Mapping the clinical adoption funnel from awareness to standard of care

Edwards Lifesciences won FDA approval for the SAPIEN transcatheter aortic valve in November 2011, for patients considered too sick for open-chest surgery. Nine years later, the 2020 ACC/AHA valvular heart disease guideline is where TAVR stopped being a rescue option and became a default choice for a wide band of aortic stenosis patients. Between those two dates: three randomized cohorts (PARTNER 1, 2 and 3), a CMS coverage decision that tied payment to heart-team staffing and case volume, and thousands of hospital-level protocol rewrites.

Every one of those steps is a gate. In front of each gate sits a cohort of clinicians and sites with a conversion rateconversion rateThe percentage of visitors or prospects who complete a desired action (purchase, sign-up, contact form), calculated as conversions divided by total opportunities.View full definition →, a dwell time, and a clock the field team only partly controls. This lesson owns that stage model, and the mechanics that push a cohort across each gate.

Why the clinical funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition → has gates a software funnel does not

Software funnels run click, sign up, activate, pay, renew, and the vendor owns every gate. Here at least three belong to someone else: a regulator decides whether you can be evaluated at all, a payer decides whether use is affordable, and a specialty society decides whether use is recommended. The buying unit is the multi-stakeholder committee with a long cycle that the acquisition-cost lesson prices.

Standard of care means the treatment guidelines and peers treat as the default, correct choice for a given condition. It is the funnel's real exit. Once you sit inside a hospital protocol and a society document, displacing you takes committee action rather than a discount.

The six clinical gates

Percentages further down are planning assumptions for modelling, not published rates. Where external reference data belongs is the benchmarks lesson.

Gate 1: Awareness (congress and peer exposure)

Entry is usually a specialty congress (ESC, ACC, ASCO, RSNA) plus the physician-influence and reference-site work its own lesson details. Metric: qualified congress leads, filtered to your target specialty and institution type.

Gate 2: Clinical interest (evaluation request)

The clinician asks for a demo, a data pack, or entry into an evaluation. First real intent signal.

Gate 3: Evaluation

Supervised use on a handful of cases, sometimes under a formal clinical evaluation framework. In the EU, ongoing clinical evidence is required under the Medical Device Regulation (2017/745). In the US, entry runs through 510(k) clearance (equivalence to a marketed device) or PMA for high-risk classes.

Gate 4: First procedure

The first real, billed procedure. Your activation event.

Gate 5: Repeat use and account penetration

Routine use by the first clinician, then spread to colleagues in the same service line.

Gate 6: Protocol embeddingembeddingAn embedding is a numerical vector that represents data (text, images, or items) in a way that captures meaning, so similar items sit close together in space.View full definition →

Written into the hospital protocol, and ideally cited in guidelines.

One structural decision before you count anything: the unit. For TAVR and for Abbott's MitraClip, Medicare coverage conditions reference heart-team composition and institutional volume, so the site is the unit of the funnel and an enthusiastic cardiologist at a non-qualifying centre cannot convert no matter how good your Gate 2 work is. For a Hologic tomosynthesis system the unit is the imaging service line and its capital cycle. Mixing clinician-level and site-level entities in one funnel is the most common reason two teams cannot reconcile their numbers.

A worked funnel calculation

A field team launching a surgical device in a mid-size European market. Illustrative assumptions.

Stage conversion = (entities exiting this stage into the next) / (entities entering this stage).

End to end, 1,000 leads produced about 13 embedded accounts: a 1.3% lead-to-standard-of-care rate. The worst gates are evaluation to first procedure (60%) and repeat to embedded (30%). Ten points on the repeat-to-embedded step adds roughly 4 embedded accounts, a 30% lift on the only output that compounds. Ten points on Gate 1 adds half an account.

These gate volumes are what the neighbouring lessons consume, and the denominator you hand them changes their answers more than their formulas do. Sixty sites reached a first procedure; about 13 embedded. Spend divided by first procedures reads roughly a fifth of the same spend divided by embedded accounts. Pick the embedded-account denominator for the committee-level acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition →, and feed the same cohort into the multi-stream account value model, so that capital, disposables and service revenue are counted against the same 13.

What moves a cohort between gates

• Gate 1 to 2 moves on evidence, not on booth traffic. Abbott's MitraClip was approved in 2013 for degenerative mitral regurgitation in prohibitive-risk patients. Interest from heart failure teams stayed thin until COAPT was published in late 2018 and the label widened in 2019. A single trial readout reprices the entire cohort standing at Gate 2 at once, which is why medical affairs spending behaves like brand investment and not like direct response.
• Gate 3 to 4 is usually a supply constraint dressed up as a persuasion problem. If a new implanter needs three proctored cases and one field clinical specialist can cover eight cases a week across a territory, that specialist onboards at most a dozen implanters a quarter before travel and cancellations. Adding congress spend on top of a saturated proctoring roster produces a longer queue, not more procedures.
• Gate 4 to 5 moves on procedure velocity and trained depth. Procedures per active site per quarter, flat or falling after first use, means the site will not embed. Count trained clinicians too: one champion is a single point of failure, three or more is institutional.
• Gate 5 to 6 runs on external clocks. Guideline committees revise on multi-year cycles, and reimbursement codes arrive when they arrive. Hologic's tomosynthesis system was approved in 2011, but the Medicare add-on payment for the 3D exam from 2015 is what let a whole cohort of US facilities justify the capital, followed by state-level coverage mandates. You can prepare for that moment; you cannot buy it.

Gates are not monotonic. Sites slide backwards when a champion moves hospital, when a group purchasing contract flips, or when a new generation of your own product forces a re-evaluation you did not ask for. Model that as a reverse flow between Gates 5 and 6, not as noise.

For grounding assumptions in device pathways and evidence expectations, the FDA's plain-language overview is a solid free start: FDA Device Approvals and Clearances.

🎬 [VIDEO: "How Medical Devices Get to Market" - youtube.com - a concise walkthrough of FDA 510(kkThe average number of new users each existing user generates through referrals. Above 1.0, growth compounds on itself and becomes exponential.View full definition →) vs PMA pathways and why evidence gates shape commercial timing]

Four ways the funnel lies to you

Evaluation tourism. Free loan units attract clinical curiosity from sites with no capital line and no committee slot for eighteen months. Those evaluations inflate Gate 3 and then rot. Qualify on budget path and committee calendar before the box ships, and report evaluations without a funded route separately.

Cohort mixing. Dividing this quarter's first procedures by this quarter's evaluation requests means nothing when the median lag is nine months and inboundinboundA strategy that attracts prospects organically via valuable content (blog, SEO, social) rather than interrupting them.View full definition → requests are growing. The ratio reads pessimistically during growth and flatters you the quarter after a quiet period. Date-stamp entry and follow the cohort as it ages.

Treating first procedure as the win. It is the midpoint. A launch that books 60 first procedures and celebrates has 47 sites that will quietly revert to the prior standard.

Assuming every product uses the same gate order. A capital imaging purchase has no free trial gate: the customer commits money before use, so reimbursement clarity and total cost of ownership have to be settled before Gate 3, not after. Reorder the funnel to match the product, then measure it, rather than forcing an implantable-device template onto a scanner.

Reading your own funnel

Reimbursement deserves its own gate on the diagram. A device can be FDA cleared and still lack a favourable CPT code or payer coverage; in Europe, national HTA bodies such as NICE and HAS gate uptake independently of the CE mark. Sites can pass Gates 1 through 4 and stall permanently at a coverage decision nobody on the commercial team owns.

Instrument every gate with dated stage transitions per account. A shared spreadsheet with entry dates beats a CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → stage field that sales reps update from memory. After 12 to 18 months your own conversion rates describe your therapy, your markets and your reimbursement position better than any external figure.

Then split the funnel by therapy area and by market, never blended. A well-reimbursed high-volume specialty and one awaiting a coverage decision produce the same-looking Gate 3 numbers and completely different Gate 6 outcomes, and the blended average hides both.

Key Takeaways

• Six gates run from congress awareness to protocol embedding, and the exit event is standard of care, not the first sale. In the worked example, 60 first procedures yielded about 13 embedded accounts.
• Fix your worst gate. Ten points on repeat-to-embedded added 4 accounts; ten points at the top of funnel added half of one.
• Decide the unit (clinician or site) and the denominator (first procedure or embedded account) before handing numbers to the acquisition-cost and account-value models.
• Gate 3 to 4 is often limited by proctoring capacity, and Gate 5 to 6 by guideline and coding cycles you can prepare for but not purchase.
• Gates run backwards when champions leave or contracts flip, and evaluation tourism plus cohort mixing are the two arithmetic errors that make a stalling funnel look healthy.