A New Approach for Understanding the Media Ecosystem
At Harmony Labs, our mission is to understand the social effects of media and to use this understanding to experiment with new solutions that serve the public good. In 2018, we saw a proliferation of early-stage startups studying media-related problems but recognized a gap in resources and entities focused on solutions. We felt compelled and well positioned to launch a program that provides mission-aligned, patient capital to entrepreneurs trying new things as a complement to our work supporting researchers focused on understanding the effects of media on society, politics, and culture.
In June of 2018, we launched a new 14-week media tech accelerator program and invested in a portfolio of companies to further this mission with a specific focus on startups linking media and civic action. In partnership with New Media Ventures and their annual Open Innovation Call, we reviewed over 300 applicants representing a wide cross section of new media startups. We narrowed in on 25 promising candidates and after two months of due diligence selected six exciting new companies to support. All the startups in our inaugural cohort are leveraging data, machine learning, and natural language processing to help clients be more effective across all aspects of the media making funnel from content creation and messaging to targeting, distribution, engagement, and analytics.
The first iteration of the program wrapped up in Fall 2018. And, as we begin to imagine the next iteration of the program, we wanted to share what each of these exciting startups has already accomplished since the program’s completion.
Please let us know if you’d like an introduction as an investor or client to any of our portfolio companies below by emailing email@example.com.
Article One is a unified cloud communications solution for government that aims to transform and modernize the way citizens and their representatives interact. They translate the exploding volume of incoming communications (advocacy, constituents, etc) to government (federal, state, local) into structured, actionable data — seamlessly integrating with legacy government technology so that elected officials can serve their constituents more effectively, responsively, and inclusively.
The problem they are solving is significant, with 97% of the public saying government is not responsive to them, and 94% of government staff saying they don’t have the tech to be responsive.
As of last fall, they have successfully piloted with 30 congressional offices, signed two congressional distribution deals with existing civic tech providers, were recognized by Twilio in their “Doers Hall of Fame” and were a featured provider by Amazon Web Services.
CivicFeed is an alert system that uses machine learning to monitor news and public data sources to discover and track changes in media coverage. Advocacy organizations, public policy professionals, and media use the platform to better understand news and government actions, and react in real-time.
CivicFeed was built in recognition of the fact that the proliferation of information is making it increasingly difficult for organizations to parse the noise, and existing processes are costly and time-intensive, while yielding results that are often too general and broad to be actionable.
To date, they’ve raised $450K in funding, launched their product, are tracking thousands of news sources and governing bodies in all 50 states, and are piloting with 10 organizations.
Countable is a platform that combines original and curated content with action opportunities, so that organizations, causes, companies, celebrities and media can power movements within their user bases. Their core mission is to lower the barrier to civic action.
With many organizations and brands struggling to find organic reach through platforms like Facebook, Countable aims to empower organizations to regain ownership of their audience relationships by embedding engagement directly into their existing communication channels via original content and the Countable Action Center platform.
The Countable platform is currently being used by brands like Starbucks, media organizations like ABC, advocacy organizations like GLAAD, and celebrities like Chelsea Handler. They recently signed their first $1M customer. Their work was recently featured in Forbes.
PredictWise is a data-driven, public opinion analytics solution that gives advocacy groups, political campaigns, digital advertisers, and companies access to deep, granular, continuously refreshing data, and algorithmically curated insights in real-time. Utilizing state-of-the-art polling methods, behavioral data, and analytic approaches, they allow their customers to address every part of an advocacy or marketing campaign funnel from topic selection to audience targeting to ROI.
PredictWise was born as a reaction to bad polling and missed predictions in the 2016 election. After co-founder David Rothschild correctly predicted election results in places like the NYT, he was inspired to turn his polling and prediction methodology into a product that could serve progressive campaigns, advocacy orgs and socially-minded brands.
Their solution was used by Elizabeth Warren’s campaign, by successful congressional campaigns like Katie Porter (CA-45), the League of Conservation Voters, and via a partnership with NowThis News launched ahead of last fall’s midterms.
Survey 160 is a software-as-a-service tool which seeks to improve traditional polling by using SMS messaging to reach audiences who are not responsive to phone surveys and are underrepresented in existing online panels.
As response rates have fallen from traditional outreach — especially including landline telephones — getting representative samples is increasingly challenging and expensive. Survey 160 supports organizations seeking to understand their audiences and ensures representative engagement for a fraction of the cost of traditional polling.
In the last year, they raised seed capital, completed and launched an MVP and delivered results for a range of paying customers in the political and corporate worlds, and sourced data that validated the core hypothesis around representativeness and costs. In 2019, they will be turning their attention to improved product features, especially making usage more turnkey for their clients, and growing adoption.
Swayable measures how effectively media content changes minds so that users can dramatically increase the impact of the stories they tell on the world’s most important issues. By utilizing randomized control trials and smart automation and machine learning, their tools allow for more accurate and efficient insights, delivered in hours in a way that is understandable by anyone, not just analysts and data scientists.
In most cases, large paid media buys are made with no upfront data on whether they will work and as a result 50%-75% of that spend is routinely wasted. Swayable was created to address these limitations and traditional methods like focus group, A/B testing and brand lift studies, to create more accurate data for advocacy groups and brands seeking to more effectively deploy ad dollars.
Through the end of 2018, Swayable was being used in over 500 tests across 65 races, by 30 partners, and their results were used to optimize more than $40 million in political advertising spend, on average more than doubling the impact of those resources. Swayable is now focused on gearing up for 2020, as well as supporting global campaigns and commercial clients, having already been deployed in 9 countries outside the US including Brazil, Mexico, France, Italy, Sweden, Austria and Poland.
Building Long-term Relationships of Mutual Value
In addition to providing critical early-stage support to these companies and furthering our understanding of media’s influence on civic action, the relationships we formed with each represents a long-term commitment — beyond the bounds of the 14-week program — to create public value by sharing data with our growing network of researchers focused on understanding the effects of media on society, politics, and culture.
These data collaborations are a unique aspect of our program and represent an exciting innovation in the work we are doing. In exchange for making important datasets available for academic research purposes, the participating startups are able to give back, further realize their missions to serve the public good, add credibility to their brands and reputations, and leverage the research insights from the Harmony Labs Research Network to inform their own product decisions and strategy.
We are excited about the possibilities that this new approach offers, in terms of better understanding the issues we face as a society, and also as a way of fueling the innovation that will lead to a healthier media ecosystem for all. Stay tuned, as we further refine our accelerator program model in 2019 and expand both our research efforts and our support for innovative entrepreneurs.
If you are interested in learning more about this program, visit harmonylabs.org/accelerator.