backable

Seven free tools for founders, a plain-language term sheet glossary and the startup numbers that are actually traceable to a source, in one Next.js app — startup-funding-planner.vercel.app

• Bootstrap — whether you need outside money at all, in two parts. First eight conditions from the bootstrapping literature — time to first revenue, gross margin, capital needed before the first sale, payment terms, how fast the market is being taken, whether founders can sell it, how long they last unpaid, and what the company is meant to become. Three of them rule it out on their own rather than costing points. Then the same question in cash: capacity caps who can be served, hiring waits for money that already exists, and the plan either funds itself or names the amount and the month to start raising.

• ARR & MRR — what your recurring revenue really is. Normalises monthly, quarterly and annual plans onto one monthly number, then builds the bridge — starting MRR plus new plus expansion minus contraction minus churn — and reports gross and net retention on the base that existed at the start of the month, not on the base after new sales flattered it. It also shows the gap between ARR taken as committed recurring revenue and the number you get by annualising everything that came in, which is the gap investors ask about.

  A second view models land and expand in ACV, for software that starts with a paid project and grows into a platform licence. Pilots are won at a rate that itself accelerates, a cohort of them steps up after a lag, and the licence carries an ACV, a billing cycle and a term — so total contract value and annual contract value stay apart, annual contracts churn at renewal rather than continuously, and billing annually upfront produces a deferred revenue balance you can see. Whether the first engagement is a one-off project or a small subscription decides whether an upgrade is new ARR or expansion, which moves net retention enormously for the same business; the model computes it both ways and shows the two numbers side by side.

• Ready to Raise — an investor-readiness board. Tick what you already have; it scores five readiness dimensions from the fundraising literature, names your binding constraint, and ranks what to fix by leverage. Also ranks which investor archetypes to target.

• Round Planner — how much to raise and what it buys. Sizes the round from the milestone the next one expects, in either direction: bottom-up from a team plan, or top-down from a revenue, user or product target. Shows both and the gap between them, with a timeline of when to start raising again.

• Dilution Planner — model dilution across FFF, Angels & VCs, handle grants and convertibles, build a cap table, and simulate exits under different liquidation preferences.

• Co-Founder Split — who should get how much, by the weighted-factor method (Demmler's Founders' Pie) rather than by handshake, plus a vesting simulation showing what each founder keeps and forfeits if they leave in a given month. Around three quarters of teams split equally; the research finding is not that this is wrong but that arriving there by default rather than by argument made teams markedly likelier to be unhappy later and less able to change it.

• Network Strategy — a weighted three-layer model (founder attributes → mechanisms → strategies) built from the empirical fundraising literature, each strategy with a step-by-step playbook.

Plus two sections that are not tools:

• Infos — the vocabulary of a term sheet in plain words, each entry answering what it is, why it exists and what it does to you, with the worked numbers where prose would be vaguer than arithmetic.
• Stats — survival rates, how concentrated venture outcomes are, what a prior success does and does not predict, and the gap between a unicorn headline and what the common shares are worth. Every figure traces to a paper or a government series and carries what it measures — including the survival series, which counts every new employer establishment and is not a startup failure rate however often it is quoted as one. There is deliberately no unicorn map and no founder-archetype chart: those counts come from commercial trackers whose inclusion rules nobody can check, and the section says so.

Each model states its weights openly and cites the papers behind them. The weights are set by hand from the literature, not fitted to data — the output is a structured argument, not a prediction.

Running it

npm install
npm run dev

Script	Does
npm run dev	Dev server
npm run build	Production build
npm run typecheck	tsc --noEmit
npm test	Model unit tests, via Node's own runner

Copy .env.example to .env.local for the optional keys.

Layout

app/
  page.tsx                  shell + footer
  globals.css               design tokens, shared component styles, print rules
  api/milestones/route.ts   server-side milestone suggestions
components/
  SiteHeader.tsx            wordmark and the one menu
  bootstrap/                bootstrap tool UI
  split/                    co-founder split UI
  infos/                    term sheet glossary cards
  stats/                    the data board
  recurring/                ARR & MRR calculator
  ScrollHero.tsx            scroll-driven word hero
  PillNav.tsx               tool switcher (dropdown on mobile)
  ToolSwitcher.tsx          shell: header, hero, seven mounted tools, glossary, stats
  MotionButton.tsx          circle-that-grows-into-a-pill link
  SiteFooter.tsx            GitHub links + disclaimer
  ready/                    ready-to-raise UI
  round/                    round planner UI
  planner/                  dilution planner UI
  network/                  network strategy UI
lib/
  months.ts                 month arithmetic, shared by the planners
  bootstrap.ts              the eight conditions, weighted, with sources
  bootstrap-cash.ts         capacity, hiring gates, the capital verdict
  glossary.ts               term sheet vocabulary
  stats.ts                  the quoted figures, their sources, and what is left out
  recurring.ts              MRR normalisation, the bridge, GRR/NRR
  land-expand.ts            pilots, ACV, renewals, deferred revenue
  equity-split.ts           founders pie, vesting, the warnings
  planner.ts                all planner maths — pure, no DOM
  network-strategy.ts       weights, citations, and the two-layer model
  ready-to-raise.ts         readiness dimensions, actions, archetypes
  round-planner.ts          round sizing, capacity model, timeline
  inplicit/                 revenue planner: model, roster, share codec

All the models live in lib/ as pure functions over a state object, with no DOM access, so they can be exercised without a browser.

Unlisted route

/inplicit is a company-specific revenue planner rather than one of the public tools. It has two views of the same company:

• Ziel — top-down. Set a target ARR; it divides back to the customers needed, compares the three pricing models (per interview, per campaign, annual licence) and reports CAC, LTV, LTV:CAC and payback. One chart shows Land, Expand and Channel as stacked layers, so the gap between two lines is that motion's contribution and each crosses the target on a different month. Plain arithmetic throughout — lib/inplicit/goal.ts.
• Monatsplan — bottom-up from the named pipeline, with delivery capacity and the financing phase switch — lib/inplicit/plan.ts.

It has no access control: it is excluded in robots.txt, absent from sitemap.xml, carries noindex and is not in the tool navigation — so search engines skip it, but anyone with the URL can read it.

lib/inplicit/roster.ts therefore holds placeholder rows only — it ships in the client bundle and lives in a public repo, so treat everything in it as published. Real account names go in through the tool itself and persist in the share link and in the browser's saved scenarios, never in the repository.

Optional: AI milestone suggestions

The "Suggest with AI" button calls /api/milestones. Set

OPENAI_API_KEY=sk-...

in .env.local (or as a Vercel environment variable) to enable it. Without a key the route returns a curated fallback set rather than failing, so the feature degrades quietly. The key is only ever read server-side.

Deployment

Deployed on Vercel. vercel.json pins framework: "nextjs" so the build runs next build regardless of the project's dashboard preset.