A path forward for the edited media industry by Philippe Barbe 27 Jan 2021

Part 12 of 20 in a series examining the interplay of Data Science, AI, the media and advertising.

Advertising is roughly a $600B global industry with slightly more than half of that generated by various forms of digital advertising. Google’s 2019 revenue was about $150B if you include their cloud business. Facebook’s was about $70B. While these giants have a large share of the total market there is plenty of opportunity for the edited media industry to grab more of the rest of that $600B of advertising revenue.

The opportunity is particularly obvious when you consider that Google and Facebook ad revenues are partly derived from content created by the edited media industry! Google and Facebook get the revenue simply because they offer a better workflow for advertisers.

For the traditional media industry, regaining control of its advertising value will require considerable investments over the long run. But the opportunities are such at all levels, from regional to global, it is possible to build the proper systems with incremental and profitable investments. It requires vision, leadership, talented software and data engineers, talented data scientists, and long term commitments. But it can be done in a profitable manner.

Faced with the hyper-fragmented landscape discussed in previous articles, the only way for global brands to make rational and efficient marketing decisions is by using data science, including AI, in smart systems.

The only way for traditional media to monetize their audience and content globally is to participate in some kind of “smart” marketplace. The problem precluding such a market is that the infrastructure, systems and automated workflows simply do not exist today.

But, they must be built if the edited media industry is to survive. This requires long term commitments (20 years), but the needed investments can be profitable within a much shorter horizon of 3-5 years, IF the growth strategy is well planned, incremental, and the company undertaking this effort is positioned in the right spot.

The current large tech companies have not been successful. Google left the TV Ad business in 2012, because to keep control of their business media groups and measurement companies blocked them as they will block any similar offerings from Facebook or Amazon.

The solution must come from within the traditional media, be it a media group, a large advertisement agency, a large global corporation, a global data provider in the ecosystem, a software company, or a combination of these.

To be profitable quickly, it requires participation, starting small in a specific part of the edited media industry, and growing carefully as a partnership across the industry.

It may seem idealistic and unrealistic for all these players to work together, but the diminishing returns of the industry will either force that evolution or see them merged with the tech companies. Such a merger will precipitate considerable further contraction of edited media, goes against the lessons of history, and against a deep political desire to keep the free press in democracies.

The system to be built

Whatever entity attempts to utilized Data Science and AI to solve the woes of the edited media industry will need to offer two core functionalities at any scale, from local to global, with systems far smarter than what has been currently designed:

1. advertising campaign planning, buying, execution, tracking and billing;
2. ability to transact edited content.

More precisely, what do these two requirements cover, and what do we mean by far smarter systems?

For starters, the system should be able to integrate a vast amount of data, originating from:

• the advertiser, (i.e. sales data);
• audience measurement companies;
• consumer information companies.

In these systems, a marketing department should be able to:

• integrate with data originating from the advertiser (i.e. point-of-sales data),
• specify high level goals, in terms meaningful for the advertiser (“I want to increase my sales in the North East by reaching new customers.” ; or, “I am launching a new product to play golf and want to let people know about it.”);
• present information to make decisions on the campaign design and allow the user to interact with this information… run scenarios, explore options that balance between reach, frequency, and budget… all in meaningful terms for the user;
• specify contract terms in a machine interpretable manner;
• click ‘buy’ and have the whole buying process initiated within a few seconds;
• then be able to negotiate at scale with very little effort (the key here is to allow for substitution and running scenarios to make informed decision);
• once the whole campaign is bought, upload all the ads and have the system taking care of executing the campaign;
• see meaningful dashboards on how the campaign is doing in near real time, with a forecast on its’ future to evaluate what is at risk and take counter measures or change course.

On the media operator side, the system should be smart enough to:

• help make informed sales decisions and negotiations on a very large scale;
• help media operators meet their budget by providing them information on their past, current, and forecast performances, as well as information on their prospective customers (this information has to be delivered in a particularly coherent, effective and interactive way that is helpful at whatever level, from individual media outlet to full fledge media group);
• offer a user experience that allows a single person to negotiate with hundreds of advertisers each day and in the case of broadcasting capability, allow also to the sale if broadcast time to companies paying for one-way data transmission.

On the content creator and media side, the systems should be smart enough so that:

• content vetted by a media somewhere is available to any media for purchase;
• translations are done, legal copyright compliance is maintained, a fair price is paid (this allows hyper-local content to be monetized globally);
• content is associated with the advertisers goals;
• content is repackages as needed (and allowed by copyright terms);
• content is piped directly to production systems.

And, the system should know how to operate under the various regulations that cover edited media in every single country on the planet.

All this is a lot to ask!

This smart marketplace cannot be built in a year. But it can be built! And, in a way that is profitable in 2 to 5 years.

Who can build the exchange?

But is building such a system creating a big monopoly that the industry does not want?

Yes, if it is an independent company aiming to be everything from an ad agency to a media operator to a content producer.

No, if it is an exchange built by, and for, the industry.

No, if it grows out of partnerships created around data and workflows which no single company has enough resources to build.

The exchange concept is not about replacing advertisers, agencies, media operators, content creators, and software vendors. It is not a platform to dictate what the players do. It is about enabling them to work more efficiently. It is about allowing them to focus on decisions instead of information gathering and workflow.

On the advertiser and media side, it is a platform to operate faster on a much larger scale.

Overall the exchange would be a tool enabling both sides to operate at a higher level, where humans keep control of algorithms from high level specification of the objectives to specification of custom algorithms enabling humans to manipulate the information in a much more efficient manner. It would serve up information to make decisions and automate the non-decisioning processes. It would support decision making, but not making decisions.

On the content side, it is a platform to monetize edited content across outlets globally.

Think of as the “Pro” version of YouTube, able to

• translate written or spoken content into any language,
• able to tag content in a meaningful way to make it easily searchable by even complex descriptions,
• able to pull pieces of content relevant to each other, and
• to which is added a transacting capability so that this content can be sold and purchased with little human intervention, yet used in a way that the copyright owners deem suitable.

There are several economic models for such exchanges, sometimes termed cooperatives.

A striking example of in a different industry is the SWIFT banking network which has been handling the bulk of international financial transactions since 1973.

In 2019, SWIFT had more than 11,000 members doing 33.6M daily transactions. Despite all these financial institutions competing with each other, they nevertheless realized that it was in their best interest to have systems that allows them to exchange information in a very effective manner.

While the scale that SWIFT is dealing with is very much in line with that is needed for a global exchange of edited content and advertisements, the complexity of transacting ads and edited content is far greater than that of financial transactions, Much smarter systems, leveraging the full power of data science and AI, would need to be built.

Many in the industry think this solution is unrealistic. At one time the majority of the systems we use daily were deemed unthinkable and unrealistic yet here they are.

What is unrealistic is to think that the traditional media industry will not contract to an extent that will make it unrecognizable if a solution is not found and implemented.

Conclusion The traditional media industry and its entire eco-system of advertisers, media operators, content creators, and data companies have a choice to make:

1. Keep in-fighting, maintain antiquated business processes, keep losing to tech companies, and die

OR

2. Cooperate in establishing processes that allow them to scale even more and fragment even more to reclaim ad dollars that shifted to the tech companies and find new growth opportunities by developing entirely new markets.

Option 2 is a way forward to stay relevant.

True, this will require a massive investment of resources and money over an extended time period along with ambitious business and technical leaders but there are ways to move forward incrementally and profit quickly, as well as paths to solve the technical challenges.

In the remaining articles of this series I will discuss possible technical solutions.