Blog

Unlocking gold: how to turn your audience data into revenue

Last updated: December 2, 2024

featuring quotes from Jacob Donnelly, founder of A Media Operator and Marissa Zanetti-Crume, Global Head of Product of Bloomberg Media

Media companies are sitting on an untapped treasure trove of data. But too often, they fall short of converting it into meaningful revenue and deeper engagement. 

So we partnered with Jacob Donnelly — the founder of A Media Operator — for a webinar with Marissa Zanetti-Crume, the Global Head of Product, Bloomberg Media to learn how they’re using data to grow revenue and engage their audience.  

Streamline your tech stack: Boost efficiency, unify data & retain talent! Get expert tips now:

Work Email *

First Name *

Last Name *

Company *

Job Title *

Read Now

Together, they shared actionable ways to use your audience data to drive more revenue, engage your readers and future-proof your media business. Read on for a summary of the session or watch the full session here. 

Editor’s note: Quotes have been slightly edited for clarity and brevity. Check out the full recording here for additional context. 

Jacob Donnelly: Can you share your perspective on the role of data in shaping the future of media businesses? And when we talk about data, are we looking at it strictly as just audience data? Is there a broader category of data that matters in your decision making among those at Bloomberg Media?

“Data means something different to everybody in an organization based on what your role is or even what industry you’re in,” Zanetti-Crume says. “And so there are teams within our organization who are keenly focused on understanding who our audience is. 

There are teams that are keenly interested in how our content is performing, what in our journalism is really sort of resonating, how are we reaching our customers through the work that we’re doing? 

And then there’s the data that’s the nuts and bolts. It’s like how we track our numbers, it’s how we track how we’re performing, it’s how we look at historicals, it’s how we drive projections. 

And so I think both perhaps the power of data, but also the thing that’s a little intimidating about data is that there’s so much of it and there’s so much that you could do with it depending on who you are and the question that you’re trying to answer. And so it’s infinitely powerful in many ways.

How do you distinguish between data that is nice to have and data that is critical?

It’s something Bloomberg is constantly revising, Zanetti-Crume say.

“I think one of the things that’s most important is to really take a step back, understand our business and identify what are the core north star metrics that we think are most important,” Zanetti-Crume says. 

“So if it’s a way we’re looking at measuring revenue, if it’s a way we’re looking at measuring engagement or measuring our audience, we as an organization are trying to drive alignment around what those key measures are. And then from there, asking ourselves,’ All right, now how are we going to hit that number?’ What are the KPIs or lever metrics that are going to help us move our numbers forward? And that’s helped us really focus in on the data and the metrics we think are critical, ultimately, to understanding the success of our business. 

I lead a product team and data is our bread and butter. We could look at any data point to measure the success of any feature or metric. But really holding ourselves accountable to the ones that we agree are going to be our core measures has been really important. 

If you think of the center of an onion or a bullseye, there’s a set of metrics that we uniquely and critically care about as a business. Then as we move outside the center of that bullseye, we’re looking at other metrics — health metrics, performance metrics, other performance indicators. 

When we’re gauging success, we want to be thoughtful and limit ourselves to the ones that matter most.” 

Part of the problem with data — and trying to unlock the value of it — is that is it exists, but it’s hard to get it in a form that’s easy to use. How does your team solve for that? 

Bloomberg has a multi-layered process of intaking, organizing, storing and activating audience data, Zanetti-Crume says.  

“The main input for us is what people are doing and how they’re engaging with our platform and products.” 

“We take all that data and it flows into our central middle layer. We collect a lot of data, everyone has a lot of data, but if you don’t know what it is and you can’t access it, it becomes less powerful.”

“So this middle layer is really thinking about how we’re managing the data, how we’re parsing the data, and then thinking about how we’re preparing that data for a variety of levels of output. That could be for our analytics team to share throughout our organization.” 

“Or we’re using that data as input for features that we’re circulating back to the front end. Personalization is a good example of somewhere you’re getting an input, then we’re providing value back to customers on the front end.” 

“Then that last layer is thinking about when we’re leveraging that data to activate customers across platforms and channels that aren’t our owned channels. And how do we think about the data that’s critical? How do we think of our data in different ways and what are the use cases?” 

Bloomberg Media has made a big push over the past half a dozen years into subscriptions. How do you use data to identify opportunities for deeper reader engagement?

“We’re continuing to just challenge ourselves about how we’re thinking of answering that question,” Zanetti-Crume says. “I think the root of it for us was that aligning on that North star metric is most important.” 

For Bloomberg, those North Star metrics were the frequency and the depth of readers’ engagement. 

“So how frequently are people engaging across our ecosystem? How frequently are they engaging with those front end platforms and products? And then when they are engaging, how deeply are they engaging? What are they engaging with? Does engagement change at different times of the day or different times of the week, or does an engagement change for different segments or populations of our audience?” 

So step one, for Bloomberg, was identifying who their users were, then looking at their engagement data to identify opportunity areas, Zanetti-Crume says.

“As you can imagine, at Bloomberg we cover financial news and we have market data,” Zanetti-Crume says. “So we start to see trends of where we see deeper engagement when people are coming to Bloomberg to understand the news that’s moving markets in that particular day. People are engaging with the things across our site and platform that are helping them answer that question of, ‘What is happening in the markets today that’s going to help me make the right decision at work?’”

That’s helped Zanetti-Crume’s team form hypotheses around what users need from Bloomberg, then develop features and products that’ll best those needs.

“How might we improve a feature for them that maybe makes it a little bit easier for them to get to the answer of that question about what’s happening in the markets and what’s moving the markets in the day to day?” Zanetti-Crume says.

“So when we’ve thought about engagement, we’ve really been asking ourselves, ‘What are those data points that indicate deeper engagement?’ Then how do we start to knit them together to understand that customer journey, and then prioritize feature products and builds based on that?” 

Can you share an example of a successful initiative that improved reader retention based off of that data?

Zanetti-Crume and her team have made several homepage improvements based on this engagement data, she says.  

“We noticed a good percentage of our subscribers are actually spending time on the homepage, which might be a bit of an anomaly for a lot of other sites,” Zanetti-Crume says. “Then it was a question of what are they clicking on our current homepage?”

From there, they talked to their customers to see how they could further optimize their homepage experience. 

“That really helped inform our homepage, which was focused on making it easier for people to understand what was breaking news, what was market breaking news, thinking about how we sort of tie together better our coverage across a news story.” 

One example is the US presidential election, she says. Bloomberg readers are really focused on how the election will impact the market, which requires many particular articles involving specific nuances and context.

“How do we show the context of how all this coverage helps ’em understand [those core issues]? And so from that, we started to ask ourselves, ‘How is subscriber behavior changing now that we’ve made those updates to our homepage?’ And so this is where data really is a constant.” 

This needs to be a continuous process, she adds, where the team uses engagement data to inform product changes, then returns to the data to evaluate their process and iterate as needed. 

“We started out with us understanding how people are spending time on our site. We talked to our consumers, we launched a new homepage, and again, we’re sort of back to the data now that we’ve launched, what are we seeing in the data?” 

Their key discovery this time: A core part of their homepage page was engaging with a newsfeed of all published articles, located below the fold of the site. This told Zanetti-Crume that a sizable part of the audience was willing to scroll down to find everything Bloomberg’s published. 

“We noticed that for this little module that was below the fold that people had to scroll to, people were engaging with that at higher rates,” Zanetti-Crume says. “So again, there’s that data point around engagement, but what do we do next? 

From there, their product managers started thinking about the user need that the newsfeed module solves for and tried to build on it. And they’ve developed the module to make it even more useful. “So now we have the option to click through and you can see everything that’s latest, and you can filter by it.” 

What does the Bloomberg subscription model look like? Is it a metered wall where you’re providing two or free three articles and then the reader has to pay? Or are you dynamically firing a paywall using individual user behaviors to dictate when they should or shouldn’t subscribe?

The Bloomberg team’s experimented with a variety of paywall types and there’s no one right answer, Zanetti-Crume says. But they’re constantly using data to evaluate their paywall performance and conversion funnel as a whole, not just across the board, but for readers from specific referral sources and audience segments. 

“We’re looking at if people are coming to us from search or social or a newsletter, for example,” Zanetti-Crume says, “How do we understand, through data, how they’re engaging with the site and how frequently they’re hitting our paywall? Are there moments in time where we might want to loosen the paywall to allow more users to consume content across the site and engage more deeply with the brand?”

The key is to stay flexible and responsive to what the audience — and their engagement data — is telling them, Zanetti-Crume adds. They’ve also been experimenting with machine learning to create more responsive paywalls, rather than providing one flat price across the board. 

“Should we be thinking about [pay]walling based on what people are reading? Should we be walling or not walling based on what is the news of the day?” Zanetti-Crume says. “Should we be walling based on what we know about you as a customer? The thing that we’ve held true is that we should continue to be dynamic in our approach rather than just providing a one-size-fits-all experience. That’s something we’re continuing to test and evaluate at any given moment.” 

How does the editorial team use data and how do they work in conjunction with the product team to make sure that they’re making the best decisions with that data?

Bloomberg’s fortunate to have an audience that engages with them across a variety of channels and content types, Zanetti-Crume says. So the editorial team’s got a strong foundation of engagement data to inform their hunches. 

“Traditionally, people are coming [to Bloomberg] when something is happening in the world, and reading about how it’s going to impact technology, how it’s going to impact regulation, how it’s going to impact politics. So we’ve got an organization or a user base that’s really consuming content across the breadth of what we cover.” 

“The analytics around that is really important, “Zanetti-Crume continues, “And it’s about understanding not just based on what the headline is of a particular story of what’s happening in the world, but the kinds of stories that are resonating with users.” 

That’s informed the work that their newsroom is doing. Now they’re not just reporting and understanding what topics work. They’re also understanding what kind of content resonates with their audience. That’s helped the editorial team experiment with new, more interactive content types more strategically. 

“So how do we think about lists, graphics, or deep dives into wealth that resonate really well with our customer base?” Zanetti-Crume says. “Then we have audiences who are asking more basic questions like, ‘What is the yield curve? What is this thing? Help me understand this.’ And so you start to see we have more content that’s a little bit more in the vein of an explainer.” 

And from there, the editorial team’s begun to combine content types to deliver informed but accessible coverage of their stories. That’s especially important for someone like Bloomberg, which covers complex economic news that requires quite a bit of nuance. 

“I think our newsroom does a great job of thinking about, ‘Here’s a story that happened today — it’s really complex. Here’s the explainer that’s going to help you understand it,’” Zanetti-Crume says. “Data is not just informing how we’re writing stories and how we’re reporting, but also how we’re packaging content together.”

How do you see AI transforming the way media companies monetize their data?

It remains to be seen, Zanetti-Crume says. Artificial intelligence has already changed so much in two years that it’s hard to know what it will evolve in the short term, to any degree of uncertainty, let alone how Bloomberg will use it to any degree of detail.

“Candidly, I think you see a lot of people trying lots of things in the market, and I’m not sure where we’ll land.”

For Zanetti-Crume, it all comes back to the audience. She and her team are considering how to use AI in a way that’ll respect and enhance their audiences’ experiences.

“One of the things that we’ve held true to as we think about our audience is, again, our audience is engaging with us and we want to be respectful of them, and we want to be mindful of how this will drive value within our organization. And so I think it’s about. ‘How do we think about what we value as an organization?’ and then, ‘How do we evaluate our usage of that data?’” 

“That’s something that we’re constantly thinking about for every particular use case. Is AI right for this? Is it not? Is it something else? And so I don’t have a good answer other than to say that I think it’s going to probably change 15 times over in the next six months and we’ll see where we land.” 

What is one piece of advice that you would give media operators looking to start or improve their data monetization strategies?

Zanetti-Crume encourages people to start small. Otherwise, you’ll fall victim to analysis by paralysis. 

“Think about one question that you have that you can answer with data. Answer that question — What’s one metric that you want your organization to really rally around?” 

A progressive approach will help you become more comfortable with data, answer your most important questions, and form the infrastructure you need to enable the use of data throughout your organization.

“By starting small, you’re going to start to get more comfortable with your data. You’re going to start to learn what are the tools that you need to enable your organization through that data? And is it training? Is it tooling? What do you actually need to allow people to use that data?” 

“One of the things that is really great about Bloomberg is that we’re using data day in and day out to make decisions, and we’re using it to start discussions about how we should decide to move things forward, even if it’s the smallest feature or if we’re thinking about broader strategies that we want to tackle.” 

That’ll help you advocate for data-driven decision making throughout your organization.  

“Building that muscle of using data in those conversations is going to be really important. Once you start to do that and people get comfortable with it, then you’ll have some tangible examples of, ‘Hey, we had this data, we made this decision and this is what we saw and the numbers, and it was net positive.’ You start to get more people wanting to engage with data and wanting to inform themselves with the data. So just starting and figuring it out I think is the most important thing.”

Subscribe to our newsletter

Sign up to get our latest articles sent directly to your inbox.

Email *

Subscribe