Testing our hypotheses: Here’s what we learned from our 7 Local News Business Model Challenge grantees

Our mission at the Lenfest Institute is to find and develop sustainable business models for local journalism. 

Since we are working on a problem that doesn’t have a clear end goal or a single one-size fits all approach, we believe that the best way to learn and make our mission more clear is to think through our assumptions about what it takes for a news organization to be sustainable and then test those hypotheses. The process we take, both as an organization and for the projects of all of our grantees, is a scientific method approach. We use our assumptions to create a hypothesis, test it via a project, and then evaluate it to either prove or disprove the hypothesis. This method helps us to share lessons of both what did and didn’t work so that the industry can use this data to inform the work in their own organizations.

Our organizational hypothesis is that in order for a news organization to be sustainable it must have three key ingredients: 

  • Relevant, high-quality journalism that interests audiences and serves the public good. 
  • Efficient, sensible technology that makes it easier for journalists to produce good journalism and convenient for audiences to access it.
  • Newsrooms and coverage that reflect our diverse communities. 

We put these hypotheses into practice last year when we launched the Local News Business Model Challenge, a program to support projects exploring new business models to support local journalism. Our thought was that if we provided resources for organizations to experiment with new products and revenue models then we would uncover new ideas that could make the path to sustainability more clear for the industry.

We awarded seven grants of $50,000 to organizations experimenting with projects that ranged from new ways of distributing and automating content, improved approaches to reader acquisition, and the creation of products that help publishers generate revenue from news adjacent areas.  

And although all of the projects have not completed yet, they haven’t disappointed. We have learned a ton from the grantees’ successes and failures. 

There are overarching learnings we’ve seen from the entire cohort. Finding new revenue models requires a significant time and monetary investment. Because of this, it’s critically important that news organizations are patient with themselves, and give themselves the space they need to invest in a promising project and see results before deciding whether to continue building a product. Many of our grantees had to double their timelines before they could see substantive results from their work.

Below is a short overview of each of the projects following the hypothesis approach.

In the upcoming weeks, we’ll dive deeper into a few of the projects as well. To keep up-to-date with the latest posts, sign up for a special pop-up newsletter below, and we’ll send you an email with each new report.

Distributed Media Lab

Who: Distributed Media Lab’s mission is to help establish a more viable economic framework for quality original content on the open web. It is focused on leveling the playing field and driving market opportunity for quality publishers, while providing a compelling user experience centered around quality and relevant content, distributed in a way that is designed to flow naturally across the decentralized architecture of the open web. It helps publishers become digital platforms able to aggregate and syndicate with absolute efficiency.

Hypothesis: If Distributed Media Lab develops a decentralized distributed model for syndication then it will expand advertising and membership revenue opportunities for publishers.

Test: Distributed Media Lab shared reporting from CalMatters, a nonprofit news site covering California government, to local news publishers across the state using a platform that allows media companies to syndicate content from other websites natively using a viewer that enables readers to view coverage without leaving the original one they visited.

A screenshot of what AMP looks like with a CalMatters story on the Times of San Diego website.

Results: News organizations including The Long Beach Post and The Times of San Diego  started sharing CalMatters stories on their sites using AMP. CalMatters received more traffic on its Voter Guide via  AMP pages shared by Distributed Media Lab than from Facebook.

Hypothesis Result: Inconclusive, but promising. Now that the viewer has been implemented in various newsrooms, Distributed Media Lab is waiting for results about whether this results in increased revenue for newsrooms

Key Takeaway: One participating publisher saw a 40% increase in social engagement since it started integrating an AMP experience that allows people to view the story while staying on Twitter.

Pico

Who: Pico is a customer relationship manager that empowers digital publishers to be built around audience relationships.

Hypothesis: If Pico can help local news organizations invest in paid lead acquisition tactics then newsrooms will be able to predict with strong confidence the rate of return on their investment.

Test: Pico provided grants and support to newsrooms to experiment with paid acquisition with Facebook lead generation ads and petitions on Care2, a website that helps connect mission-based brands with potential supporters. Once the publishers acquired those leads, they were added into the Pico’s customer relationship management software to then try to convert them to paid supporters. Six publishers joined the four-month long experiment.

Results: The publishes gained more than 10,000 new leads with the average cost per lead under $3. The average open rate for an automated on-boarding email series for new leads was 29%. Pico is monitoring the conversion rates now.

Hypothesis Result: Inconclusive, but promising. We’ll be able to actually assess the hypothesis once Pico is finished evaluating how many of the new leads actually convert to paying subscribers. The company is publishing its results as they emerge.

Key Takeaway: Paid acquisition with sites like Care2 helps to gain new leads at scale while keeping costs consistent.

VTDigger

Who: VTDigger is a statewide news website that publishes watchdog reports on state government, politics, consumer affairs, business and public policy in Vermont.

Hypothesis: If VTDigger launches self-serve portals for user-generated content announcements then it expects the services to generate $100,000 in the first year and $250,000 annually in subsequent years.

Test: VTDigger built a system for users to submit obituaries,  press releases, and classified ads.

Results: VTDigger launched the obituaries and PR portal and it is currently building the classifieds sections. The site decided to test the obituary product for free initially with plans to monetize it in the upcoming months. From April through October 2019, VTDigger published 54 obituaries which garnered more than 125,000 pageviews. Fifty-five percent of those users who read obituaries stay on the site for at least one more pageview.

Hypothesis Result: Inconclusive, but promising. It took longer to build the PR Portal and Classifieds sections than anticipated. Based on the public’s response to the obituary and PR portal there is a clear demand for these services.

Key Takeaway: There is a market need for these services, but each of these portals requires investment in marketing and building relationships with businesses, funeral homes and other businesses and individuals who can share these opportunities with the community.

Read more about VTDigger’s self-serve portal experiments here.

Technical.ly

Who: Technical.ly grows local technology communities by connecting organizations and people through news, events and services in Philadelphia, Baltimore, Washington D.C., and Delaware.

Hypothesis: If Technical.ly leverages reader behavior data, smart content and opt-in culture preferences to match employers and job applicants, then it will build a new revenue model for news organizations based on job placements.

Test: Technical.ly refined its talent offering by enhancing the digital jobs and company board user experience. It then launched its “matching” program and conducted user testing on its Recruiter Dashboard. To do this it used reader behavior data along with applicants’ resumes to determine if there was  any predictable model for using data to match candidates to open roles.

Results: Technical.ly made three direct placements and earned $36,000 in revenue with its match program.

Hypothesis Result: True. However, there is still work needed to parse out the best data to match readers with potential jobs, but without a major investment of time or resources it will be hard to reach to a point where it will create an actual return on investment.

Key Takeaway: Even with a tool to score candidates, it still requires a human recruiter to vet, introduce and support through the search and placement process — a single job placement can take over 40 hours. Even with the tool it can still require sourcing outside of the database.

The Associated Press & Newsday

Who: The Associated Press and Newsday, the Long Island-based daily newspaper, have both been experimenting with data and automation in recent years.

Hypothesis: If the Associated Press can evaluate the availability, structure and scalability of multiple data types, then it will be able to understand the opportunities for automated hyperlocal content and the necessary steps to scale these types of offerings to other local news outlets.

Test: The AP collaborated with Newsday on Long Island to study the viability of leveraging AP’s automation expertise with local data to produce higher volumes of locally-relevant content that engages local audiences and adds revenue through new products.

Results: The AP worked with Newsday to use data from the 124 school districts on Long Island to automate stories to an education microsite. It created templates that would translate the data into short narrative briefs.

Hypothesis Result: True. The AP learned that there is data available and the technology is available to make this a reality. It also learned that over time this technology could save time for journalists, but it will require a significant investment before newsrooms start to see a return. Newsday said it expected it to ultimately take a year to make updates to train the templates to account for the multiple data scenarios. Additionally, the products and system created around a single dataset does not necessarily translate to another.

Key Takeaway: The key challenge of this is editorial, not technological as each template must reflect the news values and style of each publication. Additionally, the newsroom’s goal in investing in AI should be to assist their reporters, not to replace them.

Read more about the AP and Newsday’s experiment here.

SembraMedia

Who: SembraMedia is a non-profit organization dedicated to increasing the diversity of voices and quality of Spanish content, helping digital media entrepreneurs to be more successful and sustainable. It offers training in business and technology, market intelligence and connections between journalists and other social entrepreneurs.

Hypothesis: If SembraMedia creates an audience growth and engagement tool kit with the most popular tools for driving revenue through donations, events, services, and digital advertising to make them more accessible in Spanish, then Spanish-language news organizations will be more financially sustainable.

Test: SembraMedia helped local Hispanic publications in the United States and Latin America overcome one of their biggest obstacles: identifying the best tools, covering prohibitively high fees (especially when just starting out), and finding the knowledge, support, and training they need in Spanish to be successful.

Results: SembraMedia created various assets for Spanish speaking media makers including a journalism technology catalogue, online classes, and case studies. It gave grants to six media startups and provided over five hours of consulting with news organizations to develop a strategic plan and implement technology.

Hypothesis Result: Inconclusive, but promising. This work is still in process and whether the training and grants lead to return on investment remains to be seen.

Key Takeaway: The grantees are using an assortment of tools. Action Button enables news organizations to provide action items for their readers, such as a button that goes with political stories that makes it easy for readers to email their political representatives. Another grantee is using Jivochat, an ecommerce tool to serve the information needs of some of the most vulnerable immigrants by more efficiently answering questions across many different channels.

WBUR BizLab

Who: Boston-based WBUR’s BizLab is an innovation lab developing and testing new models of support for public radio. Operating as a lean startup, BizLab is a team of product managers, experience designers, and business analysts. It works with public radio stations to uncover new avenues for revenue generation, distinct from traditional approaches including on-air drives, direct mail, and underwriting.

Hypothesis: If WBUR explores product recommendations and reviews for newsrooms and expands how listeners can purchase items through affiliate programs then the newsroom will generate increased revenue.

Test: BizLab launched an affiliate marketing program in partnership with WBUR. It held a workshop to brainstorm potential product review categories, established ground rules for how this project would interact with editorial and published 11 product reviews ranging from women’s shoes to fresh vegetable delivery. BizLab tested different approaches, including reviewing different product categories, affiliate programs, writing style and audience targeting.

Results: More than 39,000 unique visitors went to the WBUR Guides website. However, the station only gained $203.01 in revenue with this project, which didn’t cover expenses.

Hypothesis Result: False. To be successful, an affiliate marketing program must be able to tap into a site’s naturally-occurring users. A paid audience (which was the main audience in this experiment) is unlikely to be profitable, given affiliate revenue involves thin margins and the fact that a paid audience is more expensive and less engaged.

Key Takeaway: In order to tap into news organization’s naturally occuring users affiliate marketing has to be baked into the editorial strategy. In contrast, WBUR show On Point used affiliate links for its summer reading list, which attracted 27,000 unique visitors and generated nearly $2,000 in affiliate commissions.

Read more about WBUR BizLab’s experiment here.


Leave a Reply

Your email address will not be published. Required fields are marked *