The PR Agency That Learned to Say No

A publicist is usually hired at the precise moment a founder has become tired of being ignored. The product works. The customers are pleased. The financing deck contains a diagram with several arrows and one heroic circle. Still, nobody outside the company seems to care. HeadStart built its business in that awkward gap between doing something useful and being known for it.

The Washington communications agency opened in March 2021 with a broad promise: help startups tell their stories. Broad promises are comforting when a firm is new. Every company can be a prospect, every sector another invoice. They are also expensive. A publicist representing chicken coops on Monday, health insurance on Tuesday, and recruiting software on Wednesday must relearn the map each morning.

So HeadStart changed the map. It narrowed its practice to three labels: Health, Work, and Growth. Health covers digital care and the machinery around it. Work covers recruiting, HR technology, and the changing workplace. Growth is the connective tissue - venture firms, accelerators, corporate innovation programs, and startups with a serious next chapter to announce. The categories overlap enough that knowledge compounds. The agency can recognize the investors, reporters, arguments, and clichés before the kickoff call ends.

A young agency's most valuable client may be the one it finally decides not to chase.

The thing for sale is judgment

HeadStart's menu is long: media pitching, interview preparation, crisis communications, press releases, opinion pieces, social calendars, brand messaging, visual identity, websites, even paid opportunities. But the interesting product is the choice among them. The company says every engagement begins with at least two one-hour discovery calls. Strategy generally takes three to four weeks. Only then does the pitching, posting, writing, or rebuilding begin.

That order matters. A company asking for coverage may actually have a positioning problem. A founder asking for LinkedIn posts may have no useful point of view yet. A startup asking for a splashy launch may possess a perfectly good product and a perfectly ordinary announcement. Former journalists on HeadStart's team are meant to detect the difference. The agency's blunt formulation is that people will not care simply because a founder believes they should.

Public client accounts make the operating model unusually visible. Bellwether founder James Frawley described a team containing an account lead, writer, social-media manager, and general manager, with near-daily email and biweekly planning calls. He reported spending between $25,000 and $50,000 from February 2022 onward. Other disclosed projects fall below $10,000 or inside a $10,000-to-$49,999 range. There is no public rate card, because there is no single package pretending a website rebuild and an executive-opinion campaign are the same object.

Count the conversation, not the clippings

PR agencies have long enjoyed counting what they can easily collect: mentions, links, impressions, perhaps a suspiciously large dollar value assigned to a newspaper paragraph. HeadStart favors share of voice - how much of the relevant public conversation belongs to the client compared with competitors.

For executive recruiter Creative Talent Endeavors, the agency supported the launch of Fairantee, an algorithm designed to help price roles fairly. It combined media work with a website audit, social strategy, copywriting, and marketing consultation. HeadStart says placements followed in Business Insider, Reuters, Fast Company, the Financial Times, Newsweek, and HR Brew. More tellingly, CTE's share of voice against a leading competitor reportedly moved from 19 percent to 53 percent in three months.

Healthworx required a different sequence. The healthcare innovation organization had a parent brand, accelerator, venture arm, and funds that needed to look and sound related. HeadStart and a design partner recast the identity and messaging, then built sites and ran earned and social programs. The agency reports that applications to the accelerator increased by roughly 50 percent. First make the family recognizable. Then introduce it to the room.

The newest reader is not a person

HeadStart has recently added another audience to its mental model: the large language model. Its argument is practical. Chatbots assemble answers from sources they trust, often established publications and review sites. If a brand is absent from those sources, it may be absent when a prospective buyer asks an AI system for the important companies in a category.

For The Continuity Company, which provides temporary leadership support when founders take extended leave, HeadStart examined how models answered questions about that problem. The systems recommended fractional support but named no provider. Their citations leaned toward established business publications. That observation changed the recommendation: prioritize credible earned coverage over a social-first campaign.

The clever part is also the restraint. HeadStart does not propose that a chatbot write the campaign. It uses models as a discovery audit: repeat the same prompts, record which brands appear, inspect the cited sources, track sentiment and language, then let humans decide what story is worth reporting. The method turns AI from an inexpensive ghostwriter into a somewhat eccentric focus group.

Urgency, on a four-day week

There is a pleasing tension in HeadStart's culture. PR runs on somebody else's deadline; HeadStart advertises a four-day workweek. Its careers pages also promise remote work, unlimited paid time off, healthcare, professional development, and room for medical appointments, family obligations, and mental-health days. The company calls itself work-to-live in an industry that can make every reporter email feel like a fire alarm.

This is not separate from the strategy. Constraints make selection unavoidable. A small team on a shorter week cannot manufacture enthusiasm for every middling announcement. It must know which clients fit, which ideas travel, and which channels deserve the hours. The culture and the specialization are versions of the same decision: attention is finite, including the agency's own.

The model works best when a client contributes the raw material - access to leaders, defensible evidence, a willingness to hear that the original angle is weak, and enough patience for discovery before coverage. It is less persuasive for a company seeking a single viral moment, a substitute for product-market fit, or a guaranteed placement on command. Earned media retains the inconvenient property embedded in its name.

HeadStart's first promise was to move companies from obscurity to acclaim. Its more mature proposition is quieter and more useful. Know a few consequential worlds well. Help smart people say something worth repeating. Measure whether the conversation actually moves. Then, once it does, do not waste the proof.