Commentary

TV Automation is Already Here -- Retire The Fax Jokes

by Walt Horstman , December 12, 2014

This classic Audience Buying Insider was first published earlier this year.

Programmatic TV is the combination of audience data and automated execution to make TV planning, buying, optimizing and reporting smarter and more effective. Over the past few years, we have built a wealth of audience data from multiple online and offline sources that we now can use in TV. It seems every week there is more news about matching this data set with that one to discover new insights into what an audience is watching. We’re getting incredibly smart on using data to truly measure the way TV influences consumer behavior with closed-loop analyses, sales lift metrics and understanding of how frequency actually helps or hurts.

Then comes the invariable “BUT” -- and we hear the sound of the programmatic TV gears stripping. “But programmatic TV needs automation and TV is not automated,” or, “TV won’t be automated until at least 2020.” Further still: “Can you believe they still use fax machines in TV?”

So here is the truth: Very large swaths of the TV ecosystem have already been automated. Audience data and TV automation create a virtuous circle; the more you have of one, the more you need the other.

Automating TV is very hard, because TV systems are very different from digital. Digital is a real-time, decision on-demand medium. TV is a future, planned medium. In digital, ad servers collect information about a visitor who comes to a site, make complex decisions about which ad to serve, call a database with the right ad, and then display the ad -- all within a second or two, while a page is loading.

In TV, a program feed is prepared and scheduled weeks in advance of its airing. The TV ad inventory is managed by a TV traffic system that crunches through all the future available ad units and the needs of campaigns that will air over the coming days and weeks. The traffic system matches all those campaign needs and constraints together in the most optimal way possible. There are so many dimensions the traffic system is trying to optimize simultaneously, it can take hours each day just to process the final TV ad schedule.

A TV traffic system is a lot like UPS’ delivery route optimizer. There are so many different route options to get packages to the correct destinations each day that finding the absolute right answer is not feasible, even with today’s fastest computers. The UPS optimizer crunches as many options as it can until it reaches the best solution available.

TV traffic systems are the core of the TV ad business, so absolutely mission-critical that many TV companies actually build their own proprietary systems. If you ever want to play a cruel joke on a head of TV operations, tell him you heard his boss wants to forklift the traffic system -- and watch all the blood quickly drain from his face. When industry pundits state that TV isn’t or can’t be automated, they’re speaking of reinventing big iron TV traffic systems.

But there are ways to automate TV without reinventing existing TV traffic systems. TV ad tech companies usually pursue one of two approaches to automation, each with its own pros and cons. One option is to develop software plugins that can send and receive instructions and data directly into the traffic system to automate the execution of a programmatic campaign. The advantage of this approach is that it provides quick time to market. The disadvantage is that there are limitations to how much the plugins approach can automate the full TV workflow.

The other option is to integrate independent hardware and software alongside the traffic system and accompanying ad servers. The new hardware and software act as a sidecar system to the core traffic. The advantage is, the sidecar provides fully flexible and automated optimization for programmatic campaigns without interfering with the core functionality of the traffic system. The disadvantage is that installing new hardware and software into the TV workflow is a long and quite involved process.

The good news is that a sizable portion of the TV industry has already successfully adopted the sidecar approach for programmatic TV. They are already delivering on the vision of programmatic TV execution with the same automation and optimization that digital ad tech provides.

So, the right discussion is not if TV can be automated, but which ad tech approach is the correct one for different sectors of the industry. With all of these tech advances happening in TV, let’s finally retire the fax jokes, and raise our expectations for what will happen next.

programmatic, tv

2 comments about "TV Automation is Already Here -- Retire The Fax Jokes".

Check to receive email when comments are posted.

Steve Bottomley from sudaqo, December 14, 2014 at 2:20 p.m.
In theory any media can be automated.
The key barrier to TV automation (based on UK experience) is the imbalance between investment in campaign management on the buy and sell sides. Typically buyers spend their money on pre-purchase planning tools. Sellers have taken on responsibility for managing delivery of the campaign once bought under the scrutiny of the buyer.
Automation would probably require that buyers take responsibility for selecting, pricing (accounting for reach and frequency contribution of each spot) spots then, optimising ongoing delivery. In the UK each broadcast channel varies significantly in delivery profile requiring differing techniques.
Building the technologies required to manage these process is achievable but requires considerable investment. Broadcasters have already made the investment. Will the buying community want to spend the money to replace this? Using tech to remove the buy-side barriers to market entry (planning, optimisation, spot matching and audience analysis) places the media buyer at risk of being disintermediated.
Before digital technologists get too excited about the prospect of automated TV, I would ask what opportunities there are to improve effectiveness of the digital market? With inventory fraud, retargeting losses and a widespread lack of transparency in the digital domain, I would suggest the TV market isn't too high up the priority list for change right now.
Never say never but, don't hold your breath...
Reply

Gerard Broussard from Pre-Meditated Media, LLC, December 15, 2014 at 7:50 a.m.

Great article Walt. There are now a healthy number of marketplace options for integrating buy and sell transactions systems with each other without compromising the ability to achieve goals on either side: for the buyer it's effective targeting closer to the advertiser intention, for the seller, yield management. The appearance of several intermediary "sidecar" solutions is actually healthy for the marketplace as competition usually results in better quality product, built to the needs of both sides of the negotiation table. Steve, as for the plight of digital, even with the measurement and reporting foibles, digital still represents a great advertising value due to low cpms. Hopefully these issues will be resolved during the next couple of years, making the inventory more precious at which point we'll see a rise in cpm.