Swoop Built the Patient Target. Then It Bought the Patient Journey.

The useful thing about a pharmaceutical ad is not that someone saw it. It is that the right person learned something at the right moment, discussed it with a clinician, started appropriate treatment and kept taking it. Between impression and outcome lies a canyon of awkward handoffs. Swoop began on one rim of that canyon, building audiences for drug marketers. It is now trying to own the bridge.

The New York company uses de-identified health signals, machine learning and media infrastructure to help life-sciences brands find patients, caregivers and healthcare professionals across digital advertising, television, audio, social platforms and point-of-care channels. Its customers are pharmaceutical manufacturers and their agencies, not the people scrolling past the ads. Swoop says it works with the top 50 manufacturers and more than 400 brands.

That description was once enough. It no longer is. Since becoming a standalone New Mountain Capital company in 2024, Swoop has added MyHealthTeam's condition-specific social networks and Nimble's prescription platform. The pieces now run from predicted health need to clinician messaging, peer community and the pharmacy counter. “Omnichannel” is a word marketers have rubbed smooth. Swoop's version is interesting because it increasingly includes what happens after the media plan ends.

01 / The original machineFind the patient without naming the patient

Healthcare advertising has an unpleasant design problem. A marketer wants precision because a rheumatoid-arthritis message is wasted on most of the population. A person quite reasonably does not want an advertiser holding a folder labeled “probably has rheumatoid arthritis.” Swoop's commercial identity sits inside that contradiction.

Its DTC products create audiences of likely patients and caregivers. HCP Pro extends the work to clinicians, using prescribing behavior, referral networks, clinical interests and peer influence rather than treating every name on a National Provider Identifier list as equally valuable. Predictive Audiences look for likely milestones - an undiagnosed condition, a specialist visit, a therapy switch or future non-adherence. TV Audiences carry those segments into linear, connected and addressable television.

Swoop says the underlying health data it receives is de-identified and irreversibly encrypted. Its “Zero-Knowledge Methodology” is designed to build audience definitions without passing personal health information or conventional consumer attributes into activation. The company also points to HIPAA, newer state laws such as Washington's My Health My Data Act and certification by the National Advertising Initiative.

That does not make the privacy question disappear. It relocates the burden to methodology, contracts, security controls and legal review. “HIPAA compliant” is not a universal hall pass: state consumer-health laws can cover data and companies beyond HIPAA's traditional perimeter. A buyer should ask exactly which fields enter a model, what leaves it, how small audiences are suppressed, how identity resolution works and how consent travels across the stack.

“What we lack is that ability to directly communicate with the patient and give them information that can help them move along in their patient journey.”Ron Elwell, Swoop co-founder and CEO

02 / What broke firstThe handoff after the prediction

Swoop has not published a tidy failure story. The more revealing admission is a limit. After the Nimble deal, Elwell told Fierce Healthcare that Swoop could predict non-adherence, undiagnosed conditions and upcoming specialist visits, but lacked a direct way to communicate with the patient. In other words, the targeting engine could see around a corner while the company itself had to stop at the curb.

That gap appears to have changed the build-versus-buy calculation. In January 2025, Swoop acquired MyHealthTeam, whose communities gather people around more than 60 chronic and rare conditions at the time of the deal; Swoop now advertises more than 80. Members trade practical advice, read medically reviewed information and meet others living with the same diagnosis. For Swoop's brand customers, the acquisition added an opted-in environment, first-party signals and a route to patient feedback. For members, the promise is better-timed information and support.

Then came Nimble in May 2026. Its service lets patients receive prescription notifications, handle payment, choose delivery or pickup and manage refills. The network covers more than 5,000 independent pharmacies and 16 million users in all 50 states. Swoop can now pitch a pharmaceutical client on awareness, community engagement, treatment initiation and adherence without pretending those are four unrelated jobs.

What did the expansion cost? Swoop did not disclose the purchase prices for MyHealthTeam or Nimble. Earlier financing is only partly visible: Dealroom records a $4.8 million Series A in 2011 and $8.6 million of debt from 2012 through 2014. W2O Group, now Real Chemistry, bought Swoop in 2021 for an undisclosed sum; New Mountain separated it into an independent portfolio company in 2024. Anyone presenting a neat price tag for the current platform is doing arithmetic with missing numbers.

03 / The software shelfAudiences, AI personas and web agents

Acquisitions are only half the catalog. Swoop Piper is a generative-AI research interface where marketers question synthetic audience personas about doctors, media habits, creative ideas and treatment journeys. The company describes characters named Karen, George, Max, Darren and Lindsey as digital focus-group members grounded in active audience data. The sensible part of the pitch is its limit: Swoop explicitly says Piper should not replace traditional focus groups. It is a fast place to expose weak questions before six people spend a year on segmentation and approvals.

Swoop's conversational agents and web tools address another dead zone: the brand website. They can answer approved questions, guide visitors through content and generate first-party engagement signals. The company also sells coordinated patient-and-provider activation, so a consumer message can be reinforced through the clinician's preferred channel. This breadth differentiates Swoop from a pure programmatic audience vendor, but it also puts the company against a sprawling alternative set: IQVIA, DeepIntent, PulsePoint, Doceree, OptimizeRx, specialist agencies and a pharmaceutical company's own CRM stack.

The business model is enterprise and opaque. Pharma brands and agencies pay for some combination of audience creation, activation, measurement, software access, community programs and managed services. Swoop does not publish list prices. That makes procurement less like buying a SaaS seat and more like assembling a campaign and data partnership: scope, channels, audience rarity, integration, measurement and compliance work all matter.

04 / The useful lessonCopy the map, not the jargon

The copyable move is to map where the customer's desired outcome dies after your product completes its narrow task. Swoop's old job ended with a well-built audience. The patient's job did not. Community, clinician conversation, prescription access and adherence remained downstream. Buying MyHealthTeam and Nimble moved Swoop into those handoffs.

A smaller company need not buy two platforms. It can instrument the handoff, build a consented feedback loop and partner where ownership adds more regulatory risk than advantage. Synthetic research can be used for cheap iteration, then checked with real people. Performance claims should be framed with the denominator, study design and control group. Swoop reports a 29 percent year-over-year adherence increase in one campaign and 20 to 34 percent incremental lift across several actions in a TV partnership. Those are useful leads for diligence, not laws of nature.

Swoop's wager is that integration beats specialization when the outcome is treatment adoption rather than a click. The wager may produce better continuity. It also concentrates more responsibility in one company. The next chapter will be judged less by how cleverly Swoop predicts a patient moment than by whether the person on the other side experiences something useful, lawful and human.