By Matt Smith, Principal Evangelist for Media, Brightcove

There is no denying the growth in online video. It is an unstoppable force of nature and permeates every aspect of our daily lives. Whether you are eight or eighty, it’s likely you consume some form of video in the course of a day. As viewing shifts from traditional screens in the home to devices in our hands and elsewhere, publishers, networks, and broadcasters are actively pursuing monetization strategies for their content. The idea here isn’t new either, content drives revenue through advertising in the way it has traditionally done in the past with television. Though the concept holds promise, it isn’t without some early challenges. Just as traditional television evolved, so must online video. You didn’t expect the money to roll in with no effort nor challenges, right?

All kidding aside, television and cable adjusted to technology and improved upon approaches to generating revenue through advertising. Online video is no exception, and though many organizations are seeing profit from content, the overall process and the technology therein is always evolving. Rest assured, these initiatives and evolutions are well worth it. Revenues from digital video hit a record $9.1 billion in 2016, a 53 percent year-over-year rise from $5.9 billion in 2015, according to the Internet Advertising Bureau’s annual report. On mobile devices, video revenue skyrocketed, more than doubling up 145 percent year-over-year to nearly $4.2 billion. At Brightcove, we see these trend spikes too, as more than 400 million ads are delivered by our players, helping our customers realize nearly $5 million in revenue each month.

That said, the path to online video monetization isn’t without its curves and challenges. Here are a few of the present hurdles faced by dynamic ad insertion (DAI) workflows today.

Ad blockers

No technology generates more hot sports opinions these days than ad blocking. Millions of people worldwide have installed these browser plugins to fend off what they determine to be intrusive advertising experiences. There is some merit to this line of thinking. For a long time, digital audiences have been forced to deal with ads that did all they could to be ‘in your face’, popping up and taking over the browsing experience. A growing dialog amongst thought leaders in technology and advertising identified that improved ad experiences that more effectively target the user or viewer, the more likely they are to allow and even interact with advertising.    

From a technical perspective, ad blockers ‘see’ HTTP requests from the browser and prevent that payload from being delivered. When this process is applied to video, it is incredibly disruptive to the viewing experience. Not only is the ad not played, but the entire video viewing experience is often completely stopped. The remedy for this is to shift the ad request and delivery from the player to the cloud. This process is referred to as server side ad insertion (SSAI), which shifts the communication between the player, the ad network and the delivery infrastructure to the cloud. Here, what the browser and player ‘see’ is one contiguous stream, with no delineation between ads and video. Given that, ad blockers cannot interrupt the experience. Furthermore, this is the only way to deliver monetizable video streams to platforms like Apple TV and Roku.  

Parsing performance

Let’s face it, humans are data driven and like to measure performance. Whether you’re analyzing gas mileage on the trip to drop the kids off at summer camp, or how long it takes you to mow the yard, most of us consistently use data to determine how well we did or didn’t do on a particular job. The performance of online video is no exception. As video becomes an integral part of our daily lives, and more viewers grow up in a ‘mobile first’ world, the time it takes for us to get to the first frame of video is critical. The longer it takes for videos to start, the more likely we are to form negative opinions of that experience. You may not realize it, but as that video is starting up on your computer, mobile device or set top device, an auction is taking place. Before that video loads, ad networks are negotiating with advertisers who want their ad to play on your screen, provided you meet the established criteria. If this first negotiation doesn’t yield an ad, a second and third negotiation may take place. As you can imagine, this delay makes for a negative viewing experience. Once the ad is played, said advertiser wants to know a few things about the experience. Did the ad play? Was the entire ad watched? If not, when did the viewer close the video player? These client side scripts, ad managers and many other metrics being captured can often lead to latency and poor video experiences that no one wants.

To remedy these issues, a few things can be focused on to optimize viewing. First, publishers need to put their best content foot forward. Great content finds audiences, which yield ad spend. Sounds simple, but there is a lot of bad video content out there. This content participates in the same auctions, yet it doesn’t carry the same viewing quality as ‘good’ content. As an industry, we need to push invalid and poor inventory out of the system, doubling down on higher quality legitimate content. Conversely, it’s also time to do the same for video advertising latency. We need to get the bad latency actors out of the system and improve upon monetization strategies, so that there is no difference in the speed of watching traditional television or online video. This is something I call ‘better than broadcast’ – online video should be as good, if not better, than broadcast when it comes to the viewing experience. This applies to advertising too. As an industry, technologists and advertisers must seek and ship better methodologies for online video monetization. We must fully reduce dependency on Flash and shift demand from VPAID Flash to VAST + video or VPAID JavaScript. With the auction process, velocity can increase with the use of programmatic real time ad bidding and the aforementioned SSAI. In short, the cloud can help all these processes happen faster and more independently than they are now – tied to the Play button when the user initiates a video play.

Mangling manifest manipulation

The video technology that helps deliver viewing experiences to screens and devices of all shapes and sizes is able to shift and adapt to changing network conditions. The overarching term for this technology is adaptive bitrate, or ABR. Though Apple’s HTTP Live Streaming (HLS) spec is largely the most prevalent format, there are other ‘flavors’ of ABR video technology – MPEG-DASH, Microsoft’s Smooth Streaming, Adobe’s HDS.  

These different video formats contain two primary elements:

• Fragments: These are the actual pieces, or ‘chunks’ of video content, generally in four second increments.
• The manifest: I like to call this a treasure map, as it helps point a player to the video chunks and the order in which they are to be assembled and played. It is a text-based or XML file with specific variants depending on format.

As discussed earlier, SSAI is the most effective way to deliver conditioned, ad-ready streams to any screen. By nature, it also leverages a process called manifest manipulation, a fairly simple concept.

By changing the manifest, one or more video elements can be modified, such as: Inserting ads, Inserting alternate content (for blackouts for example), limiting or expanding the number of bit rates available, or setting initial bit rates. Though the concept is simple, manifest manipulation can often be very tricky. For a variety of reasons, most of these ABR protocols aren’t very tolerant and don’t like having the manifest rewritten. Specifically with advertising payload, if there is variance in the granular encoding settings between the ad payload and the program content and the manifest is changed, the player will most often break playback along those boundaries. Put another way, this is the equivalent of forcing the engine in your vehicle to switch from unleaded gasoline to diesel on the fly. It likely won’t end well. Though the process of manifest manipulation provides some flexibility as outlined above, it can be detrimental to the monetization aspect of an online video strategy. Fortunately, the remedy for these hurdles is fairly straightforward. First, establish standards between your programming and the advertising that plays in and around it. Specifically, ensure that the compression and encoding settings between the two are identical. This can be a challenge, especially with the ad payload. Still, if you design a process that properly encodes ad units as they are placed, just prior to playout. This process ensures that playback is uninterrupted, visual quality is maintained and monetization intact by processing all content in a ‘just in time’ process.  

Despite these near term hurdles, many media and entertainment organizations are very bullish on their online video strategies. Just like its big brother, traditional television evolved, so will online video. In fact, many would argue that the pace of change and process of improvement is faster for online video than television ever imagined being. Given these numbers, that is encouraging and even exciting.