Butterfly Is Selling a Semiconductor Story at 20x Revenue — and Filing the ATM to Match
Q3 delivered a 52% beat, a record gross margin with a tariff assist, and a $150M shelf. The Home Care narrative quietly fell out of the top of the keyword stack.
BFLY · Earnings Call · 2026-10-29
The beat is real. The run-rate is not moving.
Butterfly's third quarter was, on the headline, its best ever. “We delivered $32.6 million in revenue, beating top end of our guidance by 9%.” — Joseph DeVivo, President and CEO · 2026-10-29 Gross profit hit a record $23.8M, adjusted EBITDA loss collapsed to $1.2M from $8.1M a year ago, and full-year revenue guidance went to $123–126M. The stock, notably, did not care much: it enters the print roughly 24% below its 22 September high of $9.93, with the 90-day tape showing a modest decline and a $2.0B market cap. There is a reason the tape is suspicious of the headline growth rate. Total revenue has compounded 223% over six years, but the recent sequence is $32M (Q4 2025), $27M (Q1 2026), $33M (Q2 2026), $32.6M (Q3 2026). The business has been parked in the low-thirties for a year. The upgraded full-year range implies $31–34M in Q4 — i.e., another flat sequential quarter, against a Q4 2025 comp that already contained $7.3M of Embedded. The raise is the Q3 beat, annualised, not a new slope.Embedded did the beating — and got less transparent
Embedded revenue was $11.1M, up over 1,000%, versus the $7–8M quarterly cadence the CFO had guided to just one quarter earlier: “we expected Embedded to be somewhere around $7 million to $8 million per quarter” — John Doherty, Chief Financial Officer · 2026-07-30. So the segment is running 40%+ above the framework management set in July. New partners Merge Labs and Bridge Neurotech (both out of the Arbor Neuroscience / Caltech orbit) plus a stealth third brain-computer-interface partner contributed. The strategic shift worth marking is the shift in customer type. The pipeline is migrating from startups to large entities with mature products, where revenue could arrive fast because the partner already books product revenue. That is why five seven-figure TCV deals and ten six-figure TCV deals mattered more this quarter than the segment total. But the company explicitly declined to size Midjourney, which is the load-bearing partner. "We're not going to specifically call Midjourney revenue," the CFO said flatly — and Midjourney, a rank-15 keyword two quarters ago, has dropped out of Butterfly's own top-30 entirely. When 34% of revenue is Embedded and Embedded is Midjourney-plus-others, refusing to quantify the largest single contract is a disclosure choice, not a rounding decision. The showcase enterprise proof point is worth a raised eyebrow. Management described a top-five customer that "started off with a couple hundred probes" and is “probably at 700 or 800 probes today, and so that over a year and a half” — Joseph DeVivo, President and CEO · 2026-10-29. On the August 2025 call, the same anecdote was told as “upwards of 700 probes” — Joseph M. DeVivo, Chief Executive Officer · 2025-08-01. Same account or not, the anchor enterprise win is not visibly compounding.Record margin, with an asterisk from the global tape
Adjusted gross margin of 73% (+913bp) is a genuine milestone. But the CFO attributed it to mix "as well as a tariff refund under the International Emergency Economic Powers Act." That is not idiosyncratic — tariff refunds are this week's cross-sectional tell, flagged by ACU, AOS, ORLY, VFC and BOOM among others in the reporting window. When a company-specific record margin is partly a sector-wide one-time reclaim, you should underwrite the underlying mix improvement, not the reported level.The ATM, and what 20x revenue buys you
Filing a $150M universal shelf and a $150M at-the-market programme on the morning of your record quarter is a statement about how management sees its own paper. The framing was deliberately defensive: “This is good corporate housekeeping. We have no immediate need for cash” — John Doherty, CFO · 2026-10-29. With $122M of cash and $2.4M of quarterly burn, that is true — the ATM is optionality sold at a high price, not a lifeline. The valuation is the subtext of the whole call. Price to revenue sits at 19.9x, and price to gross profit is 37.7x. R&D, meanwhile, re-accelerated: research and development ran at roughly a third of revenue, up 27% year over year. Hence the pitch, delivered almost as a rebuke:That is a company asking to be valued on a roadmap — the fourth generation chip with tissue harmonics in early 2027, a cart entry into what the CEO calls an $8B existing market in 2028, Apollo at 20x processing power, wearables for Embedded partners, and physical AI as a bridge to brain-computer interfaces. It is a coherent and genuinely differentiated story — the competitive moat built on 600–700 patents and $300M of first-generation chip spend is not easily replicated. But it is a multi-year claim being capitalised today.Butterfly should not be looked at as a revenue multiple today. It should be looked at as the magnitude of opportunity that we are capturing for our future and for investors.