# Fundraise Readiness > Fundraise Readiness - Skill: `ingridleiria/fundraise-readiness` (Agent Skill) - Install (CLI): `npx skillmds@latest add ingridleiria/fundraise-readiness` - Raw SKILL.md: https://api.skillmd.com/api/skills/ingridleiria/fundraise-readiness/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: ingridleiria (https://skillmd.com/u/ingridleiria) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/ingridleiria/fundraise-readiness --- # Fundraise Readiness Rounds are rarely lost on the commercial story. They stall on a cap table nobody can reconcile, a metric defined one way in the deck and another in the model, contracts that cannot be found, an early contributor with an undocumented claim to equity, and a founder who answers a diligence question differently on Tuesday than on Friday. The cost is not the work of fixing those things. It is that they surface at the point of maximum leverage against the company. A cap table problem found in week two of preparation is an administrative task; the same problem found under a signed term sheet is a negotiation, and the counterparty now knows something about how the company is run. Delay in a financing is not neutral either: momentum is what keeps a competitive process competitive, and a process that goes quiet for three weeks while a company assembles its contracts often restarts at a lower price with fewer parties. ## When to use this, and when not to Use it eight to twelve weeks before going out to raise, and use it immediately if a process has already started without it. Use it for the same preparation ahead of a debt facility, an acquirer's diligence, or a secondary sale, since the document set and the discipline are the same and only the emphasis differs. Do not use it to build the argument for the raise. `pitch-deck` does that, and it should be built from the metrics pack this skill produces rather than the other way round, because a deck written first will contain numbers the pack cannot support. Do not use it for the periodic written update to existing shareholders, which is `investor-update`. Those updates are, however, diligence material: a series of them with drifting definitions is one of the first things a thorough investor notices, so fix the definitions here and keep the future updates consistent with them. Do not use it for the board pack, which is `board-deck`, or the relationship with existing directors and investors, which is `board-and-investor-management`. Both matter during a raise and both produce documents an investor may ask to see, but neither is diligence preparation. Do not use it to build the financial model, which is `financial-model-builder`; this skill only checks that the model, the pack, the deck and the accounts tell one story. ## What you need before starting **The cap table, with every instrument.** Shares, options granted and promised, convertible instruments, warrants, and anything agreed verbally. Missing: rebuild it from the register and the board consents rather than from a spreadsheet someone maintains, and treat the difference between the two as the finding. **Statutory and management accounts.** Missing: get the last two years from the accountant first, because the reconciliation in step three cannot start without them. **The source system and owner for every metric.** Missing: name an owner per metric now. A metric with no owner will be calculated twice, differently. **Customer contracts and the churn history with reasons.** Missing: list what exists and what does not, and start the search early. Missing contracts are the most common cause of a stalled commercial diligence workstream. **Employment, contractor and agency agreements, with intellectual property assignment.** Missing: this is the item most likely to stop a round, so treat an incomplete set as urgent rather than administrative. **The previous round's plan and previous data room.** Investors will compare what was promised with what happened, more often than founders expect. Missing: reconstruct the prior plan from board packs, and prepare the comparison yourself. **The target timeline and the process owner.** Missing: propose a calendar working backwards from the first partner meeting and get it agreed, and name one person who owns the data room and one who owns the answers. ## The method 1. **Set the calendar backwards from the first partner meeting.** Preparation, then a partner meeting sequence with firms grouped so the process runs in parallel rather than sequentially, then diligence, then signing. Judgement: group firms so that no group is more than two weeks ahead of another, because a lead that arrives while other firms are three weeks behind cannot be tested against an alternative. 2. **Write the metric definitions before touching any document.** One page, one definition per metric, with the source system, the owner, and the treatment of the awkward cases: trials, discounts, annual prepayments, refunds, professional services, cancelled contracts inside a notice period. This page governs the deck, the model, the data room, the updates and every verbal answer. 3. **Reconcile the metrics to the accounts.** Take recurring revenue as defined and walk it to the revenue line in the management accounts, then to the statutory accounts. Where it does not reconcile, find out why now. A defensible difference explained in advance is fine; the same difference discovered by an analyst is a workstream. 4. **Run the corporate hygiene sweep in parallel, starting immediately.** Each item below takes weeks and none can be fixed under a signed term sheet. 5. **Build the data room index before the documents.** Structure it the way a diligence team reads rather than the way documents accumulated, so that every question they would otherwise ask has a folder. Then fill it, and mark the gaps visibly in your own copy rather than leaving them silent. 6. **Assemble, version and control access.** One index, one version of each document, dated, with an access log. Judgement on staging: open the corporate and financial sections at first diligence and hold customer contracts and detailed people data until a term sheet, unless asked earlier, in which case give them rather than appearing to withhold. 7. **Write the weakness list.** Every company has three or four: concentration, a cohort that churned, a flat quarter, a departure, a dependency on one channel or one person. Write each with the honest account and the prepared answer. A weakness named by the company reads as self-awareness; the same weakness found in diligence reads as concealment and reprices the round. 8. **Rehearse the questions out loud.** Not in writing. Have someone who is not the chief executive ask them, and repeat until two founders answering separately give the same answer with the same numbers. 9. **Check that one narrative runs through everything.** Write the two or three sentences that are the argument, then read the deck, the model, the pack, the last six updates and the metric definitions against them. Any document that contradicts the argument is either wrong or the argument is. 10. **Run the process with a single point of answer.** One person owns the data room, one owns the answers, and every question routes through them and is logged. Two people answering the same question independently is how a process acquires an inconsistency that nobody can later explain. ## Corporate hygiene, where rounds actually stall The cap table reconciles to the issued instruments and to the option plan, including anything promised verbally and never papered. Every founder, employee, contractor and agency has assigned intellectual property to the company in writing. Board and shareholder consents exist for every issuance and every option grant. Any adviser or early contributor with a claim, however informal, is documented and settled. Customer contracts have been checked for change of control and assignment clauses. Nothing material, no domain, no account, no registration, is held personally by a founder. Filings are current. Work these first, before the deck and before the data room, because they are the only items on the list that can stop a round outright. ## The questions to rehearse What actually drives growth, and is it repeatable. What happens to a cohort over twenty-four months. Why did those specific customers leave. What is the real sales cycle and win rate, measured rather than estimated. Which revenue is contracted and which is not. How dependent is this on one channel, one customer or one person. What breaks first if volume triples. Why has the thing announced last year not shipped. What did the last round's plan say, and how did the outcome compare. That final question is asked more often than founders expect, and the previous plan is usually sitting in the previous data room. ## Worked example **Situation.** A company of 45 people with about 3.4m of recurring revenue, planning to open a Series A process in ten weeks. Two founders, an outsourced finance function, and a product built partly by a contract development agency in the first eighteen months. **Task.** Be ready to withstand diligence from four firms running in parallel, with a data room open on the day of the first partner meeting. **Action.** Work started with the deck and a draft data room index, because those felt like the deliverables. Three weeks in, a routine check of the agency contract found no intellectual property assignment clause: the agency had built the original billing and provisioning code and, on the face of the contract, still owned it. That discovery reset the sequence. The deck was paused and the hygiene sweep, which should have been step one, ran first. Resolving it took five weeks: locating the original engagement letter, negotiating a confirmatory assignment with an agency that had no incentive to move quickly, and paying a fee to get it signed. Found in diligence instead, it would have arrived as a condition to closing, with the agency in a much stronger position. The metrics work produced a second finding. Recurring revenue appeared as 3.4m in the deck draft, 3.28m in the model, and 3.51m in the last two investor updates. All three were defensible: one included annual prepayments at full value, one excluded a discounted pilot cohort, one counted contracted but not yet live customers. The definition page settled on the narrowest basis, 3.28m, and the history was restated on that basis for eight quarters. Restating downward before going out felt wrong to one founder and was the right decision: an investor who found three numbers would have assumed the highest was chosen deliberately. The weakness list came to four items: a largest customer at 14 percent of revenue, a paid-channel cohort that had churned at nearly double the rate of the rest, a prior year plan of 4.9m against the 3.28m delivered, and a co-founder who had reduced their time commitment for six months without that being disclosed to existing shareholders. Rehearsal exposed a fifth problem. Asked separately why the paid cohort churned, the founders gave different answers: one said pricing, the other onboarding capacity. Both had evidence. They spent a day on the cohort data, concluded it was onboarding, and agreed a single answer with the numbers behind it. **Result.** The process opened four weeks later than intended. Diligence produced 60 questions across the four firms, of which the tracker showed 51 already answered by material in the room. The assignment issue, disclosed in the first partner meeting as a resolved item, was never raised again. The restated revenue figure was queried once, and the restatement note answered it. Two term sheets arrived within six days of each other, which was the point of grouping the firms. The delay was the cost of having started with the deck rather than the hygiene sweep. It was also the cheapest possible version of that lesson. ### A second scenario, where it goes differently The same company two years later receives an unsolicited approach from a strategic acquirer with a 30 day exclusivity request. There is no ten week runway. The sequence inverts. Exclusivity is negotiated down or refused until the hygiene position is known, because signing exclusivity with an unresolved corporate defect hands the counterparty a repricing lever with no alternative bidder to check it. In the first 72 hours, only three things are checked: cap table reconciliation, intellectual property assignment, and change of control clauses in the top ten customer contracts. Anything else waits. The metrics pack is not rebuilt. It is frozen as it stands, with a disclosure note listing the known inconsistencies, because in a compressed timeline a disclosed inconsistency is survivable and a silently corrected one looks like a restatement made under pressure. What changed is that time, not completeness, is the binding constraint, so the method reduces to the three items that can stop or reprice the transaction. ## Output **The metric definition page**, which governs every other document: | Metric | Definition | Source system | Owner | Treatment of edge cases | Last changed | **The data room index**, as folders: corporate, financial, commercial, people, legal, product and technology, with a one-line description of what each holds and a visible marker on anything not yet complete. **The weakness register**, internal only: | Weakness | The honest account | Prepared answer | Evidence to hand | Who answers this | **The reconciliation note**, one page, walking the headline metric to the management accounts and then to the statutory accounts, with each difference explained. **The diligence tracker**, live through the process: | # | Firm | Question | Owner | Answered by (document or person) | Date | Consistency check | ## Failure modes **Starting with the deck.** Recognise it when the hygiene sweep has not begun in week one. The deck is fast to build and the assignments are slow to fix; do them in that order and the round waits for the assignments. **Three versions of one number.** Recognise it by comparing the deck, the model and the last three updates before anyone else does. Settle on the narrowest defensible basis and restate the history. **A definition that changed midway through the history.** Disclose it as a definition change with both bases shown. Smoothed and then discovered, it is treated as something considerably worse. **Two people answering the same question.** Recognise it when a firm asks a question already answered and gets a different reply. Route everything through one owner and log every answer. **A weakness the company hopes will not come up.** It comes up. Write it into the weakness register with a prepared answer, and raise it first. **A data room organised by how documents arrived.** Recognise it when a diligence team asks where something is. Every question they have to ask is a delay and a small deposit of doubt. The same applies to unusual terms in a previous round: preferences, ratchets and side letters go into the corporate folder and into the first meeting, because a term sheet built on a misunderstood structure gets withdrawn. ## Edge cases **A process already under way with no preparation.** Do the reconciliation and the hygiene sweep in parallel with the meetings, disclose what is being fixed, and do not open the data room until the corporate section is clean. An empty folder is worse than a delayed one. **No formal accounts, at pre-seed.** Reconcile to bank statements and the invoicing system instead, and say that is the basis. Precision about what the number is beats a number with no basis. **A regulated business.** Add licences, permissions, regulatory correspondence and inspection history as a first-class data room section, and treat any open matter with a regulator as a weakness register item with counsel's input. **An investor who asks for something you will not give.** Individual customer contact details before a term sheet, or raw personal data. Say what you will provide instead and when. A reasoned limit is normal; silence looks like concealment. ## Quality bar - Every metric has one written definition, with an owner and a source, used in every document and every verbal answer. - The headline metric reconciles to the management accounts and to the statutory accounts, with each difference explained on one page. - The cap table reconciles to the issued instruments and the option plan, including anything promised verbally. - Intellectual property assignment is documented for every person and firm that touched the product. - The data room is indexed, versioned, access controlled, and its gaps are visible to the company before the first meeting. - The three or four weaknesses are written down with prepared answers, and are raised by the company first. - Two founders, asked separately, give the same answer with the same numbers to each rehearsed question. ## Adapting this to your context The defaults come from a venture Series A process in a jurisdiction with a share register, statutory accounts and an option plan. The discipline transfers; the document set changes. - **The eight to twelve week runway.** Set by a competitive equity process. A debt facility or a grant application can be shorter, an acquirer's diligence often is not, and the second scenario covers having none at all. - **The corporate hygiene list.** Cap table, share consents and option grants assume a company limited by shares. A partnership checks the partnership deed and admission records, a charity checks trustee minutes, restricted fund accounting and the register of interests, an owner-managed firm what is held personally by the owner. - **Intellectual property assignment.** The item most likely to stop a round here. Where the value sits instead in licences, accreditations or a leased site, that is the equivalent item and deserves the same first-week urgency. - **The metric definitions.** Recurring revenue, churn and cohorts suit subscription businesses. Substitute the two or three numbers your counterparty will test: backlog and realisation, occupancy, contracted grant income, funded places. - **What not to change.** One written definition per metric used in every document and every verbal answer, the headline number reconciled to the accounts, and the three or four weaknesses raised by the company first. ## Related skills `pitch-deck` builds the argument, and should be written from the metrics pack this skill produces rather than before it. `financial-model-builder` produces the model that must reconcile to the pack, and `revenue-forecast` the commercial projections inside it. `investor-update` is diligence material: its series of metrics must match the definitions fixed here. `board-deck` supplies the historical packs a thorough investor will ask for, and any inconsistency between them and the pack will be found. `board-and-investor-management` runs the existing investor relationships through the process, including the pre-wiring that precedes a difficult disclosure. `decision-memo` is the format for the decisions the process itself forces, such as which firms to group and whether to accept exclusivity.