The Sacramento Bee’s been busy in the field of data. Recently, The Bee has been working closely with universities like Stanford on using data to create personalized approaches and systems to better serve readers and advertisers. Tom Negrete, who is the paper’s director of innovation and news operations as well as the paper’s previous Managing Editor, has been leading this project.
In addition to working with Stanford University, Tom and The Bee have developed relationships with faculty at the Annenberg Innovation Lab at the University of Southern California and at California State University in Sacramento, around the topic of developing apps aimed at young users. Tom also said he plans to chat with other universities about future projects in 2015.
Tom took the time to update API about these research projects and more. He discussed how newspapers can work with academic institutions, the level of energy these partnerships can bring to both the media organizations and the universities and why forming these relationships is easier than one might think.
Can you tell me about how The Sacramento Bee collaborates with academic institutions?
These relationships focus on getting informal advice on tech issues as well as more formal tech projects.
This spring we worked with a couple of graduate political science students from Stanford to do some A/B testing on emails to readers, testing certain calls to action and also the students designed a survey to see what impact our electronic voter guide had influencing users’ interests in election coverage.
A year ago we worked with two teams of grad students from Stanford University’s computer science and journalism departments to explore how newspapers might increase engagement through content and technology. One team developed a concept and prototype that allowed users to comment at the paragraph level. A second team mapped out for us possible ways to scan content, tag it, collect user data and weight that data to make real-time recommendations to readers on content, even possibly ads.
Could you tell us about The Bee’s collaborative big data research with Stanford University. What came out of the partnership?
They gave us a road map to personalization and we’re following it … and moving toward personalization, hopefully later this summer. The students had two academic quarters (about 4-5 months) to work on the project and did not have time to refine their prototypes but they put on paper for our tech folks how we can do this.
The students also did a lot of research to see what other newspapers were doing, and this led to looking at what European newspapers are doing with big data, and especially big data companies like Cxense and Neodata that help publishers. We are now working with Neodata to develop a scope of work. Later this summer we hope to launch recommendation tools, begin to identify our most engaged readers and hopefully develop campaigns to identify casual readers and move them to be more frequent readers, identify frequent readers and move them toward becoming subscribers.
Could you briefly explain how incorporating data driven systems such as these are beneficial to news organizations?
Well, we are in the era of big data and I strongly believe if any company wants to remain in business they have to move in that direction. I think “big data” is the new electricity. You can still operate your business by candlelight if you want, but you’re going to have a real challenge as more and more companies around you convert to electricity.
Our biggest national advertisers are using big data to better understand and anticipate customers’ needs and wants – and it’s just a matter of time before more of our bigger, local advertisers are doing the same thing. They know what their customers look like demographically, and the smartest retailers/advertisers understand that not all customers have the same value. Some spend more and some less, but advertisers can identify those customers most likely to spend more and then focus on that since it’s the best return on investment. So if you are trying to sell this advertiser ad impressions, page views or monthly uniques, your business is rapidly becoming irrelevant to these advertisers. If you can identify your audience and see if you have folks that advertiser is looking for, now you have more value and can probably charge that advertiser more to talk to readers that look like their top-spending customers.
I also think publishers need big data to help readers find the content they want and make the experience better when they come to us. We can potentially give each user a unique, far more relevant experience. If we do that, we should see better engagement.
How could a news organization go about implementing one of these systems?
The Stanford projects were all open source. Stanford worked with us, but the intent was to help all newspapers.
Anyone can look up the projects through Stanford’s journalism school and learn more about them — as well as other projects the school is working on to help our industry — and see what ideas they can implement or follow.
How do these partnerships and the research these collaborations produce enrich your news organization and/or impact The Bee’s strategic initiatives?
It’s about staying up to speed on technology. Things move too fast to just chat in-house or with other newspaper editors who barely have time to stay on top of their daily job. Working with universities is a way to take on complex projects and keep them moving forward when your company has some capacity issues – and what newspaper doesn’t?
Also, we noticed a lot of top tech companies in California all have representatives on university campuses, particularly in the computer science departments. Part of that is recruiting, but a big part of that is also getting good connections that help you keep pace with latest tech changes, trends and just being able to chat with experts who can advise how your business might leverage some new technology.
We’ve also found working with universities is great for retaining talent. It gives folks a sense of hope and inspiration to be working on a project with Stanford. Several of our best tech folks have told us working with students and faculty is really energizing. Even in the face of challenging revenues and lots of bleak talk about the future of newspapers, we have a sense of excitement about our future.
In your data research with academic institutions, could you highlight some of the best ways in which news organizations can leverage data gathered from their readers to improve readers’ experiences and create revenue opportunities?
We found European newspapers that are using big data and doing some impressive things we hope to be able to do one day. In Italy, Neodata is able to help publishers track readers’ behaviors and predict with 85 percent accuracy if that reader is female or male, and then target them with content or an ad campaign that gets more clicks.
In Norway, different newspaper companies are working together to track reader preferences, make content recommendations and share content so readers that get real-time content recommendations keep clicking, and bouncing from news site to news site. When I asked why they would do that, one editor said we are competitors and while we don’t like one another, we all hate Google more. I love that. The thinking is even more innovative than the technology they are using.
Frankly, American newspapers have trained their readers to go to Google — because you can’t easily find what you might want to read next on most news sites, so you go back to Google to search and maybe you don’t go back to a newspaper site. In Norway, [newspapers/media] are focused on creating a great news-reader experience, and they are betting if there are more engaged news-readers they will all win. Maybe not Google, but those who create the news content win.
In your opinion, what is the biggest challenge and the biggest opportunity for news organizations seeking to track their readers’ “social footprint”?
Transparency. It’s both the biggest challenge and opportunity. Our big data tools and know-how may never rival Amazon’s, Facebook’s or Target’s. But we can and should be better at transparency.
As newspapers we have an obligation to be transparent with our readers. This won’t be easy because we are entering a space not easy to understand, let alone explain. And as readers learn what Facebook is doing, there is growing public concern about privacy and how their data might be abused or put into an insecure situation. But your social media footprint is sort of the new credit score, and there is an opportunity there for us to help our readers understand the era they live in and how they can take some control over it. This is a simplistic example, but we have done coupon classes for years with our readers and made some of them very savvy shoppers. Could we do the same on big data and social footprint?
We plan to focus our initial big data efforts on content recommendations, and we think that is far easier to be transparent about and get reader advisory groups to work with us in developing some best practices and guidelines to help us move forward. We also hope to get a lot of help from academics, as they are really interested in the topics of transparency and big data.
Our attitude is that the data belongs to the user, and we get to leverage it only if we can demonstrate that we can create additional value for the reader. If we can remain focused on creating a great experience around content, than it’s much easier for us to be transparent than some retailer who merely wants you to buy another pair of shoes. I also think if we can better identify the kinds of ads you want to engage in, that allows us to stop bombarding you with drop down ads or remnant ads that you may not find relevant. You see fewer ads, but the ones you do see you want to click. That is a better experience for user and advertiser.
Do you have any advice to offer news organizations, or universities, looking to establish similar research partnerships?
Sounds overly simple, but just ask.
I was really apprehensive the first time I reached out to USC and Stanford. But, even in the computer science departments of these schools, they understand newspapers serve an important role in their communities and that publishers are struggling and need some help. You do need someone at your organization who is given some time to talk, explore possibilities and determine what relationship and project your publication should be cultivating, working to make happen.
My bosses were initially shocked that universities would help us and not charge us for the help. But the universities understand we’re a great lab to learn in; by helping us, they are indirectly having an impact on our community, too. I try to frame projects, the collaboration as focusing on improving readers’ experiences and civic engagement. If we can do that, everyone eventually wins. It’s in all of our interests to have a more informed and engaged public.
You might also be interested in:
By sending data from their targeted audiences in Adobe Analytics into MFN, Crain was able to more clearly understand what topics, categories, and even story types were engaging readers in key parts of their coverage area.
Ways to support conversations for balancing innovation and stability in your news organization, essential considerations about this often overlooked topic, and guidance to include them in your technology decisions.
Successfully and efficiently marketing your work can be hard, especially for local news teams with limited resources, but marketing yourself to your audience is an essential skill for news organizations to drive revenue and promote sustainability.