Ad Tech Raises the Stakes as Streaming and Broadcast Viewerships Converge
While broadcast advertising may have the larger budgets, streaming has very innovative solutions that viewers continue to lean into. One research study reports that ad-supported streaming is expected to reach 58.2% of US viewers, which means that while viewers may have a love-hate relationship with advertising, they are still viewing ads.
Connected TV (CTV) ad spend in 2022 is on target to double its 2020 numbers. That’s quite a leap, so CTV must be doing something right. It’s expected to grow by 39% year over year to $21.2 billion, according to the IAB’s 2021 “Video Ad Spend and 2022 Outlook” report.
Nielsen reports that viewers spent about one-third of their viewing time watching streaming in June 2022, compared to 22.4% for broadcast, and 35.1% for cable, both of which have been slipping for some time now. Streaming has reached the highest monthly share yet since Nielsen started measuring and breaking down viewership numbers in this way in 2021. Likely, this is due to CTV’s growing popularity. Who doesn’t want to view their content on the largest screen in the house?
Broadcasters are watching digital and trying to bring the best of digital ad tech into their “broadcast” environment–even if that broadcast environment is on CTV. While it is becoming more and more confusing to differentiate between broadcast and streaming, one thing that’s clear is that digital—and the ad tech supporting it—are responsible for the success of one of streaming’s standard business models.
Video buyers surveyed for the IAB report said CTV allows them to use data not available within linear TV buys, including first-party brand data (65%), location data (61%) and shopping data (50%). IAB also found that 57% of buyers felt CTV was more effective than linear TV at delivering website and sales actions.
Although CTV will account for 36% of total time spent with linear TV and CTV combined in 2022, though the dollars currently allocated to CTV are not proportionate to this viewership. Only 18% of video ad dollars are allocated to CTV, according to research published by eMarketer in January 2022.
With this in mind, here are a few ways to turn up the volume and hopefully bring the CTV ad spend more in line with its viewership share:
- Targeting without cookies
- Selling the right slot
- Doing addressable on broadcast.
Beyond Personal Data
Google search ads have proven a very profitable approach to using contextual information for targeting. However, when it comes to video, this is unstructured data and that makes media scraping infinitely harder without the right tools.
“You can get contextual information coming from different pieces of the puzzle—the application, the device, the CDN, the manifest manipulator, and the player itself,” says Mathias Guille, VP Cloud Platform, Broadpeak. “Right now, there is not really a standardized way to carry that information from any point of time back to the manifest, back to the service.
While we may be waiting a while for an open standard, Iris.TV has come up with one approach. IRIS_ID was created in 2019 by Iris.TV to create standardized, contextual data which advertisers could use to do personalized targeting. Customers using IRIS_ID obtain permission to scan a distributor’s content. Iris data partners like GumGum, Oracle, Comscore, Kerv, Reticle, and Silverbullet 4D, can use their computer vision technology to index the videos.
“This is the point where the segments are created,” says Field Garthwaite, Co-Founder and CEO, Iris.TV. “Some specialize in accurately identifying content categories, some brand safety. Some identify brands and celebrities, and others do emotion and sentiment.”
The idea behind IRIS_ID is that instead of needing to target on consumer data (which mostly means getting viewers to opt in to provide access to their personal information), it’s much safer and easier to target based on contextual details from within whatever content a viewer is watching. These contextual signals get stored in the IRIS_ID.
The IRIS_ID is passed into the bid stream, where it can be decoded by Iris-enabled partners. The segments mentioned above can be targeted by advertisers, who pay a fee if those segments result in transactions.
“Our identifier is for content only. It has nothing to do with personal identifiable data,” says Garthwaite. “We pipe all of that data into the advertising ecosystem through SSPs and DSPs so that it can be used for direct, private marketplace or for open auction.” Translation, if you don’t regularly speak ad tech: The contextual data goes to both the buyers and sellers, who can then transact. This takes on one of three methods:
- The direct, popular fill-in-the-blank popular brand, where buyers purchase from a popular streaming service.
- A group of select popular brands is invited to a private auction to buy from a popular streaming service.
- Whoever is in the right place at the right time and spends the most wins.
The result, according to Garthwaite, is higher revenue yield for publishers and guaranteed brand safety because advertisers know the exact content categories they are running their ads on.
“What we’ve done is created a standardized technology for storing all this data, creating a standardized data set for how to actually make it available,” he explains. “By working with all these companies who have standardized ways of creating the contextual data, we’re creating unified signals.” This data travels with the content, as long as the content is being shown in an environment which supports the IRIS_ID.
This tech is currently available for VOD only, but Iris hopes to have a live version in the market in 2023. Currently, Amagi is using this for their FAST channels to offer more targeted inventory. Previously, targeting was just to genre.
AI and Contextual Advertising
GumGum, one of Iris’ data partners, has an AI technology called Verity which analyzes video content to get contextual categorizations based on the IAB metadata taxonomy which has hundreds of categories, up to four levels deep. “We pick the best categories from that taxonomy with (the highest) confidence level and return that back to the client,” says GumGum CTO Ken Weiner.
Verity has been in use for a decade and has been sold to customers as a discrete product for the past two years. “We have an MRC accreditation that’s focused on content level analysis and we’ree the only company other than YouTube to have it,” Weiner says. The Media Ratings Council (MRC) measurements bring valid, relaible and effective measurements to media measurements.
GumGum’s contextual analysis is used for brand safety so ads aren’t placed in settings which aren’t congruent with the product’s values. Some of their customers use Verity as a way to normalize how all their editors tag content, so there’s consistency in their metadata tagging. “Most of the publishers use it for their own direct ad sales efforts,” Weiner says. First, they take the data and feed it into their ad server, and then, when their team sells campaigns, they can do targeting and brand safety using this data.
“You can move away from saying, ‘I need to find a woman 18 to 24 in a zip code, with so many kids in the household,’” says James Smith, EVP, Global Ads Sales & Programmatic, Amagi. With this technology, he continues, “you can look at somebody who has spent so many minutes watching a certain type of content. Or a certain number of scenes in certain types of shows and movies. It puts us into a situation where you’re actually talking about something that’s better from a value proposition and delivery perspective,” he continues. “When you can get to the scene level, you actually know their emotional state or have a much deeper context of what is interesting to the person. We can increase the value of those ads based on better targeting and hitting the right person.”
Whose Ad Slot is This?
Moving away from targeting, one way digital has influenced broadcast is within the intricacies of advertising inventory sharing. Comcast has a product called Linear Rights Metadata Management (LRM), which which helps to ensure that addressable advertising traffic goes the right way. LRM conveys linear rights metadata to distribution partners to ensure that programming contracts between programmers and distribution partners are followed.
“We’re providing a layer of standardized metadata to communicate and manage between parties which ad slots are addressable. It can be segmented per distribution point or per MVPD,” says Gregg Brown, Senior Director, Product Management, Comcast Technology Solutions.
“In the traditional cable TV space, the operator has two minutes of inventory and the programmer or content provider has 14 minutes of inventory,” Brown explains. “It’s about identifying—on an operator basis and also on an ad-pod or ad-slot basis—which ads are addressable or not.”
The content provider would mark their content with SCTE 224 markers to identify slots that are addressable. In addition, the SCTE 35 marker in the content stream triggers the ad break, and the decisioning built into LRM and SCTE 224 ensures that an addressable ad is used.
“We’re communicating addressability as well as all the ad rules down to the ad server,” says Brown. Aside from the rules identifying which ad slot belongs to whom, these rules can also include guidelines like not running a Pepsi ad after a Coke ad. “Linear addressability has not existed in the past.” This is the first time that it’s been made manageable, so that the content provider can specifically initiate addressability.
“For the content providers, their inventory is worth more because it’s addressable,” says Brown. “A huge part of this whole drive to linear addressability is the blending of traditional broadcasts and traditional cable with a digital mindset.” Moreover, enabling this linear addressability allows advertisers to reach both the digital and broadcast segments, placing them almost on the same wavelength.
Broadcast has been letting viewers skip ads for years, but this is less of a feature than a bug. To skip an ad, you just go to another channel. Maybe you come back to the first channel. Maybe you don’t. Of course, you can do this on streaming too, but it’s a bit more complicated unless you’re in a FAST environment. Broadpeak has an appealing and somewhat underused capability which can be used in multiple environments (as opposed to only on Roku, for example).
“We’re looking to enable use cases such as what YouTube is doing, letting you skip an ad right after a few seconds,” says Broadpeak’s Guille. “We let them talk with the player and tell them, ‘Okay, after three seconds, you can skip the ads.’ You can jump inside the manifest to go back to your original source. There’s no standard way of doing this. Our SmartLib APK is going to tell the app, ‘There's an ad right now, so you can display a small button, which will let you skip the ad and jump to the end of the SCTE marker.’” He adds that SmartLib is used mostly with longtail content or something which has another monetization track like sponsorship. He adds, “We do this in Israel for a company called Partner TV.”
Addressable Ad Replacement
Serverside.ai is taking another approach to blurring the lines between broadcast and streaming through ad tech, using their streaming SSAI technology to bring addressablem “Full Ad Break Replacement” to broadcast for Hybrid Broadcast Broadband TV (HbbTV) in Europe. They are also working on ATSC 3.0 as well, which requires the same proceedings, although the underlying technology stack differs.
“On a TV set’s broadcast, you usually have no back channel. You have no IP,” says Serverside.ai Co-Founder and Managing Director Leander Carell. “But with HbbTV, you have the capability to do signalling of additional IP information in your broadcast stream.”
With HbbTV, Serverside.ai is able to target viewers based on whatever opt-in information is provided. “This information can be used in order to substitute a broadcast ad with a personalized ad to make this addressable TV,” Carell explains. “We render everything that we want to show to the viewer in the back end, so that we have a full, personalized ad break.”
This client-side ad substitution provides a personalized experience within a true TV broadcast, one-to many broadcast environment. “We hand over to the client site this full ad break to substitute” for an existing broadcast advertisement, Carell says. “It’s not stitching in this case because we don’t stop the broadcast. The broadcast goes on, but the overlay client side personalizes the ad break.
“The Ad-Substitute is an ‘overlay,’ and technically it’s another video player loaded—either an AV Object or HTML5 player, depending on the HbbTV Version),” Carell continues. “Our SSAI back end is processing the ad-break stitching and then delivering a unique full ad break to the device where the client-side ad substitution takes place. That’s why we call it a hybrid ad-insertion solution.”
Carell says that Serverside.ai’s current clients include Discovery, who is implementing HbbTV in Poland and Italy. With their first test cases in the market, it will soon be apparent if broadcast advertising can really be targeted. As in the other two examples, the result is increased ad revenue.
Merging Digital and Broadcast Ad Tech
According to IAB research, 66% of linear TV/digital video buyers now have a single planning team for the two channels, and another 25% expect to have one in the future. So, since more the 90% of buyers expect to have a single team, ad tech needs to support this. Right now there is still very much a division between what you can do in a digital environment vs. what’s possible in a broadcast one.
Here’s a list of things to work on, and in which arena problems persist:
- Measurements (both)
- More diverse ad formats, like interactive ads (broadcast)
- How adverting is bought and sold (both)
- Targeting based on first-party data or contextual data (both)
- Ad loads (broadcast)
- Frequency capping (digital)
- QoE (digital)
- Fraud (digital)
Ad tech vendors will be pivotal in the merging of the two different delivery methods. In digital, we’ve gone from discussing the merits of SSAI to focusing on personalization/addressable (same concept, different wording) opportunities. In digital, the ability to target has existed for a while. Now as broadcast is starting to focus on using addressability to deliver more targeted ad inventory, there is the chance for ad tech to really influence how advertising is done.
“The ad industry doesn’t move forward quickly,” says Joe Friend, Media & Entertainment Consultant, Friend Advising. “Their business is to take money, so the VAST, VMAP, HTML, ad stack waterfall spaghetti logic that flows through the systems needs to work properly despite all the handoffs. We lose money when it breaks and you can never get that opportunity back.”
Many of the solutions here have not relied not on open standards. Rather, the companies behind them have gone ahead and come up with their own tech, which solves some very common problems. How do we target, measure, and introduce the right ads, provide assurances to brands they are in the right area, or ensure that the right thing appears in the right environment? The biggest issue here is that to get these services, you must use a vendor’s specific products.
Digital has identified the roadmap and heard the complaints over the years of what works and what doesn’t. Broadcast is now in a position to benefit from the growing pains digital has gone thru to provide a better advertising experience. The upshot is, while advertising may not have moved quickly in the past, the competition from all sides means ad tech, media, and brands really need to embrace an agile approach—or at least as agile an approach as possible when there’s only one shot at filling each ad slot.
Why is CTV one of digital advertising's fastest-growing channels, and where is CTV advertising heading? The channel is likely to keep growing fast thanks to larger audiences and viewer demographics, rich storytelling possibilities in a medium poised to capture consumer attention, and evolving measurement capabilities that will make CTV a go-to for both brand and performance marketing.
FAST has been the big story of the last two years, but there's tension growing between streaming services and TV/OS manufacturers over who owns the viewer and how to get those viewers to the content.
Advertising technology is scary. Here are a few solutions—for ad fill, multicast distribution, monitoring, contextual ad insertion, analytics, demand-side, and supply side—to help you make it out alive
The OTT video advertising world has reached a tipping point, as consumers turn to AVOD services and ad tech gets more sophisticated
As ad-supported viewing grows, the industry grapples with three key challenges: privacy, workflow fragmentation, and scale
Here's how five industry experts are finding their way through the complicated world of advertising technology.
Companies and Suppliers Mentioned