To say that the television sector is going through a revolution is an understatement: Changes in viewing habits brought on by new technology (mobile devices) and entirely new distribution tools and models (YouTube, Netflix) plus the incredible proliferation of choice for the average viewer are some of the factors that create both pitfalls and opportunity in this highly competitive industry. A broadcaster in this sector who can circumvent the former and exploit the latter will not only survive, but may thrive as well.

In order to understand this changing market, you need information: Who and where is your audience? How and when do they access your content? How can you sustain and grow your audience? Much of this information comes in the form of data—and in the case of a television broadcaster, this data can come from traditional sources, such as Nielsen ratings, but also factored in are audience analytics for your new-media properties (possibly comScore, Omniture or free tools such as Google Analytics). Regardless of the tools at your disposal, you still need relevant expertise to interpret the findings and ensure the results and the data meet at least a minimum threshold for quality and usability.

In the case of TVOntario, the Ontario-based, not-for-profit, government-funded broadcaster, they have the additional wrinkle of a business model that isn’t just dollars and cents. This adds non-standard factors into the way they measure success, such as educational impact and membership rates.

Sharon FlynnSharon Flynn, director of research at TVOntario, agreed to chat with us about “data people,” the meaning of data literacy and turning numbers into meaningful action.

I met Sharon at the Canadian Television Fund (now the Canada Media Fund), which is a federal public–private partnership with the mandate of providing funding to independent Canadian television production. At the time, the fund was just starting to awaken to the importance of audience measurement as a component of their funding formula and Sharon was brought in with the task of laying out processes and methodology that the fund could use. Since then, Sharon has co-founded the Digital Analytics Association–Southern Ontario Chapter and has worked at both Astral Media and CBC.

During our talk, Sharon was quick to act as a myth buster for a few ideas that are a challenge for the data–business landscape, including the notion of free data:

No matter how big the data is, how fantastic the tools are, or how increasingly accessible the skill set is, you still need to have objectives, have an understanding of what the data is being used for, and have the budget for it. Data is not necessarily free; there is still a great deal of data that you can get from Twitter and other high-profile sources but to interact with your customer base and understand who your key constituents are—that is going to cost you some money. So there is always a pause, a moment of disappointment where you realize the project is doable and the technology exists but it is going to cost between this amount and that amount to get it done.

A way to mitigate this cost, Sharon suggests, is to get the “data people” on board right at the beginning:

Someone actually has to capture the data: go out and question the constituents, in the small data world. Or in the big data world, where information is seemingly being gathered automatically, it actually isn’t, until the required processes are put in place to gather it automatically—so that’s where the data people come in. In the media industry, especially with traditional TV, there is a tendency to have a lot of syndicated data (Nielsen or BBM—now Numeris) that somebody else gathers for you and so the process for gathering the data is not widely understood—you pay collectively into a pool and the full cost of that process is somewhat masked, and like magic, all this clean, wonderful data arrives every day. But it’s deeply challenging to ensure that the analyst, junior analyst, CTO and the team understand the important data points to capture and when, and make sure data collection is not a retroactive decision.

Sharon realizes that her perspective might not be shared by everyone in the industry:

I will add the caveat that I’m coming from a digital data perspective, so I may have an over-grandiose perspective on the data’s importance in the industry. I have an MA in media studies—even though I am a data person I am attracted to media because it is so complex.

The numbers—your audience data—from my perspective, was always the product. But it’s not terribly sexy; it’s pretty nerdy. It’s far more exciting to think about the output as the actual shows, which in their defence is what actually drew the audience in the first place. But the entire business is based around a strong number that can convince Coca-Cola to buy ad space on your show, rather than the other guy’s. I think for some aspects of the media industry it has been a real shock to the system to become comfortable with anything on the digital realm (websites and mobile audience tracking) as it is essentially the media entity’s responsibility to collect the data.

As a data person herself, Sharon recognizes the value of better integration of data people into decision-making processes in organizations:

Initially, in the newspaper industry, for example, there was a very deep culture of separation of journalists and data people, as it should be, in a way. However, it did mean when the time to make decisions about moving content online came around, there may not have been enough of the data people in the room to remind them that they were going to give away the very thing that kept them going. So this was what, in my opinion, was catastrophic for the newspaper industry.

The newspapers went ahead and put their content online, but there wasn’t any currency for the industry; there wasn’t a way to talk about online content the way there was for traditional television and newspaper, which had been decades in the making, and they were left with selling their digital content, the numbers, in a much less sophisticated way, to an audience that was still struggling to wrap their head around the shift to digital. It was a challenge, and I felt if the larger media industry had brought a couple of the numbers people to the table, those data-informed people might’ve urged some caution, reminded the industry that the ratings data—especially on the television side—that had been the cornerstone of the financing of the entire industry came about as a result of congressional hearings.* The need for a currency was recognized as something that was so important, but it had been forgotten in the midst of time, in the history of the industry.

When I’ve observed projects I’ve been involved in, or organizations I’ve worked with, usually the person bringing the data to the table is both enthused and expected to be in the subject matter and the data. And you have at the table what I call a “power user” of data. Good quality power users of data usually have a lot of confidence in saying this is what I want to find out, this is what my hypothesis is, and I’m trusting you to test it and kick it. And they typically see you as more than just a resource; it’s a more collaborative experience—something that I think a lot of CEOs and other leaders of organizations currently have with perhaps the marketing team. And where it works best is where you have data brought in at the very beginning.

Bridging the gap between the data people and the non-data people requires an effort from all parties involved—from the data person’s point of view it works when you listen, and when you distance yourself from the findings. I find the mistake that data folk make is that they become just as connected to the findings as their client is to the hope of what the findings will be. They get disappointed when all indications say stop and the folks in the room say “no we’re going to continue for these other external reasons.” And the way I’ve come to peace with that, and the way that I’ve re-framed that in my head, is to say that the numbers, the analysis, the research—quantitative and qualitative—is one of the data points you bring to the table. And the grey matter of everyone around the table is an additional data point. Everyone involved has their respective knowledge and information and it’s important to view each other as partners in making big decisions in media.

It is difficult to communicate the impact of good data in a somewhat traditional industry that is going through so many changes in a relatively short span of time if you don’t speak their language:

It’s why I’m passionate about figuring out how to present information, how to talk to people, how to communicate—be it through a PowerPoint presentation, through email, or over coffee—to get them to a place where they go “oh you’re my advocate,” as opposed to using a lot of biz/numbers/stats speak that is impenetrable. They’re smart people, so you don’t want to be in a situation where you’re somehow creating confusion or coming across as not really helping the conversation.

My chat with Sharon brought to light a number of challenges in the TV industry—ranging from an over-optimism in what data actually is out there and how you gather it, to how useful it is. And there’s sometimes a fairly wide gap between the people who manage the data part of it, and the people who make key decisions on what kinds of media to produce and present; it’s clear we need both groups to be more organically integrated. Having been on both sides of the table, Sharon is a credible advocate for the importance of data.


Special thanks to Malavika Kumaran for additional edits and input.

* Note refers to the 1966 Harris Committee that came out of the “quiz show” scandal. The committee, which heard wide issues around TV, also looked at ratings and while falling short of recommending legislation to regulate audience measurement “it had a sobering effect on the ratings business—effects still evident today” specifically with regard to the scrupulousness of the data. Source: Webster, J. G., Phalen, P. & Lichty, L. (2006). Ratings Analysis: The Theory and Practice of Audience Research (3rd. ed., p. 103). Mahwah, NJ: Lawrence Erlbaum Associates.

Joseph Lalonde

Joseph is the Data Manager, Data Catalyst at MaRS See more…