Valuation for startups: 4 methods and when to use each.

Every startup needs a number. Investor asks, founder answers, and from there the conversation moves forward or stalls. The problem is the number isn't unique — it depends on method, assumptions and business stage. Presenting a poorly grounded valuation is the fastest path to losing credibility in the first meeting.

This article is a practical guide to the four methods most used in startups. No memorized formulas, no heavy theory — just what you need to defend a number with rigor.

Why startup valuation differs from mature company valuation

Mature companies generate stable, predictable cash flow. You take history, project forward, discount at the appropriate rate, and you have value. For startups, three things change the game:

• Short or non-existent history. You can't extrapolate three years of revenue when the company has been operating for eighteen months.
• Exponential growth in projections. Small variations in assumptions change valuation by orders of magnitude.
• Binary risk. Startups either scale dramatically or die. Linear models don't capture that distribution.

That's why specific methods exist for early stage — and why applying pure DCF to seed-stage is mostly an academic exercise.

Method 1 — DCF (Discounted Cash Flow)

The classic method. Project free cash flows for the next 5-10 years, calculate a terminal value, discount everything to present value using a discount rate that reflects risk. The result is the company's value today.

When to use

For startups with predictable recurring revenue and validated business model — generally Series A onward. In B2B SaaS with low churn and long contracts, DCF works well even with the company still operating at a loss.

Limitations

The terminal value's weight is enormous — often representing 70-80% of total valuation. Small changes in perpetual growth distort the result. Early stage, DCF tends to inflate valuation because it depends heavily on "if everything goes right".

How to present

Always with sensitivity analysis. Show valuation varying discount rate, perpetual growth and terminal margins. Serious investor will ask for that anyway — come prepared.

DCF isn't a number. It's a range. Presenting a single value is a sign of someone who hasn't yet understood the method.

Method 2 — Comparable multiples

Identify listed (or recently sold) companies similar to yours, extract relevant multiples (EV/Revenue, EV/EBITDA, P/E), apply to your company. Valuation is the average multiple times the equivalent metric.

When to use

When clear public comparables exist in the same sector and stage. B2B SaaS is the simplest case — dozens of listed companies and EV/ARR multiple is well defined. Marketplace, fintech and healthtech also have good comparables. Niche sectors or new models become more complex.

Limitations

"Comparable" is rarely truly comparable. Size, geography, margin, growth rate — everything affects multiple. Public companies trade with liquidity premium that private startups don't have. Applying raw multiple without adjustment is recipe for error.

How to present

List 5-8 comparables with justification. Show each multiple and calculate average and median. Apply illiquidity discount of 20-30% if comparing with listed. Indicate the final range, not a single number.

Method 3 — Precedent transactions

Variation of multiples method, but using sold companies or rounds rather than listed companies. The applied multiple is extracted from real transactions — generally more relevant for private startups.

When to use

When recent transaction history exists in similar companies in your region. Bases like Crunchbase and sector reports help. Particularly useful for Series A-B startups preparing rounds.

Limitations

Private transaction data is often partial. You know round value and check, but not full terms (liquidation preferences, anti-dilution, conditional vesting). Comparing cap table valuation with exit valuation is a common mistake.

How to present

Indicate source, date and context for each transaction used. If possible, separate by stage (Pre-seed, Seed, Series A) — multiples vary dramatically by stage. Don't use single transaction as reference: range matters more than the number.

Method 4 — Berkus / Scorecard

Qualitative methods for early stage. Berkus method assigns value (typically up to R$ 500k) for each of five factors: solid idea, working prototype, qualified team, strategic relationships, initial sales. Scorecard adjusts regional market median valuation based on relative scoring.

When to use

Pre-seed and very early seed, no revenue or irrelevant revenue. When quantitative methods don't make sense because there's no numerical base to project. Angels and accelerators use frequently.

Limitations

High subjectivity. Final number depends on who applies and weights assigned. Work more as "calibration" than rigorous valuation. Don't replace quantitative analysis when company already has traction.

How to present

Not as single method, but as complement. "Applying Berkus, we get R$ X. Triangulated with precedent transaction multiples in pre-seed, the range is R$ X-Y." Triangulation gives robustness.

Which method to use — quick guide by stage

There isn't a single correct method — there are appropriate combinations by stage:

• Pre-seed (no MVP or with MVP): Berkus + Scorecard. Quantitative doesn't make sense yet.
• Seed (validated MVP, first clients): Precedent transaction multiples + Scorecard. Triangulation is important.
• Series A (recurring revenue, clear metrics): Comparable multiples + precedent transactions. DCF starts making sense as sanity check.
• Series B+ (validated scale, positive unit economics): DCF + comparable multiples. Sensitivity analysis mandatory.

The most expensive valuation mistake

Presenting a single number as if it were absolute truth. Experienced investors know valuation is a range, not a point. When you say "my company is worth R$ 30 million", the investor thinks "30 with what assumption?". When you say "considering Series A multiples in B2B SaaS, comparable transactions over 18 months and DCF with 25% discount rate, the range is R$ 25-35 million — my initial proposal is the midpoint", you change the conversation.

Methodological rigor doesn't mean appearing pessimistic. It means being taken seriously.

What to do before reaching valuation

Before modeling any method, ensure your company has three foundations ready:

• Auditable financial history. If numbers don't reconcile, valuation collapses in due diligence.
• Clear unit economics metrics. CAC, LTV, payback, churn. Without these, no projection base.
• 5-year financial model with explicit assumptions. Each number needs traceable origin.

That preparatory work is what differentiates a startup that raises well from one that wastes time on rounds that don't close.