Despite the recent boom in streaming video, plenty of areas still need attention if everyone in the ecosystem is to thrive… especially with a much more difficult economic climate looming.

That’s the takeaway from the seemingly endless parade of in-person TV industry gatherings over the past couple of months– from DEG, NATPE, OTT.X, Future Media, Questex, and many others.

Here, on the other side of the July 4 holiday, several potential areas of improvement kept coming up in panels and conversations, as well as an intriguing suggestion to transform business relationships around that most vital streaming currency—data.

Many of those I talked with or listened to during all those events spent at least some time chewing over what I’ll call the “plumbing” of streaming. This includes the basic data connections between creators, services, distributors, and customers, not to mention the tracking and using of all that data.

“Measurement took over Cannes (Lions) this year. It was all that people seemed to talk about,” said an executive with one data-benchmarking firm.

A lot of the concerns revolve around the data and metadata that are supposed to be the big differentiator between legacy broadcast/cable operations and the streaming future. All that data is supposed to make for smarter programming decisions, more precisely targeted advertising, better content recommendations, and much more.

Data these days is a bit like Will Rogers joking about the weather: everyone talks about it but no one ever does anything. And right now, we need some things done. One key issue: getting reliable and sufficient access to data about your own shows’ performance on someone else’s distribution platform or service.

There are a couple of layers here. One is the challenge that show creators and performers (and their agents/managers/et al) have faced for years in prying audience data out of Netflix. It wasn’t much of an issue in traditional TV with Nielsen ratings as a third-party arbiter, however imperfect. But Nielsen isn’t what it was, and no one else has access to all the streaming data either.

Several companies are vying to become the post-modern Nielsen (including Nielsen), especially as even Netflix moves to an ad-supported tier. With a new constituency of brands and marketers clamoring for reliable, third-party data, something’s gotta give.

The streaming services themselves face a second layer of frustration—trying to find out how their shows are performing on distribution platforms such as Roku, Amazon Fire TV, Google TV, and Apple TV, the OEMs, and the aggregator channels that have popped up on many of those platforms.

Sinclair Broadcast Group’s hybrid STIRR service – which lets viewers access dozens of FAST channels alongside the local news and other content from its broadcast stations in dozens of markets – provides programming partners with reams of highly granular data, an executive with a data-services provider told me.

Compare that to Roku, which has a far larger and more lucrative platform than STIRR but also provides far less information about how its partners’ programming is performing, the executive told me.

That Roku reach is tempting indeed. But that data is important too. Are audiences watching episodes all the way through? What kind of audiences are they? When do they click off? Do they come back? How are shows doing compared to broader performance benchmarks (especially when broken down by demographics, genre and other categories)? Good luck on getting that on the most widely used U.S. streaming platform.

Then the executive I spoke with asked an interesting question—What if companies paid Roku and other companies for access to all that beautiful data, in something like cable’s retrans fees? That at least would be the beginning of industry-changing negotiations.

Yes, companies would be paying to get access to their own first-party data. But that data is valuable in the right hands, and the industry is starting to figure out how to use it. Once they figure out how to get it. And doing deals to pay for the data is the quickest way to make that happen.

One potential roadblock to that goal is the streaming industry’s many complexities which have proven challenging to brands wanting to shift money from legacy platforms. During a panel I moderated last month, Future Today co-founder Vikrant Mathur offered a solution—simplicity.

These days, Future Today focuses on its three big FAST channels: Fawesome, Happy Kids, and iFood.TV, far fewer than it used to trumpet. Last month, the long-time streaming company announced the launch of a dedicated sales team, FTI+, to help legacy brands and buyers better navigate the many-layered streaming ad ecosystem.

“We want to simplify the message and make it easier for these people to understand,” Mathur said. “Digital buyers are used to a landscape that's very, very fragmented. A big focus for us is to simplify. At the end of the day, (traditional buyers) want clean, family-friendly programming that is lean-back on a big screen.” 

This week, Future Today extended its “channel-as-a-service” platform to major advertisers so they can create their own streaming channels. Besides operating its own Big Three, Future Today already provided such white-label streaming services for about 300 channels belonging to media companies.

“Brands in every category are already investing billions in video content to better connect with and engage their target audience,” Mathur said. “With our CaaS offering, we can help them activate that content in the CTV environment, where the audience is exploding, through dedicated, branded channels.”

Such an approach can at least fix one area of data shortcomings, giving companies direct information about the audiences for their own media outlets. It doesn’t fix the Roku data question, but it seems likely to be more and more common in the future.

Finally, there’s another big area where data and metadata aren’t delivering on their vast promises: how systems search for and recommend shows to a viewer.

Monica Williams, an NBCUniversal SVP of digital products & operations, used her bully pulpit on a conference panel in Denver and in a subsequent conversation with me to point out that people don’t talk like metadata program descriptions.

Basically, normal humans use very different language than do the metadata-based search systems they’re saddled with. Such search and recommendation functions seldom capture the nuances.

For instance, a show may be labeled a “comedy,” but what kind of comedy is it? “Satire?” “Raunchy?” “Teen?” “Romantic?” “Dark?” “Animated?” “Short?” “Series” “Kid friendly?”

There’s a big difference between a standup special that lasts an hour and a scripted feature that lasts an hour and a half, but both are “comedy.” You get the idea. The result, however, is that viewers get crummy results way more often than they should.

The problem only gets worse when a show is made by one company, distributed on multiple platforms beyond its own, and licensed internationally to still others, said Williams.

There’s no consistency across all that. How one company describes its programming (and tries to measure its success) doesn’t necessarily match distributors’ formats or brand expectations, affecting programming and advertising across the industry.

These conversations all come as Wells Fargo analysts predicted a cooling ad market for the rest of 2022, particularly for ad-fueled companies such as Roku, Tubi, and Pluto.

“The length and depth of the recessionary slowdown will determine whether pain makes its way to the longer cycle areas of the ad market,” Wells Fargo analysts wrote.

The industry also needs to clear up what The Information amusingly called “a funny quirk” of the ads that are transmitted through dongles, pucks, and similar external streaming devices.

Many owners of those devices don’t turn them off when they turn off the connected TV. That means the video signal keeps playing, not incidentally creating ad impressions where no one’s watching. The result is millions of dollars are paid for ads that literally can’t be seen.

The study by GroupM and iSpot.TV used Vizio Smart TVs for its initial study of such “ghost” impressions, which go away when the TV OS is built into the Smart TV itself, rather than an external device. Next up are similar studies using the screens of other OEMs’ Smart TVs.

Given the many millions of external streaming devices, there’s reason for brand and buyer skepticism about at least some kinds of data. But also given the market opportunity here, expect this corner of the data jungle to get cleaned up sooner than later.

As Deloitte vice chair and tech sector lead Jana Arbanas put it: “In times of abundance, there is less scrutiny when it comes to individual performance of different types of media, or different products, if you will. But in a recession, that becomes a laser focus relative to the objectives of the products and whether they are meeting them.”