OTT, because of the IP-connected devices it serves and the “computer screen format” for viewership, uniquely enables broadcaster advertising to change from a generalised approach to a highly targeted approach. By design, this should increase the consumer’s interest level in the adverts they see, and therefore yield better results for advertisers.

The promise is clear, but can we take full advantage of it today?

The Vision

To answer this question, let’s define the ideal scenario from an advertiser’s perspective as follows:

1. Ad inventory is clearly defined in both Live and VOD services.
2. Ads can be aligned very precisely with the most likely customers.
3. Ads are always delivered, on-time and in the right position.
4. Reporting and analysis of Ad delivery, consumption and effectiveness is fast and straightforward.

The vision of addressable advertising in OTT services is ambitious. But it must be delivered in order to successfully transition the broadcast industry from its current advertising, subscription and licensing models to the next generation of commercial models that create win-win-win situations for broadcasters, advertisers and consumers.

While monetization of content is clearly a fundamentally important subject for any media business, this article focuses on the technical nature of advertising to an OTT audience. Not just for today’s OTT audiences but for the significantly larger audiences expected in future.

It should also be noted that this article focuses on a Broadcaster/Content Provider managing and delivering ads within their own D2C OTT service. We appreciate that advertisers work with broadcasters, content providers and MSOs to sell ads but this is not the focus of this article.

Delivering The Vision

In relation to the ideal scenario above, the industry is in a semi-advanced position with significant technical capability. The biggest challenges are operational. As OTT services scale-up and the advertising payload shifts from traditional over the air broadcasting to addressable advertising in OTT, the industry needs to be prepared to address several challenges to achieve OTT’s full monetization potential. These challenges relate to the definition of ad inventory, the delivery of ads, and reporting of ad consumption.

Advertising Inventory

We can define advertising inventory as the set of advertising spots presented by the OTT operator to the advertiser marketplace. How this inventory is defined is a critical subject because OTT video presents particular constraints and opportunities.

OTT advertising is a hybrid position between pure digital advertising (e.g. Google Ads) and pure linear broadcasting advertising (i.e. traditional playout-level ad insertion). While the ads can be inserted into OTT linear streams, they can also be inserted and updated in VOD assets, and they can be targeted, scheduled and measured in the same way as digital advertising. Given the possibility to address the individual viewer more personally than linear broadcasting previously allowed, what does this mean for how ad spots are defined in OTT?

Ad spots should be characterized by a long list of parameters about the content being streamed, the consumer who is watching it, and the time that is available for ads to be played. This combination of factors creates the most valuable ad spot. Once a spot and an ad are paired, the ad can be inserted on the client-side or server-side. But the challenge of creating an optimal pairing is largely based on how much data the OTT operator holds about the consumer. OTT operators are constrained by privacy laws, the willingness of consumers to share personal data, and even the administrative overhead of managing the data. Because of this, ad spots are generally characterized by content-related data – in other words the program’s subject. This means that OTT advertising looks a lot like linear broadcasting advertising, which is very different from the Google Ads and Facebook Ads experience.

That said, even without a lot of personal data, OTT already provides some very specific advantages for advertisers over linear broadcasting. One example is that it is easy to know where a customer is based and what language they use. Geo parameters can be collected from IP addresses and browser settings, so if a streaming request is made from Italy and from an Italian language browser, it is safe to assume that the ads can be served in Italian and in an Italian context.

OTT operators today can segment their ad inventory based on parameters collected from browsers, apps and devices. By creating these more focused sets of ad inventory the advertising opportunity opens up to more advertisers with smaller budgets.

Technically, this more granular segmentation does not present a challenge. Today’s technology is able to handle this targeting and placement of ads. To tailor this further using more personal data relies on a set of data gathering and data privacy rules to be in place, potentially relying on the consumer to willingly provide the data and to trust it will be used appropriately. From a consumer perspective this paradigm will either be rejected due to privacy/trust issues or will be accepted due to their own content consumption goals. OTT operators need to establish how to request and manage this data in order to create the most targeted ad inventory. Those that gain consumers’ trust are likely to have the most valuable ad inventory which can be used to leverage a favourable commercial position in the advertising marketplace.

Placement Algorithms & Ad Delivery

When the rules are defined, the pairing of the ad spot and the ad is a straightforward task that does not cause a technical burden on ad placement decisioning systems. It simply requires sufficient compute power to process small amounts of data quickly.

Operationally, various considerations must be addressed in relation to delivering ads to comply with legal requirements (e.g. adult-only viewing or country-specific regulations), commercial requirements (e.g. brand adjacency), and technical requirements (e.g. devices that support DASH, HLS or other). Metadata must be correctly defined and rules engines must be set-up to deliver the correct ads. Systems must of course support granular details, but this is not highly technical.

The real technical challenge is to flawlessly execute the commands to deliver potentially many different videos to thousands or millions of consumers. There are three overarching technical considerations:

1.  Latency. Delivering ads quickly requires low latency. Particularly in live OTT services where individual addressable ads create workload challenges when compared to streaming video (see below). Prefetching identifies in advance which ads will be delivered to each consumer based on the business rules. Prestaging those ads in a CDN or ad server ensures that the selected ads can be delivered at the required low latency. And as ads become more targeted and OTT audiences become larger then bandwidth, processing power and proactive decision-making algorithms will need to keep pace to maintain low latency at scale.

2.  Timing. Ads must fit the ad breaks. This is straightforward for VOD and linear streams that have pre-defined durations set by OTT operator business rules. For Live events and “join-in-progress” viewers this is more challenging. When ad breaks are unpredictable, real-time decision-making is required. This places pressure on both ad placement engines as well as streaming latency. Today the approach is to simplify the ads that can be inserted into these ad breaks and only serve pre-staged content.

3. Customer Experience. The ad viewing experience should be seamlessly integrated with the program content. Server-side ad insertion performs best for dynamic ad insertion because the video segments and ads can be aligned at the edge cache before streaming to the player. Alternatively, for client-side ad insertion, an ad player and a video player can run simultaneously on the consumer device, overlapping with each other to make the viewer experience smooth as the content changes. Server-side requires ads to already be available at the edge cache. Client-side requires ads to be available on the ad server.

Technically, as OTT services scale-up the CDN-based delivery mechanisms must also scale. For the biggest OTT operators serving content to millions of consumers, the most efficient way to flawlessly deliver personalized ads is to create a private CDN path for Ads. This will guarantee bandwidth and processing capacity to achieve lowest possible ad delivery latency. 

Ad Consumption And Reporting

OTT Operators have full control over whether an ad must be played or not. Working with advertisers they can deliver the appropriate type of experience for their viewers. Ad-supported OTT services, which continue to grow in popularity for consumers desiring cost-effective content, have a full range of options including forcing pre-rolls, not allowing fast-forwarding by disabling trickplay during ad breaks, enabling ads to be skipped after a specified number of seconds, and more.

At the moment, connected TVs present a different challenge because ad viewability standards have not yet been defined.  In “digital” formats (i.e. smartphones, tablets, PCs) the ad viewability standard means that the ad player was in 50% of the screen and remained there for at least 2 seconds. Once this is defined for connected TVs it will be possible to establish performance KPIs for advertisers to depend on.

Advertisers measure the success of their ad campaigns with various KPIs such as viewing completions, ad click-throughs, qualitative ad recall surveys, store sales in a related geography, and more. ROI is a tough metric to use because of the difficulty in directly tracing the viewing of an ad through to the sale of a product or service, or any other specific action that has a meaningful impact on the advertiser’s business. Today’s approach to reporting on ad effectiveness therefore relates primarily to delivery, consumption and immediate action. Low percentages of action are the norm – e.g. a low % of viewers clicked through from the ad – which is similar to most types of advertising engagement.

Conclusion

Improving consumer engagement is why addressable advertising in OTT is so important. Increasing marketing ROI by targeting buyers more precisely and improving the “ad to action” conversion rate is a marketeer’s mission. OTT as a platform directly supports this goal but needs to be leveraged to reach its potential. The technology is largely in place already although as OTT services move towards prime-time TV scale, it will become increasingly important to assure the flawless delivery of both program content and ad content which can be achieved through the right choice of technology platform. The broader challenge is to master new operational disciplines and overcome privacy/trust challenges so that fully personalized ads can be delivered.

As ad-supported OTT services expand, the win-win for advertisers and consumers is there for the taking. We simply need to deliver the right ads to the right people at the right time. The onus is on the OTT operators to capture their customer data, manage it well, and implement the tools that allow the best possible ad placement decision-making. Once that is in place, advertisers will have access to some of the best possible advertising inventory.