Frequency without relevance burns trust. Ad tech learned that, and life sciences is next

Dushyant SapreFounder & CEO, SwishX·9 min read
Contents
  1. Introduction
  2. Physician attention is scarce, and it is concentrating
  3. Ad tech already ran this experiment
  4. Life sciences is losing access, permission and measurement at the same time
  5. In life sciences, a wasted touch is a withdrawal, not an impression
  6. Earned frequency is the only kind that still works
  7. The case for volume, and where it breaks
  8. What commercial leaders should change now

Rep access is shrinking, consent is getting stricter and the metrics life sciences teams rely on are getting noisier. Ad tech has been here before, and its lesson is that when signals fade, the message is the only lever left.

Life sciences is about to relearn the hardest lesson ad tech ever learned. When attention gets scarce and the signals you use to find and measure it get weaker, volume stops working, and the quality of the message becomes the only lever you still fully control. Ad tech spent most of the last decade finding that out at great expense. HCP digital engagement is now walking into the same wall, from three directions at once.

I came into this industry after years in ad tech, and the pattern is familiar enough to worry me. Most commercial teams still plan digital engagement as a reach and frequency problem: how many physicians, how many touches, how many opens. I think that model is close to breaking, and the companies that notice first will end up holding a disproportionate share of the physician attention that is left.

Physician attention is scarce, and it is concentrating

Start with the supply. Veeva's Pulse data, drawn from field activity at most of the industry, showed US HCP access falling from 60% in 2022 to 45% in 2024, back to where it was before the pandemic. Half of the physicians who are accessible meet with three or fewer companies. In internal medicine, oncology, psychiatry and urology, nearly 30% restrict access to a single company.

That last number matters more than the headline. A shrinking supply of attention is one problem. A winner-take-most supply is a different one. A physician who sees three companies is not splitting her attention thinly across forty. She has chosen, and every company outside her three is competing for what is left, mostly through digital channels.

Those channels are filling up. IQVIA's inbox study for 2023 analyzed more than 150 million emails sent to US physicians in that year alone, and that is one vendor's sample, not the whole market. Its ChannelDynamics survey of HCPs in the US and the five largest European markets found that email reached a new high of 15% of HCPs' preferred channel mix in 2025, while face to face held steady at 40%. So physicians are more open to hearing from life sciences companies by email than they used to be, and they are receiving more of it than ever. Those two trends can coexist for a while. They cannot coexist forever.

Ad tech already ran this experiment

Programmatic advertising made each additional impression almost free, so the rational response to a hard-to-reach audience was to show it the same ad again. Retargeting, the banner that followed you around the web after you looked at a pair of shoes, was the logical end point. It worked until it didn't. People learned to stop seeing banners, and the industry learned, slowly, that past a certain point each extra impression has negative value. That is what frequency capping is really for: to stop paying to train your audience to ignore you.

Then the signals went. Safari began blocking third-party cookies by default in March 2020. Apple's Mail Privacy Protection, announced in June 2021, stopped senders from using invisible pixels to see who opened an email. Google spent years saying it would remove third-party cookies from Chrome, then reversed course in 2025 and, in October that year, retired most of the Privacy Sandbox technologies meant to replace them.

The contrarian read of that last episode is the useful one. The cookie never died in Chrome, and it hardly mattered. The signal loss had already happened in the places that counted, and the companies that came through it well were not the ones that rebuilt tracking. They were the ones with direct relationships with their audience and something worth looking at. When you can no longer follow people around, you have to give them a reason to come to you.

Life sciences is losing access, permission and measurement at the same time

Ad tech depended on three signals: access to the person, permission to target them, and a trustworthy read of whether it worked. The industry's version of all three is weakening together, which is what makes this moment different from the gradual decline in rep access the industry has lived with for years.

Access is the Veeva data above. Permission is mostly a European story. Under the ePrivacy Directive, promotional email to an individual generally needs prior consent, and GDPR governs the data behind it. The Commission's attempt to replace the directive with an ePrivacy Regulation was abandoned in February 2025, so the 2002 rules and their national versions still apply. The Digital Omnibus the Commission proposed in November 2025 would move cookie rules into GDPR itself, and as of November 2026 it is still being negotiated. Whatever comes out of that process, the basic fact for a European brand team will not change: consent is the channel. A physician who withdraws it is a closed door, and no amount of budget reopens it.

The US has no consent regime of that kind for HCP email, but the inbox has started to impose one. Since February 2024, Gmail has required anyone sending more than 5,000 messages a day to keep their spam complaint rate below 0.3% and to offer one-click unsubscribe. In effect, the mailbox provider now runs a frequency cap on your behalf, and it is measured in complaints.

Measurement is the quietest loss and possibly the most damaging. Because Apple Mail can load an email without the reader ever seeing it, opens get recorded that never happened. IQVIA itself says open rates are not as reliable as they once were. Many commercial dashboards in life sciences still lead with open rate, which means a lot of teams are optimizing toward a number that partly measures Apple's servers.

In life sciences, a wasted touch is a withdrawal, not an impression

There is one place where the ad tech analogy understates the risk. In consumer advertising, the audience is effectively unlimited. If you burn one person with too many impressions, there is always another. In life sciences the audience is a finite list of names. You cannot buy new rheumatologists. Every irrelevant message draws down an account of goodwill you will not be able to refill by spending more.

The data that exists points the same way. An Indegene analysis of about 7.9 million brand emails sent between April 2020 and August 2021 found that physicians were less likely to engage with brand emails delivered more than once a week, with engagement above 3% for brands sending 10 to 15 emails and about 2% for those sending more than 16. A Graphite Digital survey of 225 senior clinicians in the UK, US and Germany, published in February 2026, found that 65% had reduced or stopped engaging with pharma digital communications after poor experiences, 52% said they receive too many emails and messages, and 58% said the content feels repetitive or irrelevant.

None of these studies is perfect, and I would not put much weight on any single percentage. But they describe the same mechanism ad tech found: frequency without relevance does not just waste money, it lowers the response to the next message, including the good one. We have written separately about why physicians tune out the same email, again. The point here is economic. Trust compounds in both directions, and in a closed audience the negative compounding is permanent.

Earned frequency is the only kind that still works

The frame I find most useful is a simple grid with relevance on one axis and frequency on the other. Low relevance at low frequency is invisible: cheap, harmless and pointless. Low relevance at high frequency is burn, and it is where most broadcast programs end up. High relevance at low frequency is underplayed: the brand has something worth saying and says it once. High relevance at high frequency is earned frequency. The physician hears from you often because each touch answers something she actually needed.

Earned frequency has a hidden requirement. You need a steady supply of distinct, approved, genuinely useful things to say. That makes it a content problem before it is a media problem. Veeva's 2025 Pulse analysis found that field teams use content in fewer than half of HCP interactions and that nearly 80% of approved content is rarely or never used, while content-driven engagement more than doubles treatment adoption and shortens the time between meetings by up to 25%. The industry does not lack content. It lacks the right piece at the moment a physician has the question.

Timing alone shows how much relevance is worth. IQVIA's 2025 data put open rates for triggered emails, sent in response to an event or an action, at 33%, against 25.3% for broadcast. Even allowing for the noise in opens, a message sent because something happened beats a message sent because it was Tuesday. Veeva's own report put it plainly in 2024: the new models prioritize relevance and timeliness over traditional measures like reach and frequency.

The second-order effect is the one I would watch most closely. If relevance is the lever, then the real frequency cap is no longer set by the media plan. It is set by how many different, approved, accurate messages a brand can produce and keep current. Medical, legal and regulatory review becomes part of the engagement engine, and a brand's capacity to be relevant is only as large as the library its reviewers have cleared. That is also why orchestration fails without content behind it, a point made well in our piece on omnichannel as one conversation.

The case for volume, and where it breaks

The best argument for volume deserves a fair hearing. A physician cannot prescribe a drug she has never heard of, and at launch, reach matters. Volume is also measurable, and relevance is hard to prove in a quarterly review. A team that sends more and reports more opens looks productive. A team that sends less and claims each message mattered more has to argue for it.

I accept the launch point. Awareness has to come before preference, and there are moments, a new indication or a safety update, when you should reach everyone quickly. Where the argument breaks is in the long middle of a brand's life, when the physician already knows the drug and the next message has to earn its place. Reach without a reason in that phase is how brands end up in the burn quadrant without noticing.

Where I am less sure is what generative AI does to all this. It could make relevance cheap enough to deliver at scale, a version of every message fitted to the specialty, the patient type and the question asked. It could just as easily make noise cheap, and fill inboxes faster than before. A Genentech leader quoted in Veeva's 2025 announcement put it well: just because you can create more content doesn't always mean you should. My bet is that the tools will reward the teams that already know what is worth saying, and punish the ones that don't.

What commercial leaders should change now

Set a frequency cap per physician at the company level, across brands. The physician experiences your company, not your brand plans, and ad tech learned long ago that caps set per campaign do not add up to a sensible total. If three brands each think four emails a month is reasonable, the physician is getting twelve.

Give every touch a reason to exist: new data, a label change, an answer to a question the physician asked, a follow-up the rep promised. If a scheduled send has none of these, skip it. In ad tech terms, stop buying impressions you would not bid on if they cost real money, because in a closed audience they do.

Change what you measure. Replace open rate as the headline with actions that cannot be faked by a mail server: replies, requests for a rep or medical follow-up, content downloaded, time to the next meeting. Run holdout groups so you can see what engagement actually adds, the way performance marketers learned to measure incrementality once their tracking went dark. The same logic applies in the field, where call volume says the rep was busy, not that the call was any good.

Fund variety over volume. Treat review capacity and the approved content library as part of the media budget, because they set how often you can be relevant. In Europe, track consent retention as a core metric, since it is the size of your addressable market next year.

In ad tech, a wasted impression cost a fraction of a cent. In life sciences, it costs you the next one.

Filed underPerspectives

Keep reading

All articles

Want to see Magic Studio on your brand?

Video, visual aids, ads, approved email and web, made from your approved claims and ready for MLR review.