A founder preparing for a round rehearses the product, the numbers and the team. Those are what the meetings are about, and they decide whether an investment committee becomes interested at all. But interest is not a signature. Between the two sits a stretch of weeks in which nobody discusses the product, because the file has moved to another room, where a lawyer has opened the corporate documents.
That lawyer is not evaluating the business. They are testing whether the thing the fund proposes to buy is capable of being bought: whether the shares exist as described, whether the company owns what it is said to own, whether anyone outside the room can block the transaction or unwind it afterwards, and whether the entity receiving the money is one the fund may fund at all. Founders are consistently surprised by how little of this is negotiable, and by how much of it had to be settled before anyone offered them money.
Two files, read by different people
Every round runs on two files at once. The commercial file is the one the founder knows: market, traction, unit economics, the reason this team wins. The corporate file is thinner, duller and read by someone with no interest in the story. It holds the incorporation documents, the register, the resolutions, the shareholder agreements and the assignments of intellectual property.
The two are weighted differently at different moments. Early on the commercial file decides everything. Once a term sheet exists it has done its work, and the corporate file takes over, because the deal team has made its case internally and now needs the transaction to be executable. Enthusiasm does not survive a defect that cannot be cured.
This is why founders rarely hear the real reason a round collapsed. Nobody withdraws saying the share register was inaccurate. They say the timing was wrong, or the valuation did not clear the committee. The commercial explanation is easier to give and harder to argue with, and the founder is left improving metrics that were never the problem.
Ownership as recorded, not as remembered
The first thing opened is ownership, and the first discrepancy is almost always between what the founders believe and what the record says. Founders carry a cap table that is partly a document and partly a memory of intentions. An early adviser was promised a small percentage. A co-founder left with an understanding about what they would keep. Options were granted by email and never in a resolution.
The lawyer does not read intentions. They read the register, the deeds, the resolutions and the filings, and compare them against the spreadsheet the founder sent. Anywhere the two disagree is a finding, and findings about ownership are the least tolerable kind, because a fund is buying a defined percentage of a defined thing and cannot accept ambiguity about what the denominator is.
The cure is uncomfortable in a specific way: it needs the cooperation of whoever sits on the other side of the discrepancy. A person who believes they were promised equity has to agree in writing that they were not, or has to be given something. Approached during due diligence, they will understand exactly why. Their leverage has just been created by the timing of the request.
Where the intellectual property actually sits
The second question is whether the company owns the product. It gets asked because the answer is frequently no, and the reasons are ordinary rather than sinister. The first version was written before the company existed. A founder built it personally and never assigned it. Contractors were engaged on agreements that said what would be delivered and nothing about who would own it, and in several jurisdictions the default without an assignment favours the contractor.
Then there is the group version of the problem. Intellectual property is held in one entity while development, revenue and staff sit in another, with no licence, no charge and no rationale. That is read twice, once as an ownership defect and once as a tax exposure, because value created in one place and owned in another attracts questions from authorities as reliably as from buyers.
What the fund wants here is dull: a clean chain of title running to the entity it is capitalizing, or to a subsidiary that entity controls, with assignments signed at the time rather than reconstructed afterwards. Retrospective assignments are possible, and they are visibly retrospective, which is a different conversation from having the document.
Vesting, and what its absence says
Founder vesting is treated by many founders as an indignity investors impose. It is closer to the opposite. Vesting protects the founders who stay from the founder who leaves in the first year holding a third of the company, and its main function is exercised between founders, not between founders and the fund.
Its absence is therefore read as information. A company without founder vesting is a company whose founders never had the conversation about what happens if one of them stops. The fund is not offended by this. It simply notes that the discussion is now going to happen during the round.
That timing is the cost. Imposing vesting at the round means renegotiating the relationship between founders at the moment they are least able to be relaxed about it, in front of an investor, against a deadline. It can also carry a personal tax consequence in some jurisdictions, because restrictions placed on shares already held are not always neutral. Agreed at incorporation, when the shares are worth nothing and nobody has anything to lose, it costs a conversation.
The money that arrived without a contract
Almost every company has funding that predates its first formal round. A founder put in personal money. A parent transferred an amount nobody characterized. An early believer wired funds against emails promising shares. None of it was documented properly, because at the time it did not feel like a transaction. It felt like help.
Each of those is a claim, and claims are what due diligence exists to find. The question is not only how much is owed, but what it is. Money recorded as a loan that was never going to be repaid is a capital contribution wearing the wrong clothes, and money recorded as capital that the contributor expects back is a liability the accounts do not show. Interest free lending between related parties invites its own scrutiny, since authorities expect related parties to price as unrelated ones would.
The hardest version is the person who believes they were promised shares and holds an email ambiguous enough to argue about. That is a contingent claim on the cap table. It will be covered by a warranty, which means the founders personally stand behind it, and an informal arrangement from years ago becomes a personal exposure at the moment it is discovered.
The parent has to be somewhere the money can go
Funds are not free to invest wherever they like. Their agreements with their own investors restrict what they may hold and where, and those restrictions are not negotiable by the company on the other side of the table. A structure whose parent sits somewhere the fund cannot invest is not a bad structure in the abstract. It is simply unfundable by that fund.
The constraints operate on several levels at once. There is what the fund may hold, and what its own tax position requires, since a fund that will one day sell needs to know that distributions and proceeds can move without being taxed at every step. There is what its banks and administrators accept on onboarding. And there is whether a future acquirer, years away, will find the parent acceptable, because an investor who cannot picture the exit will not picture the entry.
Changing the parent is possible. Moving a company between jurisdictions, or placing a new holding above an existing one, is standard work. Doing it while a term sheet is running is not, because it takes time the deal does not have and because the reorganization may itself crystallize value nobody budgeted for. The moment to ask where the parent should sit is before there is a reason to ask.
The consents nobody remembers granting
Structures accumulate people with rights. An early investor was given a veto over new share issues. A previous round carries pre-emption that must be waived or offered around. A shareholders agreement contains an anti dilution mechanism nobody has modelled since it was signed. A bank facility, a public grant or a significant customer contract carries a change of control clause.
Every one of those is a person outside the room who can delay the round, extract something for a signature, or in the worst case prevent it. They are not adversaries. They were given a right when giving it seemed costless, and are now entitled to use it. The discipline is knowing they exist before anyone needs them to sign. A founder who hands the fund a list of every consent required, with the position of each holder already established, converts a risk into a schedule. A founder who discovers on a Thursday that a former investor holds a veto has handed that investor the round.
What can be fixed before, and what cannot be fixed after
The distinction that matters is not between serious and minor problems. It is between problems the company can solve alone and problems that need a counterparty. Missing resolutions, unfiled documents, an out of date register, an option plan never formally adopted: these are administrative, and a competent adviser clears them without anyone’s permission. They are also the ones founders worry about most, because they are visible.
The ones that need a counterparty harden with time. Every day closer to a transaction increases the value of a signature that is required, and the counterparty is under no obligation to be reasonable about it. That is the whole argument for reading the structure in a quiet period rather than a live one. In a quiet period, the conversation with the person who thinks they own something is a conversation. In a live one, it is a negotiation with a deadline attached.
A third category cannot be repaired at all, only disclosed. A tax position taken years ago, a transfer of assets between related entities at a price nobody justified, an entity allowed to fall out of good standing while it held something important: these are facts. What stays negotiable is how they are presented, provisioned and priced, and presenting them yourself, early, with a documented view of the exposure, is a different event from having them found.
Founders tend to hear all of this as an argument for spending money on advisers before it is necessary. It is closer to an argument about sequence. Almost everything in the corporate file is cheap to get right while it is being created and expensive to correct afterwards, and the only unusual thing about a round is that it is the first moment anybody reads the file properly.