8 Lessons Learned from Political Startups
A look back on working to make a difference from 2009–2016
Since walking away from Brigade last year, I’ve been approached by a lot of bright eyed entrepreneurs who want to make a difference this election and beyond. I love sharing my war stories, but after a few dozen runs through, I’m sharing what I’ve paid years to learn from founding or being around political startups for the past several years. Enjoy, and respond with any questions.
1: Find a discrete problem.
My journey down the rabbit hole of mixing politics and tech began in December 2007, when I saw my first episode of Meet the Press with Tim Russert, an interview of Ron Paul. From that day until November 2008, I didn’t miss a Sunday. I didn’t want to miss anything in fact. The wide open primaries of 2008: Hillary, Obama, Biden on one-side; McCain, Romney, Huckabee, Paul, Giuliani on the other provided simply too many storylines to consume. Coming out of that election I created what looked like a Bloomberg terminal for political news to try and make sense of the blogs, tweets, conversations, and news being produced almost 24 hours a day.
For a designer, this was malpractice. I wasn’t really solving anyone’s problem but my own. Normal people don’t have a compulsion to read everything, and normal people aren’t this interested in politics (more on that later.) The trend in consumer news has been towards Buzzfeed, Huffington Post, large headlines and digestible or ‘snackable’ content, not what I built. Without a product manager providing me a goal, or a customer-centric design process framing a problem, I simply mocked up dashboards for months on end. I was just writing and writing, with no story to tell.
When I explain Capitol Circle now I call it an art project, because it was an exercise in self-expression. I still see examples of this today, as well the opposite: really huge “boil the ocean” type problems, i.e. government is broken / ripe for disruption, or citizens have become apathetic, that aren’t solveable by any one solution. I worked at one of those companies, and have heard pitches from many others that share a similar vision of a fully responsive and perfectly functioning democracy. To them I say: consider finding someone’s actual pain, and figuring out how to fix it. Ask, “what’s the worst part of your day?” Meet people where they are, not where they aren’t.
Organizer, for example, has taken an important problem for campaigns, the need to canvass, and made it easier for both the volunteers knocking on doors and the staffers who used to enter the data by hand. They’ve replaced paper walk sheets and manual data entry with a Google-maps style app that’s familiar to anyone that’s used a smartphone. It’s also one of the few modern campaign tools with a solid business.
The good that came out of Capitol Circle was the excitement that led to me wanting to talk to people about it, and each person I talked to led me to one or two others. For this reason I believe, most the time, sharing your idea is a good thing. One conversation made me the first designer-in-residence at a venture capital firm, affording me time to explore this space, and another to my co-founder for Votizen — a company we started in 2010. Another conversation was with John Dickerson, now anchor of Face the Nation.
John thought there was something in here that would be useful for a journalist, but at the end of the day I didn’t end up interviewing and focusing on this type of user more deeply, as maybe there were problems here, but I didn’t see much of a market at the time. Existing solutions like Tweetdeck, while not great, were good enough. I didn’t see getting to a 10x improvement — an oft-cited benchmark by entrepreneurs and investors for justifying the creation of a startup.
2. Find a market.
The business comes from both finding a problem, and a market (a type of customer that’s willing to pay.) When your product hits both, this is called product-market fit. Add ‘social good’ or ‘impact’ into this, and now you’re making your job that much more difficult. A lot of pitches I hear start from the impact perspective, because most entrepreneurs in the space are coming at it from a desire to do good and make a difference, rather than seeing a market inefficiency or colossal money-making opportunity. We see a broken system and want to fix it, but that doesn’t mean there’s a market. We all see huge problems, but most people I see aren’t breaking them down into reasonably sized pieces. I wrote an article on problems with our democracy and the six I chose were not only huge themselves, but this was an incomplete list.
Remember what value is: benefit over cost. Value propositions articulate both benefit and cost that a particular market (i.e. plumbers, lawyers) finds acceptable.
In other words, if the benefit doesn’t exceed the cost, there’s no value or worse, negative value. And cost isn’t always in dollars. We’ve all downloaded free apps only to abandon large sign up forms. Why? The benefits were unclear. The cost is your time, maybe your email address, phone number, privacy, etc. Benefits accelerate, and sign up forms are friction. Without enough acceleration, users grind to a halt.
There’s also the positioning statement, useful in marketing. It outlines your target customer, product, differentiators and market alternatives all in concise, clear, and compelling statement. It looks like this:
For [target customers], who have [some problem], our product is a [new product category] that provides [solution to the above problem]. Unlike [competitor/substitutes], we have [this key differentiator].
Value propositions and positioning statements are not only useful in communicating with customers, but necessary to see if you’ve found a market or not.
3. Not everything is a venture fundable corporation, consider other ways of organizing or ways to achieve your goal.
If someone complains, there’s a problem. If it’s painful enough that someone is paying money to solve it, you have a market. But how much? How many people? Size it up.
For example, let’s look at campaigns. They’re sexy, but 99% of them are tiny, and chronically underfunded. If you’re selling a technology product, you can’t say you’re in the billion-dollar campaigns market when most of the dollars are spent on direct mail or advertising. If you’re not starting a media company, mail house, or search engine, you’re not a part of this market. You get what money is left over. Staffers are also averse to taking risks on new tools. One reason being, campaign jobs are temporary, and any field director or fundraiser is going to have to explain themselves to the next campaign manager. If you’ve seen Moneyball, Art Howe is the campaign manager and you’re Billy Beane with the crazy way of doing things.
“I’m playing my team in a way that I can explain in job interviews next winter.”
Going all in on a brand new tool or tactic could mean never working again, while nobody gets faulted for buying ads. Those guys continue to get campaign jobs or contributor gigs on CNN, win or lose. Additionally, they don’t have time. According to Stephanie Hannon, CTO of Hillary for America, new problems and opportunities crop up constantly. She often instructs her engineers to time-box tasks, and move onto the another if it’s not finished in time. The election, unlike a release date, can’t move. If you’re not solving a critical problem, for 1/10th the cost of what they’re paying today, or 10x the efficacy, then it’s not going to work.
Remember, not everything is a large venture-fundable business. Maybe what you’re doing could be a small lifestyle business or side project, or maybe it works better as a non-profit or volunteer project. Or maybe you’re better off committing your effort to a project or a campaign that someone else started. A lot of similar ideas tend to crop up at the same time, and there’s a better chance of making an impact with a dozen people coordinating together, than a dozen people working alone. Even more elementary than finding a problem, is finding your goal. Why are you interested in politics, government or civics in the first place? What’s the impact you want to have? How much do you care about money? If your goal is too big, can you narrow it down?
4. Don’t be afraid to start small, and iterate towards your vision.
The genius of a visionary founder often isn’t revealed until they’ve gone many, many steps from where they started. Big visions aren’t easy, but the steps need to be sequenced properly, building off each other.
My Votizen cofounder’s first product was a Twitter account, the earliest example of a chatbot I know of, called 2gov that took mentions, responded to the user, identified the issue, the voter registration status and elected representatives of the sender, before sending it on to them via twitter and physical paper. He built his own 50-state voter file, learned how to code in Python, and used oDesk to fill in the gaps. He applied the minimum amount of money necessary to validate demand, learn, and to lower the risk for a potential investor. This is what lean startups, often misunderstood as creating ‘minimal’ products, do. Remember that your idea is 100% risk at the outset, no better than gambling at a casino, and maybe worse. Everything you need to do in order to raise money, should that be your goal, is to lower your risk to an investor.
Showing real traction on some small idea or product is much more effective at lowering risk, and offering value, than a large or complicated product with no traction — or returning users or paying customers. This is true for the seed round and the later rounds.
Tilt offers an example as well. While other companies in the space such as Change.org and GoFundMe explicitly focus on crowdfunding for social good, Tilt changed its focus long ago to daily, weekly, and monthly use cases for crowdfunding, like sharing bills at college fraternities. Tilt’s CEO, James Beshara, has a degree in developmental economics and started the company as a charity-minded tool from his experiences in his first non-profit job in South Africa. I have a hunch, that over time, they’ll have the largest impact, because in solving the more daily case, each successfully funded project on the platform launches at least two new ones within 90 days — and its social nature, means there’s a massive network of users building under this who know only Tilt for payments. James isn’t sacrificing his principles at all, he’s playing the long game. That payments app for keggers, or maybe the polling app for teenage girls, doesn’t look like much today, but if you’re a consumer-focused for-profit business, the road to impact must go through traction.
The original vision for Brigade, Sean Parker’s political startup that acquired my company Votizen and succeeded Causes, was huge, and a counterpoint to my belief. The original product plan called for three products, or pillars as Sean called them, at launch as the MVP. This was to be built from scratch by a company of sixty people and funded with an initial investment of $9.3MM. One of my product and life mentors, a founder of many successful social products, told me: if the product is too complicated, too soon, it is difficult to diagnose what to fix if things aren’t going well — because there’s too much that can go wrong.
A partner at a major venture capital firm never believed it would work in large part because historically it’s not how consumer Internet companies are made, unless there’s some kind of regulatory or major technically defensible systems to build which require larger teams. The vast majority of Internet companies start small, needing no more than 10 people to achieve traction. Figure out step 1, and the sequence. Sean is an incredible visionary, but we differed on the sequencing, and Brigade’s value to consumers.
I use the example of SpaceX when explaining the difference — their vision is to colonize Mars. One way to do it is the way Elon Musk is doing it: he started with reliably delivering commerical payloads into orbit, now it’s landing those rockets on drone ships in the ocean for reuse, and this will continue into the next decade. The other way would be to skip all that and just start building the Enterprise in an Iowa cornfield today, based on theory rather than data.
5. Raising money can be a bad thing.
While 2gov offers an example of starting simple and iterating to improve each successive step in our metrics funnel, we got off track with Votizen.
Once we raised $1.5M, we created a new product on the web, hired more engineers, and created an entirely new product with a new funnel and new problems to solve.
Our first mistake was we didn’t lever, we changed the entire experiment. The amount of money raised was enough for us to think we could start fresh with a better product, one that had an actual graphical interface, allowed people to create open letters and then get friends to sign on to lobby additional members of Congress. It made sense as an evolutionary step, but we didn’t raise enough money give us the runway we needed to iterate on this product, and get to a follow-on round of financing.
Without an exponential growth curve, as a consumer product, we wouldn’t be able to raise a follow-on round. We were encouraged by our largest investor to hire more engineers and increase our burn rate — our second mistake. Given our raise, the best course of action would have been to immediately raise more before there was additional metrics we would be judged on, or keep burn low to extend the runway and find traction.
Our third mistake was who we raised from: a traditional venture capital firm. If we had ideological money that was committed to the vision of Votizen long-term, this wouldn’t have been a problem. We believed our partner, Peter Thiel, was such an individual, but it was not his own money, and Founders Fund didn’t live up to their own manifesto. Sean Parker and Reid Hoffman have made these types of long-term investments in civic technology, regardless of immediate-term traction—they deserve to be celebrated for that. However, more typically this money isn’t an investment, it’s a donation to a PAC meant to bend U.S. politics to one side or the other. For example, Tom Steyer and George Soros on the left, and Sheldon Adelson and the Koch brothers on the right.
Votizen deliberately stayed away from ideological money, because the vision was an inclusive one and tied to good governance. Conventional wisdom told us not to alienate half of our potential users, but there’s challenges in non-ideological products. Even something as non-partisan as Vote.org was the target of a recent conservative hit piece. But the more important lesson I’ve learned is that it’s very difficult to build an brand or affinity for a non-partisan product. A brand or political organization needs to stand for something, and match the aspirations of its customers. Good government centrism just doesn’t inspire passion the way something aligned would. For proof look at Fox News, MSNBC, and the rise of Breitbart—Steve Bannon is likely to build a $1B media business out of the Trump campaign.
6. Elections are misleading events.
We invested 6 months building a new web product between September 2010 and March 2011, then only had 6 months to iterate on it, before realizing we needed to pivot, taking a chance on using the election to get traction, or face certain death in September 2011 sticking with lobbying. The election was the siren on the rocks for us.
These election cycles are killers on all political startups. You can revisit what type of company might grow out of your efforts after the election, but it will almost definitely look different. The election is a springboard, but totally misleading event for any startup interested in the political process.
–Eric Paley, Managing Partner, Founder Collective
Staring at a declining balance sheet on a product designed to support advocacy, we saw the shiny object that was the 2012 presidential election — where $2B was about to be spent, and when advocacy dollars tend to decline as there’s too much noise generated by the campaigns. So we launched a product for the candidates in the San Francisco Mayor’s race to get out the vote. We had advanced our voter matching technology from being able to only claim your own record, to seeing which of your thousands of Facebook, Twitter, and LinkedIn friends were and were not registered to vote. Our user was no longer an typical concerned citizen, but professional campaign staff or passionate volunteers willing to send lots of messages. By the end of the year we had fully pivoted, putting our vision of a consumer network on hold.
For this to succeed, which for us was demonstrating enough traction to raise a round at a higher valuation / lower investor risk, we’d have to either sell to a huge number of customers at a small price, or sell to a smaller number of customers at a large price. In either scenario, you’d need to show that the total addressable market, or pure dollars if you sold to every single customer that’s similar, is significant. So if you can only sell to Congressional and Senate races, that’s about 936 customers if you’re bipartisan. (Assuming 2 campaigns for each of the 435 seats in Congress and about 33 Senate races per cycle.) If you charge $100/mo, for 3 months during the general, that’s $280,800. Since that’s only every two years, cut that in half. That’s not enough to be interesting to an investor.
Worse, I know companies that have raised money and given their tools away, desperate for a pilot or case study. It’s dangerous because you’ve proven your market size is zero dollars, and can’t lower risk for future investment if you haven’t gained customers, or users. And remember, all your customers churn on Election Day.
7. Consumers don’t care about politics.
Consumer products are fun to make, consumer products are sexy, they can generate huge returns. They’re also incredibly hard to get right from a product design perspective, as consumers are diverse, fickle, have short attention spans, and don’t fall into one behavioral segment. Consumer social products tend to be based on a theory of the human condition rather than solving a utility-like problem from a paying customer. For every home run there’s thousands of failures.
Consumer for politics, that’s even harder. These products are created from a place of idealism and a desire to make our democracy work better, as opposed to delivering value or enjoyment to a broad audience.
“We love to joke that the way to build the platform to revolutionize politics is to not make it about politics.”
–Matt Mahan, Brigade
We tried it at Votizen and Brigade. So did Ruck.us, Popvox, and a host of other companies beginning with Voter.com in 2000. I don’t believe going straight after an obvious political app for consumers will have the type of impact the entrepreneurs who start these apps want to have: which is usually acquiring millions of users and getting a regularly engaged electorate. Do you want to sit and vote on bills? Every day? Congress gets paid to do that, we don’t. Apps need either daily engagement, or they need to be an indispensable utility that’s properly positioned in the mind of the consumer so it can be readily recalled when needed.
Change.org, the popular petition site, is often raised as an example of working civic engagement for consumers. Causes invented the social petition, but died when its mechanisms for viral distribution, wall-to-wall posts, were cut off by a changing Facebook platform, as the traffic was almost wholly transactional. Change didn’t rely on Facebook, but worked over the years to hone a model of customer acquisition and monetization. It’s the online equivalent of ambulance chasing, not the regular civic engagement that I and many others envision. There’s a team hunting for petitions and stories that could be viral, and once those are identified, there’s a large marketing / PR team to elevate those petitions on search engines, social media, and earned media. Once you sign one, you’re prompted to sign a number of others in rapid succession, and your information is sold to advocacy organizations looking for leads. Your name, address, and phone number can be sold over and over again. There’s a market for qualified leads from campaigns and non-profits, but apparently not large enough even for Change to be a sustainable for-profit business. They’ve recently launched forays into crowd-funding, but the problem with most crowd-funding products is the same as petitions, most don’t start or fund a second campaign. Causes struggled to earn significant revenue from a similar product, and scrapped it years ago. As for ballot guides, people only care to do this once every 2 or 4 years, right before the election, if at all.
8. Consumers see no benefit in political participation.
The way to get more people interested in politics or voting is to create a benefit for doing so. People vote if the town is going to build a nuclear plant in their backyard. SeeClickFix found traction allowing people to report potholes and broken signs in their neighborhoods.
Right now everyone seems dead set on driving voting costs down by making your ballot decisions easier with more information or friends opinions, or making registration easier. With vote-by-mail and mandatory registration in some places, we’re arriving at the nadir of cost. As for information:
“We’ve got 40 years of data about why and how people vote. The general consensus in both academia and political consulting is that information is not the problem.”
–Nicco Mele, Kennedy School of Government
There’s tons online. No trusted information was the raison d’etre provided to me by a pitch woman for Change Politics—a ballot guide app that was launched in January and shut down a month ago without so much as a blog post. I trust the editorial board of my newspaper given they have a team of investigative journalists looking into these issues. There’s not a hot searing pain for the average voter if they don’t vote for every school board candidate or dog catcher, because there’s zero benefit for most people no matter the outcome. My friend’s opinions on it do not matter at all to me. In theory it’s an amazing way to reach voters at an optimal time, but only once, which is interesting for candidates dollars, but I don’t feel there’s a sufficient consumer need there, and candidates need to spend on multiple touch points leading up to the election, not on a moonshot.
The problem that should be attacked is a perceived lack of benefit from casting a ballot, or in other words, “my vote doesn’t count,” or, “nothing ever changes.”
Now take a look at Pokemon Go. People are walking all over the place to get virtual rewards. If the company behind it, or the Clinton campaign, starts putting rare pokemon around polling stations, it’s going to do more for voter turnout than all the civic engagement apps combined ever have — it’s a non-cyclical business that just happens to have world-saving uses. Similarly, general purpose networks like Facebook, which started in a dorm to solve the problem of knowing who was single in college, has boosted turnout. Who would have thought? The “I voted” sticker works on the same principle of vanity and showing off that you did your civic duty. I don’t believe anyone really cares what their friends think about politics, but we certainly care what others think about us.
If you want to have an impact on the election, don’t constrict the funnel of your available users to a political app where normal people are going to steer clear. Getting people who already vote 100% of the time to engage with your app will yield no marginal value, getting people who do not vote and are not interested, will.
The technology companies benefiting the most off of the political market are Facebook, Twitter, Google, Snapchat, etc., not political startups. Digital directors are looking at where the people are, and few voters are using political sites. Those that are, tend to be more reliable voters, and if you have limited dollars to spend, you either need to turn out your base (i.e. Hillary with democrats), or pursuade undecided voters. The latter is becoming an increasingly endangered species, so cheap turnout is the way of the future. This is something Vote.org understands well, as the most important aspect of their product isn’t the registration, but the texts to people leading up to the election reminding them to go to the polls, known as GOTV.
If I were starting over today…
I would be wary of consumer-focused political apps.
If you’re thinking of it, it’s probably been done before. Ballot guides, bill voting, crowdfunding, gambling, petitions and open letters, polls, you name it. An idea is nothing without timing and execution: but what about your plan would create a different outcome? Why you? Why now? Have you read the wealth of free available research on civic engagement? What’s your problem and your market? My thesis is that none of these apps deliver value or benefits to a broad enough audience to get traction, and the political diehards you may get, are already voting/participating — so there’s little impact. I’ve heard the emotional adoption curve as a counterpoint, but you still need to provide value to a broad audience over time. Consider starting very small and levering up, say in one city; or deferring your vision of real world impact in exchange for traction on a daily use case, like Tilt did. Facebook, Twitter, Tilt all built networks first. Have that, and you can lever into impact. If I were starting something consumer today, I’d try something so crazy people would have to share it and talk about it. With the $9.3MM Brigade started with, I’d build 20 crazy ideas to test hypotheses, staying nimble, and learning quickly, rather than going all in on one aircraft carrier that can’t be turned.
I would talk to customers that are already working to make a difference every day.
Find real pain. Talk to organizers, activists, professional staff to find needs, pain points, and behaviors. What’s the worst thing they have to deal with on a regular basis? Would they pay for a solution? As my friend and cofounder David is fond of saying, until the check clears, the answer is ‘no’.
I would be careful raising money.
There’s no such thing as a free lunch. Understand both how money can help, and hurt you. A for-profit business may not be the answer. A PAC, raising ideological money, or a non-profit, raising social good donations, may be the way to achieve your goals. A cofounder and I launched November, an app for organizers to register voters really fast. There’s no profit motive / business, or expectation of consumer mass-adoption. I’ll be thrilled it registers one more voter, which it has.
Glenn Beck has been incredibly insightful this cycle:
It is remarkable to me that both parties are so out of touch with the American people, both Democrats, Republicans and Independents, we all feel the same way. Three things: I don’t belong to anything, nobody’s listening to me and I don’t have any levers that control my own life.
What does it mean? Where’s the opportunity?
Related is how Donald Trump won the GOP nomination with essentially no money or endorsements, blowing up most predictive models. Trump used a huge social media and celebrity platform to not only reach voters directly, but to exploit a sensationally biased media. How could Jeb Bush’s $130MM have an impact when he’s just playing in few minutes of commercial time in between wall-to-wall Trump coverage? At Votizen we said ‘outfriending’ would replace outspending. It’s unfortunate that Trump did in 2016 what we failed to do in 2012. The media elite has been savaging Trump for months now, but it’s having little impact. Is Trump a black swan event or is he the shape of things to come?
These are my lessons learned over the past six years, but, conditions are always changing. I wish you the best of luck in your journey to make a difference. I’ll be out there with you.