The Word That Cost Nine Points

In the language of climate policy, there are words that sound responsible in a conference room and ominous at a kitchen table. “Mandate” is one. “Ban” is another. “Phaseout” can be a third. Potential Energy Coalition tested such language across 23 countries and found that frames containing those words drew, on average, nine percentage points less support than frames that spoke about upgrading, setting standards, expanding access and cutting pollution. The policy had not necessarily changed. The sentence had.

This is the small, consequential territory where Potential Energy works. Founded in 2018, the New York nonprofit behaves less like a traditional environmental group than a hybrid of an advertising agency, a polling lab and a campaign war room. It maps audiences, runs focus groups, pits messages against controls, produces advertisements, buys media and watches what happens. Since 2018, it says it has conducted 160 analyses, gathered nearly 2,000 hours of individual feedback and served and measured more than 3 billion ads across research in more than 23 countries.

Its customers and collaborators are the people with something climate-related to explain: nonprofit leaders, policymakers, corporate communicators, academics, journalists and local coalitions. Its ultimate audience is everyone else - particularly the persuadable people whom conventional climate rhetoric has failed to reach.

The first product was a change of mind

The origin story begins with an accusation from a child. In 2017, according to founder John Marshall’s account, his son Henry came home from school and asked why his father was using a long career in strategy to sell shampoo and credit cards instead of working on climate change. Henry then, in the family legend, locked him in the house for two days. Marshall got the point. He went on to create Potential Energy with Harvard climate scientist Daniel Schrag.

Marshall brought the habits of brand consulting. Schrag brought the subject expertise. The interesting choice was what they did not bring: a preferred speech. In an early newsletter, Marshall described the organization’s role as turning competing theories about hope, doom, personal behavior and policy into experiments. The work would not begin with what insiders felt ought to persuade. It would begin with what moved people.

Nobody wakes up and says ‘Today is a great day for some decarbonization.’Potential Energy Coalition

What failed first was the sector’s default voice. Climate change was presented as vast, technical, far away and politically pre-sorted. The familiar images were melting ice and lonely polar bears. The verbs demanded sacrifice. The nouns belonged to experts. People could agree the subject mattered while placing it behind the dozen alarms already going off in ordinary life.

Research changed the brief. The organization learned that people responded to near-term human consequences, trusted messengers and the protection of people and places they loved. In its Later Is Too Late project with the Yale Program on Climate Change Communication and the Meliore Foundation, nearly 60,000 people evaluated multiple framings of 18 policies. A generational message lifted strong support for action by an average of 11 points and beat pitches centered on economic opportunity, justice, health and extreme weather in every country tested.

The scientists who sounded like neighbors

Science Moms, the organization’s best-known campaign, came from a segmentation exercise. Moderate suburban mothers appeared both movable on climate and influential in their communities. But the standard messengers felt political, the subject felt unsolvable, and the timeline felt abstract. The data suggested two kinds of speaker carried weight: scientists and “people like me.” Someone noticed the overlap.

The program recruited climate scientists who were also mothers, helped them speak in the media, built their social profiles and placed family-centered advertising. One film used the milestones of a child’s life to answer a parent’s blunt question: will my kids be affected? Another followed two atmospheric scientists and their daughters. The point was not to make the science less accurate. It was to give accuracy a human address.

Potential Energy reports that Science Moms sustained support among 5 million moderate suburban women and expanded the baseline pool of strong supporters by 25 percent, creating roughly 500,000 new supporters. In 2025, the campaign put a climate advertisement into the Super Bowl. Fast Company reported a 14-to-1 positive-to-negative social sentiment ratio and a 25 percent increase in support for climate action among focus-group viewers.

From frightening future to familiar present

The same machinery produced a different answer in Michigan. Potential Energy found that “movable moderates” liked clean energy in principle but imagined the transition as expensive, difficult and somewhere in the future. The opening was local proof. A campaign called Powered by Michigan used humor, recognizable state references and projects already operating nearby. It reached 3.5 million people. Among the target audience, strong support rose from 31 percent to 46 percent in three months, which the organization calculated as 320,000 new supporters.

For rural permitting, the answer changed again. “Permitting reform” sounded to many landowners like distant authorities taking control. After more than 200 doorstep conversations in one Michigan community, the organization reframed the issue around the farmer’s right to choose what to build. Its tests showed a 37 percent increase in support. For Pennsylvania union members, it stopped describing workers as beneficiaries of a transition happening to them and cast skilled labor as the people capable of building it.

These campaigns reveal where Potential Energy fits in the market. A polling firm can tell a coalition what people believe. An agency can make the film. An advocacy group can organize local voices. Potential Energy attempts the whole loop: research, narrative, creative, distribution, measurement and a playbook other groups can reuse. Some organizations that resemble competitors - including academic communication centers and climate nonprofits - are also its research and campaign partners.

What $36.7 million buys

Potential Energy is not an agency with a climate client list. It is a tax-exempt nonprofit whose 2024 filing reported $28.1 million in revenue, 96.8 percent of it from contributions. Program-service revenue was only $22,500. Expenses were $36.7 million. The filing does not disclose the cost of each campaign, so there is no responsible price tag for Science Moms or the Michigan work. It does show the scale of the machine behind them.

That donor-funded model changes the customer relationship. A local coalition or journalist can use a guide without buying a consulting engagement. A research finding can be published rather than held as proprietary client work. The organization can also spend heavily on paid education campaigns whose “return” is measured in support, action and policy room rather than sales.

The model has attracted both money and recognition. Historical supporters named in public reporting include the Hewlett, Packard, Grantham and Quadrivium foundations. The work has received an Effie for a social-issue campaign and an Analyst Institute award for analytics. In 2026, Fast Company placed Potential Energy tenth on its list of innovative nonprofits, citing the Super Bowl work and an estimated 1.9 million personal actions prompted by campaigns over five years.

A method, not a magic sentence

There is an obvious danger in a messaging playbook: today’s tested phrase becomes tomorrow’s dead jargon. Potential Energy’s own guides make the more modest point that different people see the world differently. A line that works for a suburban mother may fail with a union electrician. A message about national energy independence may travel badly to another country. A result from a controlled survey may not survive a hostile local debate.

So the useful thing to copy is not any single slogan. It is the sequence.

This method works when an audience is genuinely persuadable, when campaigners can research it before speaking, when the messenger has earned credibility, and when there is enough time and media reach to measure change. It becomes unreliable when communicators copy a winning line into a different culture, confuse short-term approval with durable behavior, or polish the language of a policy that people dislike on substance. Marketing can clear a path for policy. It cannot make every destination popular.

In July 2026, Eric Asche became chief executive after a planned succession. He arrived with a telling résumé: years at Truth Initiative, the organization behind the anti-smoking truth campaign. Marshall moved to Executive Chair, while former climate diplomat Elan Strait became Executive Director. The leadership change suggests continuity in the central wager - public demand can be built with the same seriousness companies bring to consumer demand.

The wager matters because climate language is often treated as decoration applied after the real work is done. Potential Energy treats it as infrastructure. A policy must survive a legislature, a budget and an election. Before any of those, it must survive a sentence. Nine points is a large toll to pay for the wrong word.