (as in the “Show Me” state) about the potential net results from both companies: “The question remains, if it’s not exactly the same – where ‘the books don’t balance’ – it’s going to become some other metric. The industry doesn’t want to be confused, where they end up saying: “How do I buy and sell across these screens on divergent metrics?” – to an industry that has always relied on one metric – Nielsen’s – as the ‘gold standard.”
“Where digital video has been sold on a CPM/CPC basis and TV has been sold on GRPs – there’s coming an inflection point (at the end of 2014, if everything launches on time) as to how these offerings are packaged and sold as one media buy (with de-duping viewership across platforms). Nielsen’s been talking about selling only video, whereas comScore has been saying they’ll offer video, audio and text (which includes web content: pages, interactive content, infographics, visualizations, etc.) – as part of a total reach # – unduplicated across these five platforms (TV, desktop, smartphone, tablet and radio). We’ll see where these products are priced at, as it’s not cheap to stand up these kinds of services.”
With the future of media buying and selling moving towards automation (as the AAF’s Rochon opined) – there’s no way it can be done without a file-based workflow. Clark continued: “They’ve got people at ad agencies (sometimes interns) trying to go back and match what various distributors and ratings companies call different ads, because it’s not standardized. It’s such a waste of time on the backend, where they should be doing amazing things like data visualization and cross-platform data analysis – there are so many more interesting areas that could be addressed, but not if the naming conventions for ads continues to diverge across the ecosystem.”
WHEN STANDARDIZED, CONSISTENT AND PERSISTENT ASSET METADATA IS AVAILABLE – EVERYONE CAN DO WHAT THEY DO BEST – AND STOP REINVENTING THE WHEEL. – Harold Geller – Ad-ID
After the 2014 CIMM Summit meeting in April, Clarke feels there’s reason to be optimistic: “We’re getting all the MPM (multi-platform measurement) things we want this year. 2014 is the year. ESPN has it; the ten other CIMM partner companies will have it. John Muszynski, Chief Investment Officer of Spark (part of Starcom/SMG) had said in in 2006 “We’re going to a total video world, and we just need to get the measurement.” I told him at a recent conference that we started CIMM in 2009 with a roadmap to tackle this problem (with tools for ad campaign planning, measuring exposure, ROI, effectiveness). It was the ‘we’re not crazy – here’s what we want in a cross-platform measurement solution, and you probably want it, too’ moment.
We may be a ways from the “sabermetrics” era of TV content and ad effectiveness that universal IDs will usher in, but EIDR and Ad-ID, for their part – are working to make it happen sooner than later – using blended and sometimes alternating approaches of “carrots” and “sticks” to get the industry on-board. Falling into the “carrot” camp – the EIDR asset database, to be credible – has to be be as complete as possible. “In terms of metrics, we actually doubled our database in 2013 from ~300K to ~600K records,” says Don Dulchinos – executive director of EIDR. “Our growing list of digital member companies (Google, Microsoft, Netflix, et. al.) see the need for EIDR as obvious, but they have less cognition as to what the words ‘legacy infrastructure’ means from a veteran TV industry IT persons’ POV.
Falling into the “stick” camp – as of April 1st, 2014 – all new commercials produced for television, radio and digital platforms featuring SAG-AFTRA union members, must use Ad-ID as the sole standard commercial identifier. This mandate is the result of negotiations between SAG-AFTRA and the ANA-4A’s Joint Policy Committee (JPC) representing the advertising industry.
But beyond using Ad-ID to track which ads actors appear in so talent can get paid, the benefits of Ad-ID look much more ‘carrot’ like, says Harold Geller, Chief Growth Officer of Ad-ID. “In advertising, the metadata about an ad is the advertiser name, brand name, product name, commercial title, and commercial length. This is being shared today, but in a very analog manner. This information about ads gets retyped at least 20 times (and potentially spelled differently each time) in different systems from the time an advertiser tells its agency to produce an ad through to the time the ad completes its useful life and is archived for historical review.”
“Advertisers have been using ISCI (often pronounced “Iss-key”) codes for ads for years, but they’re inadequate for the multiplatform world we live in,” says Geller. “There’s no cloud registration, no canonical cross-reference. Ad-ID gives you that, and more. New tools from Adobe (e.g. – their XMP Manager) make it simple to inject Ad-ID metadata into an asset by pulling it from the cloud and auto-populating the required fields. Basically, when standardized, consistent and persistent asset metadata is available – everyone can do what they do best – and stop reinventing the wheel.”
In June ’74, the UPC barcode had its “Wrigley Moment” with that first barcode on the 10 pack of Juicy Fruit. In June 2010, the TV industry had their own with Canoe’s Wrigley’s Orbit Gum ITV RFI campaign. While mailing free packets of direct-response gum wasn’t enough to save the ITV portion of Canoe, the CPG and TV industries share a similar sticky ambition to automate and monetize their business. And the VOD portion of Canoe’s DAI (Dynamic Ad Insertion) tech is keenly interested in the success of CIMM’s industry efforts. (Canoe is now a member of EIDR).
Years in the making, CIMM’s patient and steady work is getting credible options on the table – so the industry will no longer be able to say they’re lacking competitive options for multi-platform content and ad audience measurement.
^ Kreth, Will. “MEASURE? NO MONEY, HONEY! – WHY MULTISCREEN ASSET ID STANDARDS WILL CHANGE EVERYTHING.” Will on Wednesdays. InteractiveTV Today, 17 June 2014. Retrieved 17 June 2014.