As little as nine months ago, trade publications were still filled with articles about the ‘death’ of TV advertising.
Since then, however, both Netflix and Disney+ have launched ad supported tiers. Amazon Prime Video is hot on their heels, not to mention the flood of niche-focussed FAST (Free Ad-supported Streaming Television) channels being launched.
Customised and personalised ad delivery
The technology at the heart of these new ad-supported services is known as Dynamic Ad Insertion (DAI). DAI is used to insert targeted advertisements into media content in real-time to monetise content and provide a more personalised advertising experience for viewers.
It allows advertisers to use data and targeting algorithms to select relevant advertisements for specific viewers based on factors like demographics, location, browsing history, and user preferences. This enables highly targeted and relevant ads to be shown to individual viewers, improving the user experience, and increasing the likelihood of ad engagement.
Audiences have already become accustomed to internet live streaming platforms, such as YouTube and Twitch, using DAI to dynamically insert ads based on their profile and behaviour.
The American department store pioneer, John Wanamaker, famously stated, ”Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.” The aim of DAI is to push that ratio over to the advertiser and audiences’ favour.
Dynamic Ad Insertion today
DAI is still a relatively new technology, and a relatively new approach for video. As such, it is still not fully adopted across the industry.
“It differs from market to market,” says Johan Bolin, Chief Business Officer at Agile Content. “In some markets, such as Nordics and the UK, DAI has been well deployed for years. In less mature markets, there is still a lot of generalised delivery. Another aspect in this [are] privacy directives such as GDPR that have limited or partly rolled back the personalisation level. But in those cases, ads are still dynamically inserted.”
Valentijn Siebrands, Solutions Architect at M2A Media, sees a variety of approaches being taken depending on the value of the content being delivered. “What we see for live sports and low tier events - the football match around the corner - is they can be monetised using FAST channels, because people do not care too much about what is being shown in terms of ads,” he says.
“But as soon as you have expensive rights, you want to be more in control of advertising, especially the experience. We see our customers really thinking about advertising and how to monetise it, and then coming back to us and saying: ‘for these premium rights, we want to be in full control… This is a Bundesliga match, going into Germany and it needs to display these ads for Germany, and [these] for Italy.’”
Read more Imagine publishes AdTech insights report
Due in part to the prevalence of online advertising, the past decade has seen a significant decline in what advertisers are willing to pay per ad impression. However, DAI is helping to turn that around. Johan Bolin has seen a willingness on the part of advertisers to pay a premium for targeted ads.
“The performance is generally better and hence the premium,” he explains. “That said, there is usually a part of the inventory that is sold on a more traditional basis and that could very well be the most premium inventory, for example in bigger live events.”
This is largely because of the increased Return on Investment seen thanks to the adoption of DAI technology. “I’d say DAI has introduced very efficient ways to increase the value of much of the [ad] inventory, but not necessarily all,” he says. “That’s why [DAI] is often used in combination. This combination, though, should offer higher ROI.”
The audience response to DAI
Bolin believes that getting the user experience ‘just right’ is key to making DAI successful. “I would expect that when you are watching a Champions League final, the last thing you’re going to do is click on a BMW ad because you want to watch the game,” he says.
“But the branding and the experience is really something you will remember. So, if you do advertising with premium sports or with sports in general, you relate the brand to your favourite sport or the thing you just watched.”
Advertisers are also having to walk a fine line between delivering targeted ads and seeming as if they are spying on their viewers. Bolin discusses some of the ways that advertisers are addressing this issue.
“A few years ago, when dynamic ad insertion was the new hot thing,” he says, “there were a lot of studies and reports arguing that viewers are more happy with the overall user experience if the ads are relevant. But what we have learned is, if you extrapolate that too far, there is a sweet spot. It should be relevant, but not personalised.
“In the beginning there were cases that got some push back since some advertisements were too well targeted. But learnings of how not to be too aggressive on targeting - in combination with privacy regulations - have resulted in a more balanced level of targeting. One way to use DAI is simply to keep track of how many impressions of one ad have been delivered. This doesn’t introduce any ‘targeting’ but reduces overhead.
“Then there are several levels of targeting, based on things like addresses, date and time of day, that don’t include any real personalisation. [The] next level would be to start separating gender and possibly age brackets… What is generally considered most compromising is if you start to go cross service/cross platform, like when you get Facebook ads about something you Googled two hours ago. This is rarely the case in streaming TV though.”
Valentijn Siebrands emphasises the way that privacy regulations like the EU’s GDPR (General Data Protection Regulation) play a role. “You have various legal obligations,” he says. “And it’s that driver that helps us to provide better content. It’s usually much simpler to make five different versions for Germany with five different playlists, and when people log in, [to] do segmentation based on subscription data, which has nothing to do with personalised data being exchanged.
“The major streaming services already know what type of customer you are, therefore, they can already segment you. This is slightly different, because it’s not really based on huge data that has been collected by other nodes like Facebook and Google. It’s leaner, meaner, and easier to manage.”
The Future of DAI
DAI is a technology and approach that is still in its infancy, and its potential has only just begun to be explored.
FAST is the area that is currently making the most use of DAI. “It’s a perfect model for this,” says Johan Bolin. “In general, revenues from FAST are very small compared to other advertising services such as YouTube, and are expected to remain relatively small in the next few years.
“One should also keep in mind that there are multiple reasons to use FAST. A TV manufacturer might want it to increase the value of its TV sets and to generate loyalty. A content owner uses it to get reach and stimulate interest in its content, leading to conversion to subscriptions for their own streaming service. And in markets used to FTA TV and relatively low penetration of Pay TV historically, DAI might be a tool to get reach that later can be converted to subscriptions.
“So FAST is not necessarily something that is expected to deliver a lot of revenue directly. The value can be indirect as well.”
Bolin also believes that the industry has only scratched the surface of what can be done with correctly targeted advertising and its potential use in a multi-device, networked environment.
“I would like the TV industry and the advertising industry to spend more time trying to figure out how we can evolve the concept of advertising,” he says. “Nowadays, on a connected device, there are so many more opportunities for how you do advertising as a part of a TV experience. I see plenty of untapped potential in using the fact that, on the internet, you have an interactive channel where you can use multiple devices more cleverly than we have in the past.”
He adds: “If you do this the right way, [it’s] a win/win - you could drastically improve the value of the ad inventory. And by doing that, you could reduce the ad load. Maybe you could cut the number of ads per hour in half If you instead increase the value of those ads.”
Brolin suggests that, if we want to push the boundaries of dynamic ad insertion technology, we should ask a different question, namely: ‘What kind of new advertising concepts and products could we develop for streaming services that take advantage of the capabilities of the internet?’
He says: “On the internet, the distance between a TV experience and an ecommerce store is just one click away. So, we have a much better starting point in building advertisement-based models. [You can go] directly from an ad impression to a transaction. There are just so many opportunities and maybe we are spending too much of our engineering cycles trying to reinvent what we have done in the past and too few trying to figure out what we should do instead.”