CG Common Ground | A medical device enters the US: diligence before the buyer (2026)
What we did and what it producedCompleted

The work, decision by decision

The work, decision by decision

  1. Name the gap between the advice and the incentive. What was known: everyone in this category recommends diligence before marketing a raise, and almost nobody is paid in a way that rewards enforcing it. The question: does this founder's plan follow the advice or the incentive. It followed the incentive. What it changed: the mandate. Not "help this raise close fast" but "find out whether it survives scrutiny before anyone commits to it closing".
  2. Set the scope to freedom to operate and economics. What was known: the deck's claims about the mark, the patents, the licensing history and the distribution relationships. The question: which of these, if wrong, is a closing condition rather than a haircut. The scope came out of that test, as above. What it changed: the diligence stayed fast and pointed instead of broad and slow, which is the only way a clock that short is enough.
  3. Build the data room downstream of the diligence, not beside it. What was known: a room had to exist before any outreach could start, and a room, unlike a conversation with a named investor, can be rewritten at no cost when a finding changes what belongs in it. The question: does the room get built around what the diligence finds, or does the diligence get run to validate a room already built. Around the findings. What it changed: every file in the room reflected what had actually been found by the time anyone outside saw it, and the founder saw the findings before he saw a deck that assumed them away.
  4. Read the five findings on two axes, not one. What was known: five distinct problems, surfaced by register searches, the agreements themselves and reference calls. The question: do they all carry the same kind of risk. They do not, and they differ on two axes that answer different questions. Severity is how bad the consequence is if the finding stays open, up to voiding the asset a buyer would be paying for. Discoverability is how fast a US investor's own diligence finds it without being told. The third-party IP question and the trademark conflict rank highest on severity, because they are freedom-to-operate questions an institutional investor's counsel will not let past a term sheet. The royalty terms and the exclusivity clause reprice the deal; they are read from the company's own agreements and are plausibly negotiable. The dead distribution deal ranks last on severity and first on discoverability: one reference call exposes it, and once an investor catches one claim in a deck that is no longer true, every other claim gets re-checked, including the accurate ones. What it changed: the founder got two useful views of the same five facts rather than one flat list, and the outreach plan below is built from the second axis.
  5. Tier the buyers by disclosure, not by the odds of getting caught. What was known: the founder still wanted a path to US investors. The question: who is the actual buyer for a raise carrying open freedom-to-operate questions, and does that buyer set change as the questions close. It does, and the answer is a sequence of buyer profiles rather than one list. A single undifferentiated outreach list puts the company in front of the investor tier least able to tolerate open IP questions first, and burns the introduction before the company is ready to use it. What it changed: outreach opens tier by tier, and the gate on every tier is written disclosure, never the hope that a lighter diligence process misses something.
  6. Build the go-to-market plan the raise was for. What was known: the founder's model was licensing-led, licensing the product to US manufacturers and retail partners rather than building US manufacturing, with two possible entry channels in front of him. The question: which channel the company could actually license into, and at what economics, given what the exclusivity and royalty findings said about the agreements it already had. What it changed: the go-to-market plan was sequenced behind the same two findings, because a channel the company is contractually barred from, or one where the royalty economics do not clear, is not a channel.
  7. Price the continued engagement, after the diligence had already run. What was known: the diligence was done and the five findings were on the table. The question asked at the time was what to charge for a defined remediation scope over the following months, and the proposal went out as one fixed monthly fee for that scope. What it changed: nothing about the findings. The founder declined to pay an upfront fee for the remediation phase and the prospect went on hold. No contract was signed.

The retrospective judgment, which I hold as a rule now rather than a regret: the diligence should have carried its own fee, locked before the work began. Running it unpaid meant the founder could reject an unwelcome answer at zero cost the moment it arrived, which is close to what happened. Nothing in the working sequence weighed whether running that diligence unpaid was itself a decision. It was a default nobody examined, and the part of the job I was best at, doing the diligence right, was exactly the part I stopped questioning.

Price the no before you go looking for it, or you are financing your own bad news.

What the sprint produced, and the go-to-market plan that came with it

The gate is the outreach plan. Read down any column and it says what has to be true before that tier is approached; read across any row and it says how one finding's demand on the company rises with the sophistication of the buyer. The two freedom-to-operate findings are the only ones that ever require full resolution, and only for the tier best able to fund the whole raise. The two economics findings are never a bar to a conversation once they are documented with a clear pricing position. The dead distribution deal is a correction to the pitch, and it is the one item that gates every tier, because it is the one item every tier will catch.

The room held fourteen files, and the red-team stress test is the one that matters most. It is written against the company, the way a skeptical investor's associate would write it, and it is where the five findings live with their sources beside them: the register a trademark conflict is checkable against, the assignment record a third-party IP question traces to, the counterparty's own current status that a reference call confirms. None of the three depends on our interpretation to exist. The royalty and exclusivity findings are read from the company's own agreements, a step closer to interpretation and still not manufactured by the process that found them. The rest of the room is what a raise needs in any case: the product and its clinical evidence, the market and the competition, the launch plan, the capital plan, the brand and the story, manufacturing and unit economics, partnerships, and the founder's own background and correspondence, read the way a buyer would read them.

The go-to-market plan

The founder's model was licensing-led: license the product to US manufacturers and retail partners and earn a royalty, rather than build manufacturing here. Two entry channels were in front of him, a regulated medical channel and a consumer channel, and the deck treated the choice as open. The plan we handed over made it a sequence rather than a choice. The exclusivity clause decides which channels the company is free to license into at all; the royalty terms decide whether the economics of a US license clear once the existing obligations are paid. So the channel decision waits on those two findings, and the outreach to manufacturing and retail partners runs behind the investor gate on the same logic: a partner who finds the exclusivity problem himself is a partner lost. The deck work that went with it was ordinary and necessary: drop the framing that leaned on an unnamed secret, replace the flattering comparison with a credible one, and correct the geographic and biographical inconsistencies before any partner or investor read it, because the dead distribution deal had already shown what one stale claim does to every other claim on the page.

The arrangement on the go-to-market side is tentative and unsigned: if Common Ground brings the founder a deal, the work is split evenly between us. It is case by case, and it has not been tested by a deal yet.

What we kept, replaced and installed

Kept. The founder's own clinical file and licensing history as the starting evidence; nothing was re-derived where a primary document existed. The founder's raise target and his licensing-led model, both of which survived the diligence intact. Prince Capital as the licensed route for the raise itself, with Common Ground's side kept to diligence, the room and the market-entry narrative, so the capital track and the advisory track never blurred.

Replaced. "Build the room and start outreach together" as the plan. It came from the founder's timeline and from the way the category normally runs, where findings surface inside investor conversations as they come up. The faulty logic is that a finding surfacing late is a normal part of the process. It is not: a finding an investor makes after commitments have formed around a close reads as a deal-breaker discovered too late, and the same finding made before outreach reads as a company that knows itself. It had to change before the first conversation, because the first investor to find a freedom-to-operate gap is the one whose read travels.

Installed. A two-axis findings register, severity and discoverability kept separate because they answer different questions and sort the same five facts into different orders. A buyer-tier gate on outreach, with written disclosure as the gate at every tier. And the rule I wrote after this engagement, which Attachment A turns into a template: the diligence phase carries its own fee, fixed and payable on delivery of the answer regardless of what the answer turns out to be, agreed before the work starts; the remediation phase is priced from the findings and offered as a menu of scopes rather than one flat number; the capital track stays separate and runs through the licensed advisor. I start from that template now on any capital-readiness engagement. How it is applied is decided case by case.

What it cost to hold the line, and what I would watch

Gating the outreach cost the founder the fast start he wanted and cost me the goodwill that comes with giving a client what he asked for. Scoping the diligence to two categories cost the thoroughness a broad checklist would have signaled. Handing the findings over straight, rather than folding them into the room, cost the engagement its next phase: the founder declined to pay for remediation, and I do not know how much of that was the findings and how much was the fee. Those were the correct costs, and I would pay them again.

The cost I would not pay again is the unpaid diligence. The diligence ran with no fee agreed and no fee protection if the answer came back unwelcome. Relax that constraint and test it against the outcome: had the fee been locked in advance, Common Ground is paid for the diligence phase, and the founder still faces the same five findings and the same decision about whether to fund fixing them. The outcome does not change. Only who bears the cost of finding out does. A check that has the power to end the thing it is checking has to be paid for regardless of its answer, or the person running it will eventually stop running it honestly or eat the cost every time it comes back negative. That is true of an inspector, an auditor and a reference check, and it has nothing to do with medical devices.

The honest limit on the finding itself: the party grading the severity of these five items is the same party that proposed a remediation scope sized to them. The register's defense against that is source class, not my word. Three of the five findings exist in a public register, an assignment record or a counterparty's own status, independent of anyone's interpretation, and a skeptical reader can check them there.

What I would watch. Whether the two freedom-to-operate findings close, because nothing above the angel tier opens until they do. Whether the exclusivity clause is renegotiated or carved out, because the go-to-market plan has no channel until it is. Whether the deck was corrected before it went in front of anyone, because the dead distribution deal is the one finding every investor at every tier will catch on the first call. And on my own side, whether the diligence-phase fee holds on the next engagement where a founder in a hurry asks for the fast version, because that is the moment the rule is worth something and the moment it is easiest to drop.

What it produced

Five findings in under two weeks, before any investor saw a page: a trademark conflict, a royalty issue, an exclusivity problem, a third-party IP question, a dead distribution deal. The founder declined the remediation fee; the decision after the findings is his.

A slice of the project list

A few related projects.