A raise rarely dies from a bad answer in a meeting. It dies quietly, over about eleven weeks, from a decision made before the first email went out. Here are the ones we see most.
01. Starting cold
The first three investors you speak to will find the weakest part of your argument. This is useful and you should want it — but not from the fund you most wanted to lead. Spend the first wave on credible investors you are willing to lose, fix what breaks, and then go to the ones who matter with a story that has already survived contact.
02. Raising the wrong number
Founders anchor on dilution and land on a figure that funds eighteen months of progress but does not retire a single major uncertainty. You arrive at the next round having spent the money and proved nothing that changes the price. Size the round against the milestone that re-rates the company, not against the cap table arithmetic. If that number is uncomfortable, that is information about the plan, not about the number.
03. The wrong instrument
Equity is the most expensive capital available and it is routinely used to buy things that should never touch the equity line: GPUs, ground stations, tooling, launch inventory. Hardware with a resale market and a contracted revenue stream can very often be financed against itself. Splitting the ask reduces dilution and widens the investor set at the same time, and almost nobody does it early enough.
04. Sequential meetings
Taking investors one at a time, at whatever pace each prefers, guarantees you never have two offers in the same week. Without simultaneity there is no competitive tension, and without tension you are negotiating with yourself. This is the single largest avoidable value leak in private fundraising and it is entirely a calendar problem.
05. Burying the government revenue
Specific to space and defense, and endemic. The contract structure — program of record, ceiling versus obligated value, option years, appropriation risk — goes in an appendix or is described in a sentence, which reads to a serious investor as though the founder does not understand their own revenue. Put it in the main narrative and explain the mechanics as though your reader has never bought anything from a government, because most of them have not.
06. The unmodelled depreciation
Endemic on the AI side. The entire investment case rests on an assumption about the useful life of a piece of hardware, and that assumption is nowhere in the deck. Somebody in the investment committee will raise it, you will not be in the room, and whatever number they pick will be worse than yours. State it, defend it, and show the case where it is two years shorter.
07. Going out too late
A raise takes three to five months from first email to money received, and longer if the first wave sends you back to rebuild. Starting with seven months of runway means starting a negotiation in which the other side can see your position. Start at twelve. Every month of runway you hold is negotiating power, and it is the only kind that compounds.
08. Confusing interest with demand
A partner meeting is not demand. A second meeting is not demand. Diligence requests are not demand. Demand is a term sheet, and until one exists the correct assumption is that every conversation ends in nothing. Founders lose entire quarters keeping warm the eleven investors who were always going to pass, while under-working the two who were not.
The pattern underneath all eight is the same. Each is a process decision, made early, usually by default, by someone whose actual job is building the thing. None of them are failures of the technology, and all of them are cheaper to fix in week one than in month four.
Draft note written during the build of this site. Rewrite in your own voice before publishing.