Voice

an experiment in direct democracy

You can do a lot with your phone.

Every day you decide what to like, what to share, what to buy.

But when it comes to politics, it feels like shouting into the void—waiting around for next November for change.

Why can't you be heard by politicians as often as you're heard by your friends, followers and advertisers?

Why can't your perspective be felt by politicians every day they're on the job?

We live in a democracy. Let's make politics as fast, easy and responsive as the rest of our lives.

Let's do an experiment.

Voice is an interactive news app where you rate your politicians on breaking news, fueling a real time, highly-public political barometer that can't be ignored.

  • All your politicians in one place, local to national - and clear, nonpartisan, breaking news about them. Your newsfeed is about your politicians — your mayor, your governor, your president.
  • As we rate, we're building a livestream of voter sentiment that is moment-to-moment and issue-to-issue. You can learn, rate, and watch as your politicians are rated, every day.
  • Voice broadcasts ratings in the news, social media and directly to your politicians. Ratings drive the conversation – they can't be ignored.

Voice has one mission – strengthen the feedback loop between voters and politicians for a more modern, responsive democracy.

So, let's do something new with our phones.

Let's vote every day.

Thanks for joining.

- The Voice team

To achieve this vision, we follow some important rules.

(tap to jump to a section)

Politics + Editorial

Gathering Data

Sharing Data

Politics + Editorial

Small-d democracy

In a democracy, informed citizens and responsive institutions are the core drivers of social progress. Each citizen knows best how society should improve for them. No one is qualified to make that decision for them.

So Voice must be a nonpartisan resource that never veers to the left or right. We give you the information you need to make decisions and make sure they're loud.

That said, politics is about persuasion. Out in the world, Voice ratings will likely be spun or misrepresented to fit political narratives. And yes, of course the folks that build this app have their own perspectives.

But impartiality is existential for Voice - the company only works if it remains a credible source of political information and data. If voters, the media, or politicians believe Voice's content is biased or data is manipulated, they will stop listening, and Voice ratings will stop leading the conversation.

Voice is a place for “small-d democracy,” for better or worse. The idea will simply not survive otherwise.

Editorial + Content Curation

Voice's editorial philosophy is that political content should be clear and actionable - not noisy, misleading and pointless.

Voice is not another social algorithm or media outlet that tells you what you want to hear so it can sell your attention to advertisers. Voice doesn't do “doomscrolling.” That's not how our business works ( more on the business model below ).

That said, Voice is not a news organization or media company and is not attempting to fix the many problems of bias or conflicts of interest in this industry. Sensationalism and biased persuasion is inevitable in the news business.

But with Voice, you can cut through the noise, hear the different personalities and biases equally, and be the judge yourself. Voice simply packages content in a fair way, and shares your reaction, whatever it is, with the people who need to hear it.

Voice has a very simple curation system that prioritizes three things:

  • Actionability: You'll mostly see content about your politicians, so that you are rating them as their constituent. That means when you rate them, it matters.
  • Simplification: You don't need to read 50 news headlines and 30 social posts to understand what your politicians are up to. Voice uses AI and human editors to provide a four-point summary for each political “moment” which you are asked to rate.
  • Citations: Voice gathers credible political content from across the internet, including news outlets, podcasts, and social media, and will always cite sources so you can dive deeper.

Gathering Data

Personal Data

To use Voice to its full potential, Voice asks for just two details about Voters:

  • Email address - Simply so we can stay in touch. When your ratings have made the news, or your politicians saw and responded, we'll email you. When you want to log-in and see your ratings and those of your community, you'll use your email.
  • Physical address - Voice uses your address to populate your app with the politicians you vote for, and give you relevant news about what they're doing. This is crucial because your ratings matter 10x more to your politicians if they know you're a real constituent.

Ratings Data

When Voters rate their politicians, Voice gathers those ratings, anonymizes them, and correlates them with voting districts (like city, county, state), so that the data is insightful to politicians in those districts, as well as journalists, academics and you.

Then we make the data publicly available. Here's how we publicize ratings data to different audiences:

  • Voters like you: As a user of the app, you'll be able to see ratings data for your politicians on the Politicians + Sentiment screen. We're always adding interesting ways to parse the ratings, so you can see how ratings change across issues, regions, and over time.
  • Media: Voice publicizes ratings with journalists and on social media so that your ratings are front-and-center in the political conversation.
  • Politicians: Voice makes ratings data available to politicians according to our business model (below).

A note on trolls and bots

Tolls

Voice's philosophy is the more data, the better. Trolling, swarming, and annoyingly loud people would indicate that it matters what Voice ratings show. That's politics!

But if you sign-up and add your address, your ratings are much more important. That's because politicians and other Voters know you're a real constituent that votes, and can filter out the noise.

Bots

Voice has strong protections against bots, including reCaptcha and VPN detection. The ratings you see on Voice will rarely come from bots.

Sharing Data

Business Model

Voice makes money by licensing real-time, anonymized ratings data - the most dynamic, detailed, and abundant source of political sentiment data available to politicians and the media.

Politicians, journalists and academics are invited to license Voice datasets in support of their work.

To be crystal-clear: Licensors of Voice data will never have access to Voters' personal identifying information. In fact, Voice doesn't even collect Voter's names, just emails and street addresses.

Interested in licensing Voice ratings data? Reach out to kyle@voicenews.app

Data Ethics

Our ethical standards are based on the American Association for Public Opinion Research's Code of Professional Ethics and Practices.

  • We will never engage in data falsification or fabrication.
  • We will act in accordance with all relevant laws and regulations governing the collection, use, and disclosure of information obtained from individuals.
  • We will never knowingly make interpretations of results that are inconsistent with the underlying data.
  • We will store data in a secure, confidential environment.
  • We will disclose the methodology used to generate our own publicly disseminated research results (below).
  • We will promptly correct any errors that are brought to our attention. See something? Say something!

User ethics

  • We will avoid any practices that may harm, endanger, humiliate or mislead voters, politicians, the media etc.
  • We maintain that participation in our platform is entirely voluntary and can be revoked at any time.
  • We will never attempt to identify anonymous contributors without consent, nor disclose any information that could be used alone or in combination with readily available information to identify contributors with their data.
  • We will return the aggregated results of our general research with the contributing users.

Customer ethics

  • We will confidentially hold and store all proprietary customer information, except when the dissemination of the information is authorized by the customer.
  • We will be cognizant of the limitations of our expertise and capabilities to conduct various types of research and will accept only assignments we can reasonably expect to accomplish.
  • We will work together with media organizations to ensure consistent interpretations of our data.

Methodology

Voice's mission is to provide timely, actionable political insights to Voters, Politicians, the Media - anyone who wants to understand political sentiment as it changes day-to-day.

We follow these methodological guidelines to deliver transparency and common insight:

Public In-App Ratings Data

The “top layer” of Voice ratings data is public on the app's Politicians screens. Voters can see the impact of their ratings, how their politician's ratings change from moment-to-moment, and interesting things like how their ratings compare to others in their city, state and country.

  • All aggregate results (ie. rating percentages of specific Moments or Politicians) must have a sample size of at least 30 ratings. After passing this benchmark, we will show the average at 90% and 95% confidence intervals.
  • For in-app live data (as opposed to customer reports below), no weighting adjustment is made to the sample by population demographics unless otherwise stated.
  • All possible attempts are made to remove responses from automated agents (bots), by assessing signatures such as geographical location, and response timing.

Licensed Customer Reports

Voice Customer Reports are individual reports for a specific Customer, and as such will use various methodologies that best serve their specific research questions. Any specific methodologies therein will be described in context, especially as they deviate from the usual approaches below.

  • All reasonable attempts will be made to limit responses to the location of the voter demographic unless otherwise stated.
  • Where demographic weighting is used, it will be labelled and the particular methodology employed.
  • Demographic weighting is the procedure of up-weighting responses from under-sampled population demographics to match the true proportions of the target population. For example, if 50% of the target population is male but 60% of the respondents are male, female responses would be allocated a proportionally marginal higher weight to create the weighted estimates.
  • Post-stratification weighting may also be employed and noted, to weight or adjust a sample where appropriate in order to obtain a more representative result of the assessed demographic, using a methodology known as multi-level regression and post-stratification (MRP).
  • Quality control of research outputs will be employed to remove responses from automated agents (bots), by assessing signatures such as geographical location, device platform, and response timing. This methodology will adapt as threats to data integrity develop and change.
  • Sample sizes of aggregate reports will be clearly noted.
  • Standard statistical software and methodologies will be employed and noted, and custom statistical procedures will be made public.
  • Responses will not be as a result of recontact due to a particular response (ie. positive or negative #) to a previous poll, unless explicitly noted.