The Condition

A consequential decision is made once, in a specific room, at a specific time, by specific people acting on whatever understanding they had assembled in that moment. If the decision is later challenged — whether by a shareholder, a regulator, or an internal successor trying to understand what happened — that understanding cannot simply be recalled. It has to be reconstructed. And it can only be reconstructed from what remains: the documents, the minutes, the memos, the emails that were kept.

This creates an asymmetry worth naming carefully. The people who made the decision generally had a fuller, more immediate view of it than anyone examining it years later ever will — the tone of the discussion, the concerns raised and set aside, the judgment that tied it together. That does not mean the original decision-makers themselves had complete or perfect visibility; they may have worked from information that was itself fragmented, incomplete, or unevenly distributed across the organization. What can be said with more confidence is narrower: whatever view they had, however complete or incomplete, was almost certainly fuller than what survives for anyone examining the decision afterward. If what was written down is thin, incomplete, or reconstructed after the fact, that later examination proceeds from a materially poorer vantage point than the decision itself was made from — regardless of how good or limited the original deliberation actually was.

This piece examines what Delaware courts and federal banking regulators have, in specific and documented instances, treated the record as capable of establishing — and, just as importantly, what they have not treated it as capable of establishing.

What the Record Established — And What It Could Not

Netsmart Technologies (2007). In 2007, the Delaware Court of Chancery examined the sale process by which Netsmart Technologies, a small public company, was sold to a private equity buyer. The court's concern was not simply whether the board made the right decision, but whether the process by which it reached that decision could be reconstructed from the board's own records. According to legal commentary quoting the opinion directly, it could not, in critical respects: the meeting at which directors "claimed to have actually decided to sell" was informal and was never memorialized in minutes at all, and minutes for ten subsequent meetings spanning four months of deliberation were not approved until after litigation had already been filed. The same commentary quotes then-Vice Chancellor Strine's own characterization of the board's minute-keeping practices as "not confidence-inspiring" — and separate commentary on the same opinion notes that the court went so far as to "cast doubt on the veracity of the company's record" itself.

What this establishes: in this specific, well-documented instance, where a company's own record did not capture a decision its own directors later described as the actual moment of decision, the court examining that decision afterward had materially less to work with than the directors themselves had. The absence of the record became, itself, a fact the court weighed.

What this does not establish: that no genuine deliberation occurred in that informal, unminuted meeting. Neither the court nor this piece can determine that with confidence either way — and that is the point worth sitting with. A thin record does not prove thin thinking. It demonstrates only that, from the outside, thin thinking and thin documentation become difficult or impossible to distinguish. Nor does this case establish a general rule that every informal or unminuted meeting is legally problematic; it reflects the court's assessment of this particular sale process, not a universal minute-taking standard.

Walt Disney Company Derivative Litigation (2005, affirmed 2006). By contrast, when Disney's board was examined for its 1995 hiring and 1996 termination of Michael Ovitz — one of the most extensively litigated corporate governance matters in Delaware history — the Court of Chancery's own opinion describes reviewing twenty-four witnesses' testimony, thousands of pages of deposition transcripts, and more than a thousand trial exhibits. After this review, the court concluded, in its own words, that while the defendants' conduct "fell significantly short of the best practices of ideal corporate governance," the directors had not breached their fiduciary duties or acted in bad faith. The Delaware Supreme Court affirmed, holding that fiduciary duties "do not change over time," independent of evolving aspirational standards of practice.

What this establishes: even where a board's contemporaneous process was later criticized as falling short of best practice, a sufficiently developed record — assembled here substantially through litigation discovery and trial testimony rather than contemporaneous minutes alone — allowed the court to reach a considered judgment about what had occurred.

What this does not establish: that the volume or scale of the record is what protected the directors, or that comparably extensive records would produce the same outcome in a different case. The Supreme Court was explicit that directors are not held to an aspirational "best practices" ideal as the measure of liability. This case does not establish that large litigation records generally insulate boards from liability findings; it reflects what this particular, extensively developed record allowed this particular court to conclude.

Marchand v. Barnhill (2019). A sharper illustration of what a record can establish comes from an entirely different context: a 2015 listeria outbreak traced to Blue Bell Creameries, which killed three people. When the Delaware Supreme Court considered whether a derivative claim against the board's oversight of food safety could proceed, it did not need to determine whether internal testing had detected contamination — the record showed plainly that it had. Internal reports had identified listeria in company plants before the outbreak became public. The court's finding, central to allowing the claim to proceed past a motion to dismiss, turned on a narrower fact: the board's own meeting minutes reflected, in the court's words, "no board-level discussion" of those reports at all. The court further rejected the directors' argument that general operational discussions with the board were sufficient to satisfy this requirement.

It is important to be precise about what this decision resolved and what it left open. Chief Justice Strine, writing for a unanimous court, began the opinion by describing Caremark claims generally as "difficult to plead and ultimately to prove out" — "possibly the most difficult theory in corporation law upon which a plaintiff might hope to win a judgment." The court held that directors "have great discretion to design context- and industry-specific approaches," but that Caremark's core requirement is that the board make "a good faith effort — i.e., try — to put in place a reasonable board-level system of monitoring and compliance." The Supreme Court's ruling permitted the claim to proceed to further litigation; it was not a final adjudication that the board had, in fact, breached its duty of oversight. That question remained for further proceedings, not this opinion.

What this establishes: where information demonstrably reached management, the absence of any board-level record addressing it was, in this case, sufficient to support an inference relevant to allowing a derivative claim to proceed past a motion to dismiss.

What this does not establish: that the board's oversight was, as a final matter, actually deficient, or that every case involving a thin board record will similarly survive dismissal. The court's own description of how difficult these claims are to ultimately prove should be read alongside its willingness to let this particular claim proceed.

TD Bank OCC Consent Order (2024). Outside the courts entirely, federal banking regulators have arrived at a related but distinct requirement. In its 2024 consent order resolving anti-money-laundering violations at TD Bank, the Office of the Comptroller of the Currency did not merely require the bank to fix its compliance program. The order's own text requires, as a specific and enforceable term, that the bank's board document its review of the required remediation plan in official meeting minutes, with a copy furnished to the OCC's Examiner-in-Charge within thirty days of that review.

What this establishes: a federal regulator, operating independently of Delaware corporate law, arrived at a related expectation in this specific order — that board-level review of a consequential, ongoing regulatory matter should be reflected in the board's own contemporaneous record.

What this does not establish: that documentation alone satisfies the underlying substantive remediation obligation, or that this specific documentary requirement — crafted for this specific consent order and this specific violation — generalizes as a standard applicable outside its own terms. The order requires both the substantive remediation and its documentation as separate, cumulative obligations.

The Distinction

Read together, these four sources — two shareholder derivative matters, one oversight-doctrine case, and one federal enforcement order — do not all say the same thing, and it would overstate the evidence to collapse them into a single rule. But they do converge on a distinction worth stating precisely.

Record existence is not the same as decision reconstruction.

A record can exist — a minute can be taken, a memo can be filed, an email can be sent — without preserving enough of the decision's actual basis to allow someone examining it later to understand what leadership knew, what it weighed, who held the authority to decide, and why the decision came out the way it did. Netsmart had minutes for some meetings; it did not have a record of the meeting its own directors described as the actual point of decision. A record's mere existence does not answer the harder question of whether it preserves the decision's reconstructable basis.

Conversely, the absence of a pristine, contemporaneous record does not, by itself, prove that no genuine deliberation occurred. Disney's directors were found to have fallen short of ideal governance practice, and the court still found no breach of duty — because enough of the underlying substance could be reconstructed through other means, even without a complete real-time paper trail.

What the evidence shows is narrower and more useful than a rule that says "always keep better records." It shows that the specific, reconstructable connection between a decision and its documented basis — not documentation volume for its own sake — is what later examination has, in these cases, actually depended on.

Where the Evidence Stops

It is worth being explicit about the limits of what this comparison can responsibly support.

This is not evidence that poor documentation automatically establishes a breach of fiduciary duty. Marchand's holding permitted a claim to proceed past a motion to dismiss; it did not itself resolve, as a final matter, whether the board's oversight was deficient, and the same court described claims of this kind as exceptionally difficult to ultimately win.

This is not evidence that any particular volume or format of documentation guarantees defensibility. Disney's record was extraordinarily extensive and assembled substantially through litigation rather than contemporaneous practice, and the court was explicit that Delaware law does not impose an aspirational "best practices" standard as the measure of liability.

This is not evidence that every consequential decision requires the same documentation standard. A board's oversight of food safety at a company where food safety is admittedly central to the business is a different fact pattern than a merger negotiation, which is different again from a banking remediation plan under active regulatory supervision.

This is not evidence that courts require exhaustive contemporaneous records for every decision. The Disney matter shows a case where a record built substantially after the fact was nonetheless sufficient for the court to reach a considered judgment.

And this is not a claim that a strong record, once created, resolves every future question about a decision. A record narrows the space of what can later be reconstructed. It does not eliminate the underlying uncertainty about what happened in a room that no later examination can ever fully re-enter.

What Remains for the Executive to Consider

None of this reduces to a documentation checklist, and this piece does not attempt to supply one. The specific standard that would satisfy a Delaware court examining a listeria outbreak is not the same standard that would satisfy an OCC examiner reviewing an anti-money-laundering remediation plan, and neither is necessarily the standard appropriate to a decision your organization is making today, in a context these cases did not address.

What the evidence does support is a question, not a prescription: before a consequential decision is considered closed, is enough of its basis — what was known, who decided, and why — actually visible in a form that would survive the decision-makers' own departure from the room? That question does not have a universal answer. Delaware's courts, examining fact patterns as different as a private equity sale and a fatal product contamination, arrived at different conclusions about what the record needed to show. What they share is not a common documentation standard but a common method: when the decision was later examined, the record — whatever it consisted of — was what stood between the moment of judgment and everyone who tried to understand it afterward.

Close

A decision, once made, does not stay in the room where it happened. It travels forward as whatever was written down, and it is that traveling record — not the original deliberation itself — that any later examination actually encounters. Netsmart's directors may have deliberated carefully in an informal meeting that was never minuted; no one examining that decision today, including this piece, can know for certain. That uncertainty is not a flaw in the analysis. It is the condition itself: the record is what remains, and what remains is not always what happened.

The question worth carrying forward is not how much to document. It is whether, for this specific consequential decision, enough of its basis will still be visible after the room has emptied — to someone who was never in it.

—