Most medical device sales training programs are built backwards. They start with the product — anatomy, indications, technique, competitive positioning — and assume that a rep who knows the device will be able to sell it. Then the rep gets into a hospital and discovers that the person who decides whether the device gets bought never once asked about the device.
The training gap in medical device sales is not a product knowledge gap. It is a gap between what reps are taught and the room they actually have to survive.
What it costs to get this wrong
The most reliable number on the cost of losing a salesperson comes from DePaul University's Center for Sales Leadership, whose Sales Effectiveness & Sales Acceleration Survey put the average cost to replace a salesperson at $97,690, against a planned training investment of $9,721 per rep per year. Those are 2016 figures from 127 responding firms, so treat them as directional — but the ratio is the point. Companies spend roughly a tenth of the replacement cost on preventing the replacement.
In medical device specifically, the arithmetic is worse, because the scarce resource is not the rep's time. It is access. A rep who is not credentialed cannot enter the hospital. A rep who is credentialed but unprepared gets one meeting with a value analysis committee and does not get a second. Every week of unproductive ramp is a week of burned access, and access does not come back on demand. That is the case for treating sales training and onboarding as infrastructure rather than an event.
Start with what the regulation actually requires — and what it doesn't
Almost every article on this topic tells you that FDA requires medical device sales reps to be trained, and cites 21 CFR 820.25 as the authority. Both halves of that are wrong, and the second half is now provably wrong.
FDA's Quality Management System Regulation took effect on 2 February 2026, amending Part 820 to incorporate ISO 13485:2016 by reference. In the process, § 820.25 — the old personnel and training section — was removed. Sections 820.20 through 820.30 now read [Reserved] in the eCFR. If your training deck still cites 820.25, it is citing a section that no longer exists.
What replaced it is § 820.10, which requires a documented quality management system complying with ISO 13485 — including clause 6.2 on competence and training. But that obligation attaches to personnel whose work affects product quality. Sales reps fall inside it only where a manufacturer's own QMS gives them quality-affecting duties: complaint intake, field service, device handling, returns.
The real training burden on reps comes from somewhere else entirely, and it is contractual rather than federal. Hospital vendor credentialing programs set the bar — immunizations, background checks, and a course list that typically runs to a dozen or more modules covering HIPAA, bloodborne pathogens, OR protocols, radiation safety, aseptic technique, the Sunshine Act, and the AdvaMed and PhRMA codes. AdvaMed has endorsed ANSI/NEMA SC 1-2020 as a voluntary national credentialing standard precisely because the unstandardized version costs the industry over a billion dollars a year in duplicated effort.
Two practical consequences for your program. First, credentialing content is compliance overhead, not selling capability — budget for it separately and do not let it eat the training calendar. Second, because credentialing requirements vary by facility and expire on a rolling basis, the records problem is permanent. This is the part of the program that has to run on a system with completion tracking and audit-ready records, not a shared drive.
Train for the room, not the device
The single biggest structural change in medical device selling is who is in the room. Gartner's 2025 sales survey, fielded across 632 B2B buyers, found buying groups ranging from five to 16 people across as many as four functions, with 74% of buyer teams showing what Gartner calls unhealthy conflict during the decision.
In medtech the shift is sharper. Bain's Front Line of Healthcare research, surveying 1,156 physicians and 154 procurement officers, found that nearly 60% of physicians said equipment purchasing was predominantly led by procurement with some physician input, and only 17% of surgeons reported a large degree of discretion. That study is a European sample and now several years old, so do not quote it as a US benchmark — but the direction it identified has not reversed.
A rep trained only to talk to the clinician is trained for a conversation that no longer decides anything on its own. The program has to cover the economic buyer, the value analysis committee, and the supply chain function as first-class audiences, with their own materials and their own objections.
What surgeons actually say matters
Here is the finding that should shape your program more than any other, and it is peer-reviewed rather than borrowed from a vendor deck. Burns and colleagues surveyed 201 orthopedic surgeons in Pennsylvania performing at least 12 implant procedures a year, published in Medical Devices: Evidence and Research, and asked what drives their choice of implant.
On a five-point scale, patient outcomes rated 4.49. Implant longevity rated 4.34. Scientific evidence rated 4.29. And the sales rep's follow-up and thoroughness rated 4.39 — above implant longevity. Rep knowledge rated 4.36. Rep availability rated 4.16.
Read that again: surgeons scored the rep's thoroughness alongside the clinical performance of the device. Meanwhile consulting arrangements rated 2.10, and only 12% of surgeons reported any economic relationship with a vendor. The relationship that moves the decision is a competence relationship, not a commercial one — and competence is trainable.
Ignore the retention statistics you have been sold
You have seen the claim that 87% of sales training is forgotten within 30 days. You may also have seen 70% within a week, or 84% after 90 days. None of them have a study behind them. Trace any of the three and the trail runs from one vendor blog to another, each citing the other, occasionally gesturing at unnamed research — a dead end documented in detail by the sales-training firm Clario, which went looking for the primary source and found that it does not exist.
What does exist is Murre and Dros's 2015 replication of the Ebbinghaus forgetting curve in PLOS ONE. It is real, peer-reviewed, and open access. It is also one person memorizing nonsense syllables, and its numbers are savings scores rather than percentage-remembered. It tells you the shape of forgetting — steep at first, flattening out — and nothing at all about how much of your hinge-mechanism training a rep retains.
The practical takeaway does not depend on a number. Forgetting is front-loaded, so reinforcement has to be too. Build spaced retrieval into the first 30 days rather than staging a single certification event at the end of week two, and let reps practice the conversation before they have it rather than re-reading the deck.
A five-stage program you can actually run
| Stage | Timing | What it covers | How you know it worked |
|---|---|---|---|
| 1. Credential and clear | Pre-start to week 2 | Vendor credentialing modules, immunizations, background checks, AdvaMed Code, facility-specific policies | Rep is cleared for the target accounts. Records are audit-ready and expiry-tracked. |
| 2. Product and procedure | Weeks 1–4 | Anatomy, indications, technique, competitive landscape, the clinical evidence base | Rep can walk the procedure and cite the evidence without notes. |
| 3. The room | Weeks 3–6 | Value analysis committee mechanics, procurement and supply chain priorities, economic modeling, GPO and contract status | Rep can build and defend an economic case, not just a clinical one. |
| 4. Rehearsal | Weeks 4–10, recurring | Role play against real personas: the skeptical surgeon, the price-driven materials manager, the VAC economic challenge, the incumbent-competitor objection | Rep handles each scenario without freezing. Manager has seen it, not just heard about it. |
| 5. Reinforcement | Ongoing | Spaced retrieval on the highest-stakes content, refreshers on new evidence and label changes, re-credentialing before expiry | Performance holds at 90 and 180 days, not just at certification. |
The sequencing matters more than the content list. Stage 3 before stage 4 — reps cannot rehearse a conversation they have not been taught to have. And stage 1 genuinely gates everything: an uncredentialed rep with perfect product knowledge is a rep who cannot get in the building.
Measure ramp, not completion
Course completion tells you a rep clicked through. It does not tell you they are ready. The measures worth tracking are time to first credentialed account visit, time to first independent call, time to first VAC presentation, and 90-day retention of the material that matters most — checked by retrieval, not by asking.
You will notice that none of those are content-consumption metrics. That is deliberate. Forrester's research, originating with SiriusDecisions, has long held that roughly 65% of B2B content goes unused. A training program measured by material delivered will always look healthier than it is.
Where to start
If you are building this from nothing, start at stage 3. It is the stage most programs skip entirely, it is the one the research says decides the outcome, and it is the cheapest to add because the material already exists somewhere in your organization — it is just sitting in finance and health economics rather than in training.
If you already have stages 1 and 2 running and reps are still ramping slowly, the problem is almost certainly stage 5. Front-load the reinforcement. And if you want a fuller view of what the delivery layer has to do — offline access, SCORM, learning paths, audit-ready records — we wrote about what a field-ready LMS actually needs to handle, and about improving rep onboarding more broadly. Both are written for medical device field teams specifically, which is where the generic sales training playbook tends to fall apart.