How Condé Nast Democratized Data Across Brand and Sales TeamsMay 8, 2014 by Abe Selig
Condé Nast is one of the world’s leading media powerhouses. From its inception over a century ago with the purchase of a single fashion magazine, the company has grown to include a readership of over 160 million across 20 print and digital publications.
This success hasn’t come without challenges, one of which is analyzing the company’s increasingly large data sets to identify readership trends and behaviors. Accessing and understanding this information, however, was eating up the bandwidth of the company’s data and analytics team and complicating the work of the editorial and marketing departments.
To help address these challenges, Condé Nast adopted Microsoft Power BI in 2013. To learn more about the effect of the new software, we spoke with Chris Reynolds, the company’s VP of data and marketing analytics, who oversaw the implementation process. Here, we illustrate how the software helped ease the workload of Reynolds’ analytics team and empowered the editorial and marketing departments to access the information they need most.
A Wealth of Information Needed for Publishing Decisions
For a publishing company like Condé Nast, key performance indicators (KPIs) are crucial for understanding how its many audiences respond to and interact with content. This information is also important for providing the company with the leverage it needs to negotiate successfully with advertisers.
For Condé Nast, these KPIs have traditionally come in the form of syndicated data, such as marketing and survey results, common usage patterns and other types of forecasting data. As the company has expanded its digital footprint, however, these KPIs have grown to include real-time web analytics, which looks at online behaviors such as bounce and conversion rates and time on page.
This data helps the editorial and marketing teams make certain decisions about publishing content. For example, audience engagement metrics may lead staff to decide that an article should be published on a magazine’s website, but not in the actual print edition, or to include certain types of articles in international editions of magazines, but not the U.S. versions.
Outdated Tools Prevented Easy Access to Key Data
Prior to Power BI, Condé Nast was using outdated data processing methods, which meant that accessing this information didn’t always translate into better insights. As the syndicated data was supplemented with the constant flow of data from the company’s Web platforms, the data and analytics team found itself increasingly bogged down with query requests from sales and marketing.
These queries sometimes took a day or longer to process, forcing the analytics team to focus the bulk of their time and energy on these tactical tasks and preventing them from engaging in more strategic projects.
“I wanted my team to have the time and space to think more creatively, but since they were constantly busy with query requests, that time and space just wasn’t there,” Reynolds explains. As a result, the team eventually came to be seen as “report runners” who were isolated from other departments.
Another big issue was that the process for submitting these queries was extremely inefficient. “It was basically a glorified email inbox,” Reynolds explains. “People would submit questions through an inbox, somebody from my team would see it come in, pick it up, run the answer in whatever data environment they needed to run it in, and then send it back.”
Meanwhile, without the proper tools to allow the editorial and marketing teams to visualize data, these departments found it extremely difficult to decipher key trends. “One of the biggest issues, especially around syndicated data, was the tools available to produce that data and to make it available to other people hadn’t really changed in 10, maybe even 20 years,” Reynolds says.
He explains that cross-tabulation, a statistical process that summarizes categorical information, is often the most common function used to examine syndicated data, and was one of the functions Condé Nast was using to dig into the audience information in their old database system.
“It’s a great way to do analysis of survey data like that,” he says, “but for a non-research person, accessing data in that format is not necessarily easy.”
Implementing a New BI Tool to Improve Workflow
The amount of time this process was taking finally tipped the scales for Reynolds, who had already been considering implementing a new business intelligence (BI) system to assist with the legacy processes being used to query Condé Nast’s databases.
Microsoft Power BI was a particularly attractive option given its user-friendly interface, which Reynolds hoped would empower both his and the editorial and marketing teams to better visualize and understand the data they needed to harness. In adopting this tool, Reynolds says his main goal was delivering better, easier-to-understand information and improving the workflow between each department.
The first order of business for Reynolds’ team was training employees on the editorial and marketing teams—what Reynolds terms the “brand side”—on how to access and use the new system. A big help here, he says, was the brand side’s familiarity with Microsoft Office, which allowed them to quickly learn their way around the Power BI environment.
Reynolds also had his team members engage in one-on-one sessions with editorial and marketing employees to show them how to manipulate data environments and customize query results. This was done as-needed, but Reynolds says he lauded both his team and the brand side for their eagerness to get the process rolling.
“These concepts can be difficult for people who haven’t encountered them yet,” he explains. “But once they see it, it’s not a very hard sell. They’re like, ‘Okay, how do we learn to start using that?’”
Templated Visualizations Reduce Requests by 30%
One particular tactic that helped brand side employees learn the new system quickly was a method Reynolds and his team determined to be a best practice early on: templating data visualizations with common queries that had previously been filling up the request tool.
“A lot of the visualizations bring together data about our consumers or subscribers, our syndicated audience data and sales data, which includes advertising sales data,” Reynolds says. These three categories had never previously been grouped together in a single platform, which meant combining them for particular queries was a request his team often received.
With the templated visualizations, however, the editorial and marketing teams are able to easily access all of this information, and then customize these visualizations across different data sets (e.g. to compare subscriber data alongside advertising data, or pair advertising data with audience data).
Now, Reynold says, both his team and the brand side are suddenly able to access more detailed information, more quickly. “It’s something that would have taken days before,” he says, adding that he’s seen a 30 percent decrease in the volume of requests his team receives since the templated visualizations were created.
“To me, this says that members of the brand teams are preemptively answering their questions without having to go through our process,” Reynolds says.
Automatic Data Import Frees Up Additional Time
While the data visualizations go a long way in helping Condé Nast’s non-BI professionals understand what the various sets of information mean, Power BI first requires information to be fed into it in order to build these visualizations.
For Reynolds and his team, this need presented an easy way to put to their batch files to good use. These batch files are massives stores of syndicated data that were creating additional query requests and taking up a lot of his team’s time
“In the really old days, batch files were literally binders with paper,” Reynolds says. Currently, these files are in the form of large Microsoft Excel workbooks. Since Power BI integrates seamlessly with Excel, these files can be automatically uploaded into the system, which greatly reduces the amount of time Reynold’s team must spend sifting through data.
Once the data is automatically uploaded, the end user—be it someone on Reynolds’ immediate team or an employee from the brand side—can then use the templated data visualizations to find the answers they need.
“They can adjust how, what and when they’re looking at the data,” Reynolds says. “And they can do this easily, without having to go back through my team to submit an entirely new request because the first one wasn’t exactly what they wanted.”
New Insights Provide Greater Leverage With Advertisers
Once Power BI was in place and employees had become sufficiently familiar with it, new insights were quickly uncovered. For example, Reynolds says, brand side employees were able to use the system to combine audience and consumer data and determine that one of Condé Nast’s brands had specific strengths with consumers who buy photography equipment.
“Not only was there deeper analysis of the kind of content that attracts that kind of user, there was also the ability to apply it to that segment of the audience to see how we perform compared to our competitive publishers,” he explains.
From there, Reynold’s team was in a better position to speak with advertisers—in this case, those representing photography vendors—about how this particular content was attracting the prospective buyers they wanted to target.
“It really helped the sales side understand who the advertisers are that we should be approaching, and how to position themselves competitively,” Reynolds says.
A Tool That Helps Democratize BI Across Departments
This sort of actionable information is precisely the way Reynolds had envisioned Power BI being put to use. Allowing members of the brand teams to cross over into basic data analysis is one part of that vision; freeing up his analytics team to think more strategically about challenges in the marketplace is another.
Reynolds admits that, prior to implementing the new software, he viewed the request tool his team was using to field questions from the brand side as somewhat detrimental. “I always said to my team that it helps us manage the workflow, but it’s a bad thing in that it isolates us from our internal end clients and makes them think we’re just machines sitting somewhere typing numbers out,” he says.
Now, however, less volume through that tool means Reynold’s team can spend more time sitting with the brand side, answering their questions directly and working with them to answer more significant questions, rather than just running time-consuming data queries.
This result, Reynolds says, is a clear example of the “democratization” of BI, where information and insights gleaned from analytics tools are increasingly accessed and used across multiple departments, by new groups of employees. It’s a prevalent theme in the BI marketplace today—one that is gaining traction thanks to tools like Power BI, which make it easy for those without technical or analytics expertise to easily work with data.
Reynolds says that the accessibility of this type of data is leading to the formation of a type of hybrid employee: one who can handle both the tactical, in the form of basic data analysis, and the strategic, in the form of big-picture thinking.
The benefits don’t stop there. Reynolds says the success of Power BI across his data analytics team and the brand side has led Condé Nast to explore its potential in other departments.
“We’re looking at other ends of the business, in environments where we need more BI democratization,” Reynolds says. “If you think about the production of our magazines, our own internal consumer marketing—there are other angles of the business that have plenty of different data sets that need people to share them across huge groups of people, and that’s where I see this going in the short term.”
Image courtesy of Wikemedia Commons.