AI-Nudged Messaging Lifted Vaccination Uptake Across a Large Pharmacy Network
Behavioral messaging pushed vaccination rates higher in a large retail pharmacy program, with AI-powered nudges outperforming standard pharmacy outreach by more than 20%. The program reached millions of patients and paired behavioral science with individualized timing rules, a combination designed to do more than repeat reminders. It tried to match the message, the moment, and the patient’s own posture toward vaccination.
The resulting uptake matters because vaccination decisions rarely hinge on a single fact alone. Patients weigh convenience, prior experience, perceived need, and social discomfort, then respond unevenly to generic reminders that arrive too often or at the wrong time. A system that learns from individual signals can adjust not only what gets sent, but whether a nudge is sent at all.
Personalization Drove Uptake Across Five Vaccine Types
Lirio’s Precision Nudging technology was deployed across flu, COVID-19, RSV, pneumococcal, and shingles vaccination efforts. Each of those vaccines carries a different clinical context, different eligibility considerations, and different public attitudes. A single template rarely fits all of them well, especially when the target population includes people who are ambivalent, already vaccinated, or firmly uninterested.
The program found that 40% of patients reached through the system got vaccinated. Nearly half of those who ultimately received a vaccine did so after just two nudges. The fact pattern points to a short decision window for many patients: a small number of well-timed prompts may be enough when the outreach respects autonomy and avoids fatigue.
One especially important signal was the drop in unsubscribe rates, which fell 62%. In practice, that suggests the messaging was not simply louder. It was more tolerable. In patient engagement, tolerance is not a soft metric; it determines whether a communication channel stays open long enough to influence behavior over time.
Behavioral Science Shaped the Message Before the Algorithm Chose the Moment
The outreach strategy began with consumer research on changing attitudes toward vaccination. That research informed how the messaging was framed for the pharmacy network. Personalized content that reinforced patients’ right to make their own health decisions resonated regardless of whether people were pro-vaccine, hesitant, or undecided.
This framing is clinically relevant because resistance often rises when outreach feels coercive. A message that defends autonomy can lower psychological reactance, the tendency to push back when people feel pressured. In vaccination campaigns, reactance can be as important as information deficits. A well-designed nudge reduces friction instead of adding it.
The AI component then used individual signals to estimate who was likely to act and who was likely to opt out. That distinction allowed the system to decide when a message might help and when silence was the better choice. In a crowded communication environment, restraint can preserve trust while still moving people toward care.
A Different Model of Engagement for Preventive Care
The scale of the program gives the findings practical weight. Million-level outreach exposes a communications strategy to the same real-world barriers that blunt many preventive care efforts: message fatigue, low salience, and uneven readiness to act. When a system can improve vaccination uptake while also reducing unsubscribe behavior, it suggests the engagement model is becoming more precise, not merely more aggressive.
The broader lesson extends beyond vaccines. Preventive care often depends on persuading patients to complete an action that feels optional in the moment but important over time. Behavioral science offers a way to make that action easier by reducing cognitive load and matching outreach to the patient’s decision stage. AI then operationalizes that approach at scale, sorting through signals faster than a human team could manage alone.
The program also reflects a key shift in digital health: the goal is no longer just to send reminders, but to sequence communication intelligently. One patient may need an early prompt followed by a second nudge. Another may need no further outreach after the first contact. The value of the system lies in recognizing those differences before communication becomes noise.
What the Vaccination Data Suggests About Future Outreach
The 20% relative improvement over standard pharmacy messaging offers a concrete benchmark for health systems trying to improve preventive uptake. It also suggests that personalization can be effective without relying on invasive tactics. Patients were not simply bombarded with more messages; instead, the system attempted to read readiness and calibrate the response.
That approach may prove especially useful for vaccines with uneven public familiarity or mixed emotional valence, such as RSV, pneumococcal, and shingles vaccination. Those decisions are often shaped by age, risk perception, and previous experience with medical care. A generic reminder cannot account for all of that. A tailored nudge can at least try.
Even so, the program raises familiar questions about generalizability. A large retail pharmacy network offers reach, but it also has a distinct patient population and a particular communication infrastructure. Whether the same results hold in smaller systems, primary care clinics, or populations with different trust patterns is not answered here. The intervention also depends on the quality of the underlying signals, and AI only performs as well as the data feeding it.
The most concrete finding remains the same: targeted, autonomy-preserving messaging increased vaccination uptake, and it did so while reducing opt-outs. In a field where too many reminders are ignored, that combination is hard to dismiss.