Mobilizing KOLs and reference sites

# Mobilizing KOLs and reference sites

In the early 2000s, urologists at Henry Ford Hospital in Detroit began doing radical prostatectomies with Intuitive Surgical's da Vinci system, publishing their series and letting visiting surgeons watch. Within a decade most radical prostatectomies in the United States were robot-assisted and Intuitive's installed base ran into the thousands of systems. Almost none of that came from advertising. It came from a short list of named surgeons, and a short list of operating rooms, being credible and visitable.

That is the machine this lesson builds: KOLs (Key Opinion Leaders, the clinicians whose judgment other clinicians borrow) and reference sites (institutions that have made your technology routine and will let peers come look). The evidence itself is another lesson's object. This one is about the humans who carry it.

Why KOLs and reference sites decide adoption

Clinicians are trained to be skeptical, and the risk they carry is personal: patient harm, a malpractice file, a reputation among colleagues who will hear about the case at the next society meeting. Two questions run underneath every decision to change practice:

• "Who respected has used this and stands behind it?"
• "What serious institution has committed to it, and can I go and see it?"

There is also a procedural reason peer proof matters, and it gets underestimated in marketing plans. Before a surgeon uses a new device at a hospital, the medical staff credentialing committee grants privileges for it, and those committees routinely ask for documented training and a number of proctored cases. Proctors come from your reference sites. If no trained proctor sits within a day's travel of a new account, credentialing stalls for months whatever the sales team agreed. The peer-proof engine doubles as a supply chain for qualified trainers, and its capacity, not clinician enthusiasm, is usually what caps your growth rate.

All of this runs inside the disclosure and engagement limits the fair-treatment lesson sets out. Assume them, budget the compliance time, and never treat a relationship as a way to buy an implant decision.

Step 1: MapMapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → the influence landscape

Before you recruit anyone, you need a mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition →. Not "who is famous," but who moves opinion in your specific indication.

Build it across a few dimensions:

• Clinical authority. Publication volume, trial leadership (a PI, or Principal Investigator, runs a study), guideline committee seats.
• Procedure or assay volume. A prolific author who no longer operates cannot be a proctor.
• Reach. Podium presence, society roles, teaching load, fellows trained.
• Relevance to your claim. A famous interventional cardiologist is useless if your device is a peripheral tool.
• Sentiment. Curious, neutral, or already committed to a competitor.

Tiering helps: Tier 1 shapes guidelines and society opinion, Tier 2 moves a metro area or hospital network, Tier 3 is the high-volume local practitioner who is more accessible and hungry for a platform. Do not over-index on Tier 1. Rising stars are often the better early partners, and they answer email.

In research tools the mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → has a different shape. For Oxford Nanopore or 10x Genomics the person who moves a market is a PI whose lab publishes a method other labs copy, so influence shows up as citations, protocols posted on public repositories and workshop attendance rather than guideline seats. 10x, which sells single-cell and spatial instruments, has built its case largely on the thousands of peer-reviewed papers its chemistry appears in.

Veeva and H1 sell platforms that mine publications, trials and claims data for this; they sell the mapping, so treat their tiering as a starting point rather than a verdict. For a narrow specialty, an analyst with PubMed, five years of conference agendas and the public payment disclosures can build a usable first mapmapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.View full definition → in a fortnight.

Step 2: Recruit with value, not money

The rookie mistake is leading with a consulting fee. The clinicians who matter are motivated by the science, their own reputation, and patient outcomes. Offer a platform instead: early access to data, a role in a registry or an investigator-initiated study, authorship, a speaking slot.

Oxford Nanopore ran the cleanest version of this in 2014. The MinION Access Programme shipped pocket-sized sequencers to hundreds of labs that applied, on terms that were mostly about data sharing, and participants were free to publish whatever they found. What they found early was unflattering: per-read accuracy well below established platforms, reported openly in the literature. The company kept the programme running anyway. That same community then produced the work that made the platform's name, including real-time Ebola sequencing in Guinea in 2015 with instruments carried in hand luggage, Zika surveillance in Brazil, and the ARTIC amplicon protocols later used for SARS-CoV-2 sequencing at national scale.

A KOL who is unable to report a bad result is worth nothing to peers. Unanimous praise gets discounted on sight.

🎬 [VIDEO: "How Pharma and Medtech Work With Key Opinion Leaders" - youtube.com - a plain-language overview of KOL engagement models and compliance guardrails]

Step 3: Build reference sites, not just individuals

A KOL is a person. A reference site is an institution that has operationalized your product, which answers a different question: can a serious hospital run this at scale, in their workflow, with their staff and their turnover?

Turning an account into a reference site means the technology sits in protocols and order sets, the site tracks outcomes it is willing to share (complications, throughput, cost per case), and the team will take peer visits and blunt phone calls.

The form it takes depends on who buys. Intuitive designated high-volume centers where visiting surgeons observed live cases and took hands-on instruction, and required training before a system went live. Straumann faces a market of two-chair practices with no flagship academic hospital to visit, so its version is educational: it has backed the ITI (International Team for Implantology, founded 1980) for decades, an academic network whose study clubs and scholarship centers reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → tens of thousands of clinicians, which makes the "site" a study club evening and a training course. 10x Genomics uses certified service providers, labs that run its chemistry for other labs, producing revenue and an institutional proof point for buyers who cannot yet justify an instrument.

Concentration is the risk nobody prices. If two sites generate most of your published proof, you are one job move away from a hole in your evidence base, and champions do move: the institution keeps the logo, the individual keeps the following and often the case volume. Decide whether you are investing in a person or a department, and if it is a person, get a second clinician at the same site trained and publishing. A neglected reference site is worse than none, because it becomes a reference against you, delivered by someone peers already trust.

Step 4: Sequence the mobilization

Activating advocates before you have data, or dumping all your evidence into one conference, wastes the asset.

1. Evidence first. No data, no credible advocacy.

2. Engage Tier 1 quietly as advisors and investigators.

3. Build reference sites in parallel so there is somewhere to send the skeptics.

4. Publish. Peer-reviewed papers outlast any podium.

5. Take the stage, timed to the flagship meeting in your field.

6. Cascade to Tier 2 and 3, using the papers and the sites as the proof for the cautious followers the adoption funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition → lesson maps.

Cascading faster than you can train is the most expensive error in this play. Learning-curve studies of robot-assisted prostatectomy put the case volume needed before margin and continence results settle in the low hundreds. A surgeon twenty cases in, compared against a colleague's long open series, generates exactly the data your competitor will quote for the next five years. Gate the cascade on proctor capacity, not on quarterly targets.

Common failure modes

Buying voices instead of building them. Payments create disclosure obligations and suspicion, not conviction. Peers can look the payments up.

One-way relationships. Call only when you need a favor and engagement decays within two cycles.

Ignoring the detractor. Robotic surgery drew sustained academic criticism of its outcome claims from senior open-surgery authorities, and that argument was not won by rebuttal letters. It was won by the volume of published series and by surgeons who could show their own numbers. Engage the respected skeptic directly; converting or neutralizing one is worth more than adding a third supporter.

Forgetting your advocate is also someone else's. In sequencing and single-cell work, the same lab will benchmark two platforms in one paper. Read the draft, offer method corrections, and do not try to edit the conclusion.

Reference sites with no proof. A prestigious logo means nothing if the site cannot show outcomes or will not host a visit.

Treating compliance as a handover. Marketing co-owns these programs with medical and legal from the first invitation, not at contract signature.

Measuring whether it worked

• Leading: active relationships by tier, publications submitted and accepted, podium slots secured, sites onboarded, peer visits hosted, and proctor coverage (trained proctors per region against accounts sitting in credentialing).
• Lagging: adoption in follower accounts that met your KOLs or visited a site, sales cycle length, competitive switch rate, and how long a paper keeps generating inboundinboundA strategy that attracts prospects organically via valuable content (blog, SEO, social) rather than interrupting them.View full definition → questions.

The clean test: a follower says "I want to talk to the team at that hospital before we decide." Your marketing has gone invisible because the peer proof is doing the work.

Key takeaways

• Peers, not ads, drive clinical adoption. KOLs answer who respected backs this; reference sites answer whether a serious institution can run it, and supply the proctors credentialing committees ask for.
• Map before you recruit, and weight procedure volume alongside publications. Rising stars answer the phone.
• Recruit with a platform, not a fee, and protect the freedom to publish bad results. Oxford Nanopore's early access programme survived unflattering accuracy papers and gained a community that made its reputation.
• Match the site format to the buyer: training centers for hospital capital equipment, study clubs and courses for fragmented practice markets, certified service labs for research tools.
• Never cascade faster than you can train. Early adopters climbing a learning curve in public produce your competitor's best evidence.