Startup Valuation - 8 Methods, Stage Aware

Turn one set of figures into a defensible valuation range, the assumptions that carry it, and a value bridge showing which lever moves the number by how much.

No API keys, no LLM calls, no data sent anywhere. The arithmetic is deterministic: the same inputs always produce the same range, and every method shows what it was computed from.

What it does

A pre-seed company and a profitable one do not get valued the same way, so the Actor picks the method set from the stage you give it and weights the results accordingly.

Stage	Methods used
Pre-seed - idea or prototype, no revenue	Berkus, Scorecard, risk factor summation
Seed - product live, first revenue	Scorecard, risk factor summation, VC method, revenue multiple
Early growth - repeatable revenue	VC method, revenue multiple, EBITDA multiple, DCF
Growth - profitable or near break-even	Revenue multiple, EBITDA multiple, DCF, adjusted net assets

Eight methods are implemented in total: Berkus, Scorecard (Payne), risk factor summation, VC method, revenue multiple, EBITDA multiple, DCF, adjusted net asset value.

A method whose inputs are missing is reported as not computed — never silently scored as zero. Its weight is redistributed across the methods that were computed, and the report says which ones dropped out and why. The effective share column tells you how much each method actually carried.

Enterprise value is not equity value. Multiples and DCF produce enterprise value; the report bridges to equity value line by line (+ cash − financial debt) in every method detail.

The early-stage anchor. Berkus, Scorecard and risk factor summation are all relative methods — they scale a single anchor: the median pre-money valuation of comparable rounds in the same sector, stage and geography. Without an anchor those three are not computed and nothing is guessed.

What you get

Every run writes these to the run's key-value store:

File	What it is
value-card.md	One-screen summary: range, mid value, method-by-method figures, round maths
report.html	Full report — method comparison, every method's detail, DCF and sensitivity, value bridge, inputs, limits
report.pdf	The same report, print-ready
table.xlsx	Summary, methods, DCF, sensitivity, value bridge, inputs and the parameter set, as a spreadsheet
data.json	Everything above as structured data
input.json	The exact input the engine used

The dataset row carries the low / mid / high value, each method's result and effective share, confidence, post-money and investor stake, the one-line diagnosis and the full value bridge — so you can chain this into a sheet, a database or another Actor.

Input

Tick Run with sample data for a full example run before you enter anything.

Otherwise fill Valuation data. Field names can be English or Turkish. Anything you do not have, leave as null.

{
  "company": "Example Technology Inc.",
  "period": "Last 12 months - end of June 2026",
  "currency": "USD",
  "stage": "seed",
  "sector": "saas",
  "anchorValuation": 25000000,
  "financials": {
    "netSales": 4800000,
    "grossProfit": 3600000,
    "ebitda": -900000,
    "netProfit": -1250000,
    "cash": 2500000,
    "financialDebt": 600000,
    "equity": 1900000
  },
  "growth": { "priorYearNetSales": 2100000 },
  "scorecard": {
    "team": 1.2, "marketSize": 1.15, "product": 1.1, "competition": 0.9,
    "salesChannel": 0.95, "additionalInvestment": 0.9, "other": 1.0
  },
  "riskFactors": {
    "management": 1, "stageRisk": 0, "regulation": 0, "manufacturing": 1,
    "sales": -1, "funding": -1, "competitionRisk": 0, "technology": 1,
    "litigation": 0, "international": -1, "reputation": 0, "exit": 1
  },
  "vc": {
    "exitYear": 5, "exitRevenue": 120000000, "exitMultiple": 4,
    "targetReturnMultiple": 12, "dilution": 30
  },
  "dcf": { "revenueGrowth": 60, "targetEbitdaMargin": 22 },
  "round": { "roundSize": 6000000 }
}

Notes:

• stage — pre-seed, seed, early growth or growth. When in doubt pick the lower one; a higher stage inflates the value.
• sector — general, saas, marketplace, e-commerce, fintech, deep tech, services, manufacturing, health. It selects the revenue and EBITDA multiple bands; position inside the band is set by growth (60%) and margin (40%). Supply multiples.revenueMultiple or multiples.ebitdaMultiple to override.
• anchorValuation — the median pre-money valuation of comparable rounds. Publicly announced figures are usually post-money; deduct the round size first. If you cannot source it, leave it null.
• berkus — five criteria, 0-100 each. scorecard — seven factors, 0.50-1.50, where 1.00 means "same as the median company". riskFactors — twelve items, -2 to +2, where 0 is neutral. Score against evidence, not adjectives.
• dcf — give freeCashFlow as five years directly, or just revenueGrowth and targetEbitdaMargin and let the engine project it. Tax, capex, working capital and D&A ratios fall back to defaults (25% · 3% · 5% · 3%); a blank discount rate falls back to the stage default (24%-45%).
• round.roundSize — supply it and post-money and investor stake are computed. Without it the VC method result is flagged as post-money.
• Amounts must all be in the same currency. Use plain numbers.
• Instead of the JSON field you can pass Input file URL — a public link to a JSON or CSV file with the same data.

Report language

language: "en" or "tr". Every output — value card, HTML, PDF, spreadsheet — is produced in the chosen language.

Typical uses

• Founders — walk into a round with a range you can defend and the three assumptions that carry it, instead of a single number from a napkin.
• Investors and angels — screen inbound decks consistently: the same method set, the same weights, every time.
• Accelerators — value a whole cohort on one rubric and see where each company sits.
• Advisors and accountants — produce a client-facing valuation report from figures you already have.

Cost

The Actor is free to use; you pay only Apify platform usage for the run. A run takes roughly 15-40 seconds, most of it PDF rendering. Turn Generate PDF off for the cheapest possible run.

Limits and honesty

• A valuation is a range. The single mid value is a defence line, not a price — the market sets the price.
• The multiple bands in the parameter set are documented defaults, not market data. They ship in table.xlsx so you can disagree with them explicitly and replace them with your own transaction evidence.
• The report is for information only: it is not an audit, a statutory or regulatory valuation report, a legal opinion or investment advice.
• Garbage in, garbage out. A DCF built on a plan nobody believes produces a confident-looking number that means nothing — which is why the report shows the terminal value share and warns above 75%.

About

Built by Bloom Consulting, a fractional CFO practice working with startups and SMEs. The paid desktop edition adds valuation-over-valuation comparison, your own multiple bands and weights, and branded report templates.

• Web: bloomconsulting.com.tr
• Contact: iletisim@bloomconsulting.com.tr

Found a bug or want a method added? Open an issue on the Actor page.