Seed Funding: How It Works, Examples, and Raise Calculator

Seed funding is early capital used to turn a working startup hypothesis into evidence that can support a larger round or a sustainable business. It usually pays for product development, initial hires, customer acquisition, and the proof points a company must reach before Series A.

For example, a B2B software startup might raise $1.8 million to fund 18 months of engineering, security work, and sales experiments. The important number is not $1.8 million itself. It is the amount required to reach a defensible milestone—such as repeatable customer acquisition—without selling more ownership than the milestone justifies.

Key takeaways

• A seed round should be sized from milestones, burn, runway, one-time costs, and cash already available.
• “Seed” describes the company's evidence and financing purpose more reliably than a fixed dollar range.
• Seed capital can come through a SAFE, convertible note, or priced equity round. Those instruments create different ownership, debt, governance, and legal consequences.
• Founders should model dilution and clean the cap table before signing, not after the round closes.

What seed funding is—and what it is not

Seed funding is often the first institutional capital a startup raises, although an experienced team may attract a seed VC earlier and a capital-intensive company may use several small rounds before calling one “seed.” The label is useful only when it communicates what has already been proved and what the new money is expected to prove next.

These are patterns, not legal definitions. A deep-tech startup may be seed-stage before revenue because technical validation is the relevant evidence. A low-cost software business may reach meaningful revenue before raising any institutional capital. Calling a round “seed” does not make it suitable; the operating plan and financing terms must still hold together.

Seed funding is also not automatically a loan. Whether it must be repaid, converts later, or buys shares immediately depends on the instrument used.

How much seed funding should a startup raise?

Start with the next milestone, then work backward to the capital required to reach it. A practical first-pass formula is:

Target seed raise = monthly net burn × runway months + one-time milestone costs − cash available for the plan

Where:

• Monthly net burn is monthly cash outflow minus reliable monthly cash inflow. Do not count unsigned contracts or an aggressive revenue forecast as cash.
• Runway months should cover the operating plan and leave time to measure results. If another round may be needed, the plan must also allow for preparation and fundraising before cash reaches zero.
• One-time milestone costs include items such as a prototype build, regulatory testing, security audit, tooling, launch inventory, or a critical senior hire.
• Cash available is unrestricted cash the company can actually allocate to this plan after near-term obligations.

The result is a funding need, not yet a final ask. Test it against four questions:

1. Does each major use of funds buy evidence? “Hire six people” is an expense list. “Cut implementation time enough to run ten paid pilots” is a milestone plan.
2. Can the company operate if revenue arrives later than expected? Build a downside case rather than solving the model with optimistic sales.
3. Can the ownership cost be justified? A larger round may reduce financing risk but increase dilution and raise expectations for the next round.
4. Can a smaller close still produce a coherent plan? If not, state the minimum viable raise and what changes below it.

Current data is useful as context, but not as a target. In a Carta sample published in July 2026, more than 1,000 software-company rounds completed in the prior six months showed a median seed raise of $4.1 million, a $24.3 million median valuation, and 18% median dilution. The sample excludes bridges and extensions and is heavily influenced by software and AI companies. Carta explicitly cautions that medians describe the middle of a wide range; they do not tell a particular founder how much cash the business needs.

A hardware company with certification and manufacturing costs may need more runway and one-time capital than a services-enabled software company. A capital-efficient startup with revenue may need less than the market median and retain more ownership. The model should explain the amount—not the stage label or a benchmark screenshot.

Three seed funding examples

These hypothetical examples show how the same formula produces different rounds. They are operating cases, not market quotations or fundraising promises.

Example 1: B2B SaaS company raising $1.8 million

The company has a working product, eight paid customers, and evidence that a narrow buyer segment converts. The round funds four engineering/product hires, a security audit, and a focused sales motion.

• Use of funds: improve reliability and onboarding, complete the audit, and test whether one sales playbook can acquire and retain customers efficiently.
• Next milestone: enough renewal, expansion, and sales-cycle evidence to show that customer acquisition can become repeatable.
• Main risk: hiring ahead of evidence. If the company builds a large sales team before the founder-led motion works, the raise buys burn rather than learning.

Example 2: Marketplace raising $1 million

The company has completed transactions in one city but does not yet know whether supply and demand can remain balanced without heavy subsidies. The round funds local operations, trust and safety, and measured acquisition experiments.

• Use of funds: reach reliable liquidity in one market, improve repeat usage, and understand contribution economics.
• Next milestone: a dense market where transactions repeat and subsidies decline without damaging service quality.
• Main risk: expanding geographically before solving liquidity. More cities can make topline metrics look larger while weakening the core model.

Example 3: Deep-tech company raising $4.6 million

The company has laboratory proof but needs a production-representative prototype and third-party testing. Its longer development cycle makes a short runway impractical.

• Use of funds: engineering, specialist hires, fabrication, testing, and customer validation with design partners.
• Next milestone: independently validated performance plus a credible path to pilot production.
• Main risk: treating technical completion as commercial proof. The plan needs customer and manufacturing evidence, not only a successful test.

The amount differs because the work differs. A good example makes the assumptions visible so a founder or investor can challenge them line by line.

Who provides seed funding?

The right capital source depends on check size, speed, sector expertise, governance expectations, and how much support the team needs after closing.

Choose investors as part of the financing design, not after deciding the amount. A founder who needs regulated-industry introductions should value relevant expertise differently from a company that mainly needs a fast, clean close.

The same distinction matters when evaluating the round from the investor side. Angels invest their own capital; VC firms invest a fund raised from limited partners and usually follow a defined mandate. The angel investor versus venture capitalist comparison explains the decision processes in more detail.

Founders building an investor map can use the Venture Capital Careers companies directory to identify firms, then verify each firm's current stage, sector, geography, check size, and portfolio directly before outreach. A logo list is not an investor strategy.

SAFE, convertible note, or priced equity?

Seed funding describes the purpose and stage of the capital. The financing instrument determines what the investor receives.

Y Combinator's official SAFE documents include three US post-money forms and a user guide. YC emphasizes two features: ownership sold is more immediately calculable than under the original pre-money SAFE, and the standard form has no interest or maturity. It also warns that a SAFE is not suitable for every situation and recommends legal review in the company's jurisdiction.

Use the instrument that matches the transaction, not the one with the shortest template. A SAFE can reduce closing friction but does not eliminate valuation economics. A note can defer share pricing but remains debt until conversion. A priced round costs more to execute but may create cleaner governance and ownership clarity for a substantial institutional financing.

For the mechanics behind each choice, see the VCC explainers on SAFEs, SAFE versus convertible note, and term sheets. Founders should have qualified counsel review the actual documents; this comparison is not legal or tax advice.

What dilution looks like in a seed round

Dilution is the reduction in an existing holder's percentage ownership when new equity or equity-linked securities are issued. It is not automatically bad: selling part of the company can create more value than preserving a larger percentage of an underfunded business. The mistake is accepting dilution that the company has not modeled or that the milestone does not justify.

Simplified post-money SAFE example

Assume a startup raises $1 million on a post-money SAFE with a $10 million valuation cap and no discount. In the simplest cap-only case:

Estimated ownership sold = investment ÷ post-money valuation cap

$1m ÷ $10m = 10%

That makes the immediate ownership economics easier to see, which is a core purpose of the post-money SAFE. It does not mean the investor will own 10% forever. The next priced round, new options, later SAFEs, and the actual conversion mechanics can dilute both founders and SAFE holders. If a lower-priced conversion or other terms apply, the result may differ.

Simplified priced equity example

Assume the company raises $2 million at an $8 million pre-money valuation:

This 20% result applies before considering any option-pool increase, converting notes or SAFEs, warrants, or other negotiated adjustments. If the term sheet requires a larger employee option pool to be created before the financing, existing holders usually absorb that additional dilution.

Model the transaction on a fully diluted basis before signing. A clean pro forma cap table should reconcile current shares and options, every convertible security, the financing, the post-close option pool, and each holder's ownership. The post-money valuation explainer covers the valuation formulas and round mechanics in more detail.

The practical decision is not “Is 10% or 20% normal?” It is “What ownership and control will every stakeholder have after all securities and negotiated changes are included, and does the new capital buy enough progress to justify that result?”

Is the startup ready to raise seed funding?

A startup is seed-ready when it can present a coherent chain from evidence to capital to milestone. Score each area as ready, partial, or missing.

Not every company needs revenue before a seed round. Evidence should fit the business model: retention for a software product, transaction density for a marketplace, technical validation for deep tech, or regulatory and clinical progress for a life-sciences company. The common requirement is that outsiders can examine the evidence rather than take the story on faith.

When not to raise yet

Pause and fix the foundation when any of these conditions is true:

• The team cannot name the milestone the round is buying.
• Runway works only if uncontracted revenue arrives on schedule.
• The cap table cannot be reconciled.
• The investor pitch contains no evidence beyond founder conviction.
• The financing instrument is being chosen without understanding ownership, repayment, conversion, or control consequences.
• The company wants capital mainly because competitors raised, not because it has a high-value use for the money.

Waiting is not always safer—some products require capital before meaningful proof is possible. But the founder should be able to explain why capital unlocks the next evidence and why the proposed round is the least expensive credible way to obtain it.

How the seed funding process works

Treat fundraising as a sequence of decisions with an artifact at each step.

1. Define the next milestone. Write the operating claim the company must prove, the metric or evidence that will prove it, and the date by which it should be visible.
2. Build the cash model. Calculate base and downside burn, one-time costs, current cash, minimum close, target raise, and the actions required if the round closes below target.
3. Clean the cap table. Reconcile founder shares, grants, promises, options, SAFEs, notes, warrants, and any side letters. Model the proposed round and option pool.
4. Assemble the evidence. Prepare a concise deck, product demonstration, operating model, customer or technical evidence, incorporation records, intellectual-property records, and a controlled data room.
5. Map investors by fit. Filter for stage, sector, geography, check size, ownership target, lead/follow behavior, portfolio conflicts, and ability to support the next milestone.
6. Run meetings and diligence. Track questions, refine weak evidence, manage references, compare complete terms rather than valuation alone, and keep enough momentum to make a real decision.
7. Close, allocate, and report. Complete legal work, update the cap table, assign budget owners, set milestone gates, and establish an investor update cadence.

The process is iterative. An investor objection may reveal a missing customer test; diligence may expose a cap-table issue; a smaller first close may require a narrower hiring plan. What should not change casually is the connection between capital and milestone.

Before signing, review the full term sheet, not only the valuation cap or pre-money valuation. A seed investor may also examine many of the same areas covered in the VCC venture capital due diligence checklist: market, product, customers, team, legal structure, financial model, and cap table.

How a seed investor reviews the round

At seed, the evidence is incomplete by definition. The investor's job is not to pretend uncertainty has disappeared. It is to decide whether the team has found a valuable problem, learns unusually fast, and can turn this amount of capital into a much stronger company.

For a founder, this table is a preparation tool. For an analyst or associate, it is a first-pass investment memo structure. It prevents a common mistake on both sides: spending most of the discussion on the market size while leaving the use of funds, ownership, and milestone logic vague.

A strong seed memo should end with a falsifiable statement: If the company receives this capital, what evidence should exist before the money runs low, and what would make the investment thesis wrong?

What happens after seed funding?

Closing converts a financing plan into an operating obligation. The company should translate the round into monthly budget ownership, hiring gates, and a small set of milestone metrics. Investor updates should distinguish outputs—features shipped, people hired, meetings held—from outcomes such as retention, technical performance, revenue quality, transaction density, or verified customer demand.

Founders should preserve enough runway to respond when evidence arrives later than planned. That may mean delaying a hire, narrowing a market test, or killing a product line. Staged spending is not a lack of ambition; it is how a seed company protects the ability to learn.

Series A is one possible next step, not an entitlement. Some companies become sustainable, sell, use debt or strategic capital, extend the seed stage, or close. A company should approach Series A funding when the seed plan has produced evidence that additional capital can scale—not merely continue—the model.

Seed funding FAQs

Is seed funding a loan?

Not necessarily. A priced equity round sells shares. A SAFE is a contract for future equity and the standard YC form has no interest or maturity. A convertible note is debt intended to convert into equity and normally has interest and a maturity date.

Do founders pay seed funding back?

Equity is not repaid like a loan, but investors own part of the company and may have negotiated rights. A convertible note can create repayment or maturity consequences if it does not convert. The documents control the answer.

How much equity should a startup give up in a seed round?

There is no universal correct percentage. Current market data can provide context, but founders should model every security, option-pool change, and later dilution, then ask whether the round buys enough progress to justify the ownership sold. Carta's July 2026 software-only sample reported 18% median dilution, but that is a descriptive median—not a recommendation for a particular company.

Can a startup raise more than one seed round?

Yes. Companies use seed extensions, multiple SAFE closes, or more than one named seed round. Each additional financing should still have a defined purpose, and stacked convertibles must be modeled together.

How long should seed funding last?

Long enough to reach and measure the next milestone with room to adjust. The correct runway depends on sales cycles, product development, technical or regulatory work, hiring, and financing risk. A founder should model both the operating plan and a downside case rather than rely on one universal month count.

What comes after seed funding?

Usually a period of product, customer, and team development. If the company proves a scalable model, it may pursue Series A. Other outcomes include profitability, another seed financing, strategic funding, acquisition, or closure.

Make the milestone earn the round

The strongest seed plan is simple to interrogate: this evidence exists today; this capital funds these actions; these actions should produce this milestone; and the resulting ownership is acceptable under the base and downside cases.

Before contacting investors, build the cash model, reconcile the cap table, and write the one-sentence milestone the round must buy. If those three pieces do not agree, the financing is not ready—regardless of what similar companies have raised.