The most revealing fact about New Blue Interactive is not that it uses artificial intelligence, programmatic advertising, universal IDs, automatic content recognition, or more than 200 data sources. Plenty of agencies own impressive nouns. The revealing fact is that a client once had no television campaign at all.
In 2024, Monroe Nichols was running for mayor of Tulsa in a three-way, nonpartisan race. His budget was bootstrapped. His opponents were on television. Nichols was not. This was the sort of constraint that normally arrives wearing a tasteful black suit and carrying bad news. New Blue treated it as a media plan.
The Washington agency built a digital-only campaign around search, YouTube, connected television, and a sharply defined electorate. Nichols needed Democratic voters, certainly, but also Republican support and Democratic women who could see significance in electing Tulsa's first Black mayor. The work won industry awards. More importantly, Nichols won the election.
The constraintThe campaign that could not buy television
A political consultancy can make nearly any result sound preordained after the ballots are counted. This case matters for the opposite reason: the missing piece was obvious before Election Day. No TV meant New Blue could not blur budget across the traditional media mix. It had to decide who mattered, where those people could actually be reached, how often they should see the message, and which creative could carry the weight.
Taryn Rosenkranz founded New Blue in 2011 after running finance and new-media programs at the Democratic Congressional Campaign Committee. The company now works across fundraising, persuasion, media buying, data, creative, texting, social media, influencers, ballot initiatives, and email deliverability. Its customers range from congressional and mayoral campaigns to state parties, PACs, unions, and nonprofits. The client list includes the DCCC, the ACLU, Brady PAC, EDF Action, SEIU affiliates, and the March on Washington Film Festival.
The fractureThe first thing to fail is the list
Political digital fundraising has a temptation built into it. An email costs almost nothing to send, urgency is easy to manufacture, and the reporting arrives fast. So campaigns send more. Subject lines grow louder. Deadlines multiply. Every news event becomes the most important news event of your lifetime. The short-term numbers may rise even as the audience slowly learns to stop listening.
New Blue calls that approach “churn and burn.” Its preferred phrase is “digital with dignity,” which sounds soft until you recognize it as a claim about unit economics. A donor who trusts the sender can give again. A supporter who feels used becomes an unsubscribe, a spam complaint, or a name that still appears in the database but no longer behaves like a person.
The old list fails in a second way. A voter file can tell a campaign who is registered, where that person lives, and some modeled likelihoods. It cannot, by itself, explain what someone is reading this afternoon, which streaming service has their attention tonight, or whether a climate message delivered by a familiar creator will land better than another glossy ad. New Blue's response was to treat the voter file as a starting point and layer it with consumer, behavioral, contextual, and first-party information.
How a signal becomes an action
The balancing actTwo hundred signals, one human voice
This is where the company fits in the market. A campaign can hire one shop for fundraising, another for advertising, a third for data, and a fourth for creative. New Blue sells the integrated alternative. Its advantage, if the arrangement works, is not merely convenience. It is the ability to notice that an acquisition ad brought in a certain kind of donor, change the email welcome sequence, adjust the media audience, and keep the candidate's voice intact without four firms passing a spreadsheet around.
The technology is substantial. New Blue says it invested $250,000 in automated reporting. Its NBI Network buys programmatic inventory across multiple demand-side platforms. Its donor modeling draws on more than 200 sources. Its connected-TV work has used automatic content recognition to reach households exposed to opposing ads. Yet the company returns, almost compulsively, to authenticity. The machine is there to decide who should hear the candidate. It is not supposed to invent who the candidate is.
That distinction explains some of the stranger, better details in the portfolio. An award-winning email carried the subject line “We have to talk to your dentist.” Another fundraising series borrowed the visual language of comic books because the candidate genuinely loved Marvel and bold clothes. The company used Twitter direct messages as a fundraising channel, and Arabic audio creative to speak both to Arabic-only voters and to a broader community that wanted to feel recognized. Specificity was not decoration. It was the targeting strategy made visible.
The receiptsWhat the numbers actually say
Agency case studies are selected because they went well, so they are evidence, not a universal forecast. Still, New Blue publishes enough figures to see the machinery. A 2021 program for Melanie Stansbury raised $500,000 digitally, reached more than 190,000 New Mexicans, grew its active audience by nearly 300 percent, and reported paid-acquisition return above 300 percent. A separate 2022 program started at zero and reached 127,000 subscribers, including 70,000 active members, with 130 percent return on ad spend within 60 days.
Indexed for display, not a like-for-like comparison. Each metric comes from a different campaign and objective.
In connected television, a Roku home-screen unit reached 852,000 households during a 2025 award-winning program. More than 20,000 people clicked a phone-number overlay and more than 2,000 texts were sent. Some ads and media markets posted video completion rates of 99.5 percent. Those figures do not prove persuasion on their own, but they show the difference between buying impressions and building an observable path from exposure to action.
The costs are less public. New Blue discloses that reporting investment and says it reinvests 10 to 20 percent of its operating budget in staff training and professional development. It does not publish a standard price card. The business runs like a private full-service agency: custom retainers and campaign engagements, with media, strategy, data, and creative assembled around the client. Texting is included in fundraising retainers, and pricing is described as flexible and scalable.
The transferable partA playbook small teams can borrow
What changed was not a single revelation in a conference room. The environment changed first. Inboxes saturated. Cookies receded. Voters acquired more screens. Static audience files aged faster. The firm responded by joining the channels together, investing in reporting, and making relevance the test for whether another message deserved to exist.
- Begin with the constraint. “No television” is more useful than “be everywhere.” It forces a choice about audience, channel, and purpose.
- Treat acquisition and cultivation as one system. A cheap new name is expensive if the welcome program turns that person off.
- Measure movement, not just delivery. Connect impressions to clicks, sign-ups, replies, donations, and votes where the data allows it.
- Use data to remove irrelevance. More signals should mean fewer wasted messages, not more excuses to interrupt everyone.
- Protect the principal's voice. The dentist line and comic-book series worked because they were particular. Generic urgency is not a personality.
There is an organizational corollary. New Blue is woman-founded, describes itself as a proud union shop, and says it invests a meaningful share of its operating budget in professional development. Political work arrives in brutal cycles. An agency cannot credibly promise long-term supporter relationships while treating its own people as election-season consumables. Culture, in this context, is part of delivery.
The conditions matter
A data-rich, integrated program will not rescue a candidate with no persuasive message, a client unwilling to test, dirty first-party data, or too little time for learning. It also loses its advantage when measurement is mistaken for causation. A completed video is not a changed vote, and a large list is not a community unless people keep opening, acting, and giving.
New Blue's bet is therefore narrower, and more interesting, than “technology wins elections.” Technology can make a campaign more observant. It can spot the people being missed, coordinate the screens they actually use, and tell a team when an idea is failing. The campaign still has to say something worth hearing. Respect is not a substitute for persuasion. It is the condition that gives persuasion a chance.