Paid acquisition into clinical and channel buyers
Where are new device competitors actually entering?
Every device manufacturer eventually asks some version of the same question before a budget cycle: is this category getting more crowded, or does it just feel that way? The usual answer comes from the sales team’s impressions, a competitor’s press releases, and whatever an analyst report happened to cover.
There is a better source, and it is free. Every device cleared for the American market through the premarket notification pathway is filed publicly, with the review panel that cleared it and the date it was cleared. That file is a register of market entry, and almost nobody in medtech marketing reads it.
What does the clearance data actually show?
The total is flat and the mix is not. Across all review panels, clearances rose from 9218 in the 2016-2018 period to 9700 in 2023-2025, an increase of 5.2 percent over roughly seven years. Underneath that near-static total, individual categories moved by 60 percent in both directions.
This is the whole finding, and it is the reason category-level intuition fails here. A manufacturer watching the aggregate would reasonably conclude that competitive entry is stable. A manufacturer in Physical Medicine or Ophthalmic would be experiencing something entirely different, and the aggregate would give them no warning either way.
Which categories are gaining, and which are losing?
Physical Medicine grew fastest, from 184 clearances to 295, a rise of 60.3 percent. Ophthalmic fell hardest, from 204 to 140, a drop of 31.4 percent. Orthopedic, still the single largest panel by volume, lost 13.5 percent of its clearance activity while remaining the busiest category in the file.
The growth side has a pattern worth naming. General and Plastic Surgery added 234 clearances to reach 1144, a rise of 25.7 percent. Neurology rose 23.7 percent, from 497 to 615. Radiology, already one of the largest panels, still managed 13.7 percent growth from 1184 to 1346 — meaningful because growth is much harder to achieve from a large base than a small one.
The decline side is more mixed. Ophthalmic, Hematology, down 29.5 percent, and Immunology, down 25.3 percent, are small panels where a modest absolute change produces a dramatic rate. Clinical Chemistry fell 16.2 percent. Cardiovascular barely moved at all, down 3.8 percent from 1109 to 1067, which for a panel of that size is closer to stable than declining.
Orthopedic is the one that should interest a growth-stage manufacturer most. It is simultaneously the largest panel in the file and one of the fastest shrinking, which is an unusual combination. A category that large does not lose 13.5 percent of its clearance volume by accident.
Why does a flat total hide so much movement?
Because the aggregate is a sum of offsetting movements, not a description of any particular market. Orthopedic shed 213 clearances while General and Plastic Surgery added 234. Those two facts cancel almost exactly in the total, and the total is the number most industry commentary reports.
This matters more than it sounds. The premarket notification pathway has a practical ceiling set by review capacity, not by market demand, so the annual total tends to sit in a narrow band regardless of what is happening commercially. Reading that stability as market stability is a category error.
What the file actually measures is where new products are being brought to market, panel by panel. When a panel’s clearance count rises, more products are arriving that a buyer could choose instead of yours. When it falls, either the category is consolidating, or the innovation has moved to a pathway this file does not cover.
Both explanations are plausible for Orthopedic, and the data here cannot distinguish between them. That distinction matters enormously for a marketing plan, which is exactly why the honest version of this analysis stops short of asserting one.
Are more clearances the same as more competitors?
Not necessarily, and the difference decides what a rising panel means. Checking the applicant names inside the two most interesting panels shows they are structurally opposite. Physical Medicine’s growth comes from a long tail of small filers. Orthopedic’s larger volume sits with established names filing repeatedly.
This is the check that makes the trend readable rather than merely true, so it is worth walking through the numbers.
In Physical Medicine across 2023-2025, the hundred largest applicants account for 193 clearances between them. The single largest filed 9. The top five hold 15.5 percent of that total and the top ten hold 26.4 percent. There is no dominant filer anywhere in the panel — the growth is genuinely distributed across many separate entrants, most of them clearing one or two products.
Orthopedic looks different in kind, not just in size. Its hundred largest applicants account for 808 clearances, and the largest single filer holds 41 of them. Yet the concentration ratios are almost identical: the top five hold 16.5 percent and the top ten hold 25.1 percent.
That similarity is the interesting part, because it means the two panels are equally concentrated in statistical terms while being completely different commercially. Orthopedic’s leading filers are recognisable multi-product manufacturers iterating on established lines. Physical Medicine’s are largely first-time entrants, and a substantial share of the names in that panel are contract manufacturers based in a single manufacturing region.
For a marketer the practical translation is direct. Growth driven by many small new entrants usually arrives as price pressure and channel noise rather than as a head-to-head competitive threat. Growth driven by incumbents filing more often usually arrives as feature parity and a harder differentiation problem. The clearance count alone cannot tell those apart. The applicant list can.
How was this calculated?
Two queries against the openFDA 510(k) Clearances API, one for each three-year window, counted by advisory committee. Panels with fewer than 60 clearances in the earlier window are excluded from the rate comparison because a percentage change on a small base is noise. All 21 panels remain in the downloadable file.
The exact queries are recorded in the sources file for this article, and the underlying counts are in the CSV above. Anyone can rerun them.
Three-year windows rather than single years, because single-year clearance counts bounce around for administrative reasons — a review backlog clearing in January rather than December moves a year’s number without anything having changed commercially. Comparing two three-year blocks separated by a gap smooths that out without smoothing away the trend.
The comparison periods are 2016-2018 and 2023-2025. The gap between them is deliberate: adjacent windows would mostly measure noise, and including the gap years would blend the pandemic period into both ends of the comparison.
What can this data not tell you?
It counts clearances, not manufacturers, not revenue, and not units sold. One company clearing eight variants of a product produces the same eight rows as eight companies each entering once. Nothing in this file distinguishes those two situations, and they mean opposite things competitively.
That limitation is severe enough to state twice. A rising panel count is evidence of more products, not necessarily more competitors. A manufacturer using this to size competitive threat should pull the applicant names within their own panel — the same API exposes them — rather than reasoning from the panel total.
The review panel is also a coarse category. Radiology covers an imaging workstation and an algorithmic triage tool, which compete for nothing in common. Physical Medicine spans rehabilitation equipment and a range of therapy devices. Any panel-level reading is a reading of an administrative grouping, not of a market.
Clearance is also not launch. A cleared device may reach the market a year later, in a limited geography, or never — clearance permits commercial distribution, it does not commit anyone to it. Treating a clearance count as a count of products a buyer can actually purchase today overstates the competitive picture by an unknown margin, and the file gives no way to close that gap.
One panel in the data is labelled Unknown, holding 206 clearances in the later period. Those are real clearances with an unresolved panel assignment. They are retained in the CSV rather than silently dropped, because quietly discarding records is how a dataset stops being reproducible, but they belong to no category and should not be read as one.
Finally, the premarket notification pathway is not the whole story. Devices entering through premarket approval, and those exempt from clearance entirely, appear nowhere in this file. A category could look quiet here while a great deal happens through another route. Software-driven products in particular have shifted across pathways more than once in the period this analysis covers, which makes any long-run comparison in the software-heavy panels weaker than the same comparison in, say, Orthopedic.
How do you run this for your own panel?
Two requests, no key, no account. The API is open and rate-limited rather than gated, so a browser address bar is enough to reproduce every number in this piece. Substitute the panel name that matches the products being analysed, then compare the two windows.
# clearance counts by review panel, one three-year window
https://api.fda.gov/device/510k.json?search=decision_date:[2016-01-01+TO+2018-12-31]&count=advisory_committee_description.exact
# who is filing inside one panel
https://api.fda.gov/device/510k.json?search=decision_date:[2023-01-01+TO+2025-12-31]+AND+advisory_committee_description.exact:"Orthopedic"&count=applicant.exact&limit=100The panel names are the values returned by the first query, so run it once and read the list rather than guessing at the spelling. Note that the applicant count returns at most a hundred terms, which is enough to characterise the top of a panel but never the full tail.
The most useful version of this analysis is not the one in this article. It is the one restricted to the specific panel a company competes in, run once a year before planning, and compared against the previous run. Absolute levels matter much less than direction.
What should this change about where a device marketer spends?
Look up your own panel before setting next year’s acquisition budget, and treat the direction as a signal about competitive cost rather than about demand. A panel adding products faster than the market grows is a panel where auction prices and buyer attention both get more expensive, usually before anyone internally notices.
The practical version of that is short. Pull your panel’s clearance count for the last three years and the three before. If it is rising faster than the overall 5.2 percent, assume the cost of reaching your buyer is rising too, and that the increase will show up in acquisition efficiency before it shows up in win rates. If it is falling, the interesting question is why — consolidation and category decline demand very different responses.
For a growth-stage company in a shrinking panel, the mistake to avoid is reading the decline as room to breathe. Orthopedic losing clearance volume while remaining the largest panel in the file is consistent with consolidation, and a consolidating category is one where the surviving competitors have more budget each, not less.
There is a second-order effect worth planning around. Paid acquisition costs in a category tend to move with the number of parties bidding, not with the number of buyers, and clearance activity is a leading indicator of the former. A manufacturer that watches its panel fill up has roughly a year of warning before that shows up in cost per qualified lead, which is enough time to build the organic and lifecycle assets that make paid acquisition cheaper. A manufacturer that waits for the cost increase has to solve the problem while paying for it.
The same logic applies to content and positioning. Categories where entrants are arriving fastest are categories where the buyer’s shortlist is getting longer and their evaluation harder. That raises the value of anything that shortens evaluation — comparison material, technical documentation, real specifications — and lowers the value of broad awareness spending that assumes the buyer only has to be reminded you exist.
None of this replaces a real competitive analysis. It does something narrower and more useful at the planning stage: it tells you whether the assumption underneath your plan — that your category is roughly as crowded as it was — is true. For a meaningful number of panels in this file, it is not.
The broader point is the one worth carrying into any category analysis. Public regulatory data is unusually good at describing structure and unusually bad at describing demand. Used for the first, it is close to free. Used for the second, it will mislead you confidently.