Mastering HCP Targeting Best Practices

Most digital advertising is an exercise in scale, the wider the net, the more likely you are to find the person you want. HCP marketing inverts that logic entirely. For a specialty drug, the universe of relevant prescribers might be a few hundred rheumatologists or a few thousand oncologists nationwide. The job is not to reach millions cheaply; it is to reach exactly the right clinicians, repeatedly enough to matter, without burning budget on anyone else. That requires a different mental model, a different data stack, and a different optimization discipline than almost any other programmatic category. For the broader foundation of how pharma programmatic works, start with The Ultimate Guide to Pharma Programmatic Advertising.

Why HCP targeting is different

Consumer targeting tolerates a degree of waste. If a household-goods campaign reaches five people who will never buy the product, the economics still pencil if the price per incremental customer is low enough. HCP media does not have that buffer. A physician whose prescribing behavior is irrelevant to your brand is not a rounding error, in a small target universe, wasted impressions meaningfully reduce in-target delivery percentage and inflate effective cost per reached prescriber.

The other distinction is the nature of the conversion. A physician does not click a banner and write a prescription. Influence accumulates across multiple touchpoints, awareness at a conference, a detail from a rep, endemic display at the point of care, a CTV spot during evening hours at home. Programmatic’s role is to hold frequency against the right individuals across those environments so that by the time the clinical moment arrives, the brand is already present in the prescriber’s consideration set.

This means every decision, list composition, channel mix, creative sequencing, reporting cadence, runs on the assumption that the audience is small, high-value, and worth the additional precision overhead.

Deterministic NPI targeting

The National Provider Identifier is a unique ten-digit number assigned to every licensed healthcare provider in the United States. It is the closest thing HCP marketing has to a first-party identifier: it is stable, publicly available, and unambiguous about who a person is and what specialty they practice.

Deterministic NPI targeting works by matching that list of named providers to digital identifiers, hashed emails, device IDs, IP addresses tied to known practice locations, through healthcare-specialized identity graphs. The match is validated: when the graph reports a match, it has confirmed that a specific NPI is associated with a specific digital touchpoint, not inferred it from browsing behavior. That is the critical distinction from probabilistic HCP audiences, which use contextual signals and lookalike modeling to estimate that a given user might be a physician. Probabilistic methods have their place for extending reach, but they should never anchor a plan where in-target delivery is the primary success metric.

Deterministic targeting also enables specialty- and account-level precision. Rather than buying “all cardiologists,” you can target interventional cardiologists at health systems with formulary influence, or general practitioners within a specific regional payer network. That granularity matters enormously for brands where formulary access or account-level dynamics shape prescribing behavior. The first-to-market HCP YouTube targeting program I helped execute, which secured a $60K Google partnership credit, was viable specifically because deterministic NPI matching made it possible to validate that physicians were actually being reached in a non-endemic environment.

Building the target list

The NPI list is a starting point, not a finished product. How you segment and prioritize that list determines whether your campaign delivers against the accounts and prescribers that will actually move volume.

• → Specialty segmentation: Filter by the exact specialties relevant to the indication. For a branded drug with a narrow label, this might be a single specialty; for a primary-care product, it could span internal medicine, family medicine, and general practice. Resist the temptation to over-broaden here; each specialty you add without a strong rationale dilutes the plan.
• → Decile and prescribing-potential tiers: Claims data partners can enrich the raw NPI list with prescribing-volume tiers, often structured as deciles one through ten. High-decile prescribers typically warrant heavier frequency and premium placements; low-decile or new-to-brand prescribers warrant awareness sequencing before benefit messaging.
• → Account-based overlays (ABM): For hospital-influencing or health-system campaigns, layer in account-level targeting so that every touchpoint reaching a physician is also reinforcing the system-level message. ABM overlays are especially useful when a brand’s formulary position is in flux and institutional pull-through matters as much as individual prescriber persuasion.
• → New-to-brand vs. current prescribers: Separate these audiences in your DSP. Current low-decile prescribers need nudging toward volume; new-to-brand physicians need foundational awareness and trial messaging. Serving the same creative to both populations wastes budget and dilutes measurement signal.

Channel selection

Once the list is built, the question becomes where to reach those providers. Channel selection in HCP programmatic is a tradeoff between match rate certainty, context relevance, and incremental reach.

• → Endemic medical and point-of-care publishers: Platforms like Doceree, PulsePoint, and the endemic inventory within healthcare DSPs provide the strongest match rates and the most clinically relevant context. A physician checking a drug reference at the moment of prescribing is a different kind of attention than the same physician watching streaming video at 9 pm. Endemic placements earn higher CPMs but typically deliver lower waste and stronger engagement with complex clinical messaging.
• → Extended open web: After endemic inventory is saturated, extending to the open web through the same NPI-matched audience adds incremental reach against the same validated prescribers. This is where the balance between deterministic and probabilistic targeting becomes a live decision: use probabilistic extension carefully, with clear in-target percentage thresholds that trigger a pause before spend escalates.
• → CTV and online video for HCPs: Physicians watch television. Connected TV now supports NPI-matched audience targeting at meaningful scale, enabling the same deterministic precision in a lean-back, high-attention environment. The tradeoff is cost, CTV CPMs for validated HCP audiences run substantially higher than display, but for brand campaigns where emotional storytelling and recall matter, the premium is often justified. Video sequencing also opens the door to narrative arcs across the journey that static display simply cannot carry.

Frequency, sequencing, and creative

Frequency management in HCP campaigns is more consequential than in most programmatic contexts because the audience is small and the impression budget per provider is finite. Over-frequency burns dollars on a prescriber who has already been saturated; under-frequency means the brand never accumulates enough presence to move through the consideration funnel.

A reasonable starting framework is to think in weekly impressions per reached provider rather than campaign-total frequency. Aiming for a moderate weekly cap, often in the range of three to five impressions per provider across channels, gives the algorithm room to optimize while preventing the fatigue that erodes engagement rates and wastes the tail of your budget.

Sequential messaging compounds the value of that frequency. Rather than serving the same awareness creative throughout the flight, design a three-stage sequence: an awareness unit that establishes the brand and indication, a mechanism-of-action or clinical-differentiation unit for prescribers who have already been exposed, and a reinforcement or call-to-action unit for those who have consumed both prior stages. Most healthcare DSPs support sequential logic natively; the constraint is usually creative production, not platform capability. The cross-channel sequencing approach is something I used to help drive $2.1M in incremental investment across Sanofi vaccine programs by demonstrating that structured journeys produced measurably better downstream Rx outcomes than undifferentiated reach.

Optimization cadence

HCP campaigns move on a slower optimization cycle than consumer campaigns because the data feedback loop is longer. You will not see Rx signal weekly. What you can monitor weekly is the leading-indicator layer that predicts whether the campaign is positioned to convert.

• → Match rate: The percentage of your target NPI list that the DSP has successfully matched to digital identifiers. If match rate is below expectations, investigate whether the list format is correct, whether the identity graph has adequate coverage for this specialty, and whether supplementary data partners can fill gaps.
• → In-target delivery percentage: Of total impressions served, what share actually reached a validated member of the NPI target? This is the single most important efficiency metric in HCP programmatic. Falling in-target percentages often signal that the DSP is supplementing deterministic delivery with probabilistic fill, which may be acceptable at the margin but needs a defined ceiling.
• → Frequency distribution: Are impressions concentrating on a small subset of the target list, or are they distributing across the full universe? Heavy concentration suggests the algorithm is optimizing toward easy-to-reach providers rather than the full target; adjust bid floors and frequency caps to broaden distribution.
• → Pacing and waste signals: Weekly pacing checks should flag any publisher or placement delivering disproportionate volume at low in-target rates. Those placements are candidates for exclusion or bid reduction regardless of their raw performance metrics.

Measuring against Rx outcomes

All of the above optimization is in service of one question: did exposure to this campaign change prescribing behavior? The answer requires connecting media delivery data to de-identified prescription records, a linkage that platforms like Veeva Crossix make possible without exposing any individual patient’s health information.

The measurement framework begins at campaign setup, not at reporting time. The exposed and unexposed control populations need to be defined before the campaign launches so that any Rx lift can be attributed to the media rather than to baseline prescribing trends. Key metrics to track through the Crossix or comparable measurement layer include: reach among the verified target NPI universe, incremental Rx lift versus control, time-to-first-new-prescription among newly reached prescribers, and, the metric that most directly ties media to ROI, visit-to-Rx conversion rate. On a Sanofi vaccine program built on this discipline, that conversion rate reached 31.5%, a result that would have been invisible without the measurement infrastructure in place before a single impression was served.

For the full measurement methodology, including how to structure control groups and interpret Crossix output in campaign reviews, see How Programmatic Media Drives Rx Outcomes.