Console lock-in is symmetric — it protects the incumbent for weeks and excludes it for years

Claim

The standard read on a console-plus-consumable business is that the installed base is a moat: the customer has bought the capital equipment, trained on it, and credentialed staff to it, so the consumable revenue is annuity-like. True — but the analysis is usually stopped one step too early.

The friction is a property of the console slot, not of the vendor. Whoever occupies the slot enjoys it. So the same mechanism produces two opposite conclusions depending on timing:

Incumbent occupies the slot → a two-to-three-week consumable stockout does not cause switching. The customer defers the job, falls back to older equipment, or uses a second console it already owns. This is genuine protection and it is why short outages are usually harmless.
A rival console lands in the slot → the incumbent is locked out for the equipment's life, typically three to five years, and no amount of consumable price or clinical evidence gets it back before then.

Therefore the risk from any supply interruption scales with the fraction of customers who already hold a competitor's console — the ones who can substitute instantly rather than defer. That fraction is the number to track, and it is usually disclosed by the competitor as installed-base growth, not by the incumbent.

The BSX case, 2026-08-26

FARAWAVE catheters only fire from a FARASTAR generator through a FARADRIVE sheath, so a lab cannot run a Medtronic Affera case in week one of a Boston Scientific outage unless an Affera system is already installed with proctored operators. On paper that is strong protection.

But Medtronic's US cardiac-ablation revenue grew +124% in FY2026 with the Affera US installed base up ~40% sequentially in Q4 alone, it took 8 points of US ablation share, and Abbott moved Volt 2.0 to full US market availability in the same quarter as the outage. (Inference, not sourced: the population of US labs holding a second PFA console is therefore far larger in August 2026 than in August 2025.) BSX's protective friction had been eroding for four quarters, and the outage arrived when it was at its weakest.

The learning curve is the asset being transferred

The switching cost in these markets is not the box — it is the operator's competence with it. A physician who runs eight to twelve competitor cases over three weeks because the incumbent's consumable was unavailable emerges proctored, comfortable, with staff trained on the rival workflow — at zero customer-acquisition cost to the rival. That is the durable transfer. The console placement is merely what makes it possible.

How to apply

When assessing supply risk at a razor-and-blade company, ask "what fraction of customers already hold a rival console?" not "how long is the outage?" The second question only matters through the first.
Read the competitor's installed-base disclosures as the incumbent's risk metric. Rivals report console placements enthusiastically; incumbents do not report displacement at all.
Score switching costs with an explicit direction and confidence, not a single number. On BSX the score stayed 5/10 while confidence in it fell sharply — the mechanism was confirmed and the variance became asymmetric to the downside. A flat score hid that; the note is what carries it.
Treat a rival's launch window overlapping a supply interruption as a compounding event, not two events. Abbott's full US Volt launch and BSX's outage occupying the same weeks is worth more than either alone.
Symmetry cuts the other way too. Where an incumbent wins a slot during a rival's stumble, that gain is equally durable — the mechanism is directionless. J&J's Varipulse safety pause in Jan-2025 made Farapulse "the biggest beneficiary," and nineteen months later J&J's EP was still a stated drag.
Applies beyond medtech wherever consumables run on installed capital with certification or validation friction: sequencers and reagents, clinical analysers, semiconductor tools, industrial printers.

Confidence and limits

Marked medium, not high. The mechanism is well-evidenced and the BSX-specific installed-base inference is explicitly mine rather than sourced. The counter-argument is real: Farapulse's single-shot, mapping-optional workflow is genuinely faster, high-volume labs have a throughput reason to return, and three weeks is short relative to how slowly lab defaults actually change. The pattern predicts elevated tail risk, not a base case — on BSX it moved the permanent-transfer probability to ~30%, not to a majority.

What would falsify this

A documented case of a consumable supplier regaining share inside a console generation after a rival system was installed during an outage — i.e. evidence that the slot is contestable on consumable merit alone. Alternatively, BSX's Q4-2026 recapturing the Q3 shortfall in full despite competitor console placements during the outage.

[[pattern-supply-outage-is-availability-not-quality]] — the base rate this pattern is the named exception to.
[[pitfall-logo-retention-masks-dollar-retention]] — same family: a retention metric that looks one-directional and is not.

History

2026-08-26 — written during /analyze BSX, when the Moat analyst identified that the FARASTAR console dependency protecting BSX through a short outage is exactly the thing that would exclude it from any lab where an Affera or Volt console landed during one.