Pre-Seed Startup Metrics: What to Prove Before Raising (2026)

There is no universal set of pre-seed startup metrics that guarantees a fundraise. Some companies raise before revenue or even before a working product; others need meaningful usage or revenue evidence because of the market, founder profile, capital required, investor strategy, or current funding environment.

The useful question is not “What number do investors require?” It is “Which risks have we reduced, what evidence supports that claim, and what specific milestone will this capital fund?”

This guide helps you build that evidence using model-specific metrics, clear formulas, worked examples, and a pre-seed dashboard. It is educational, not investment, legal, accounting, or fundraising advice.

What pre-seed metrics are supposed to prove

At pre-seed, a metric is valuable when it helps answer one of six questions:

1. Problem: Does a specific customer experience a consequential problem?
2. Demand: Will that customer commit money, time, data, reputation, or workflow change?
3. Value: Does the product or service deliver the promised outcome?
4. Repeat: Does use or purchase recur at the interval implied by the product?
5. Reach: Is there an increasingly credible path to more of the same customer?
6. Execution: Does the team learn and ship against the important constraint?

Pre-seed evidence often combines numbers with qualitative records: customer interviews, pilot scopes, usage sessions, sales notes, product logs, letters of intent, contracts, and delivery observations. A large top-of-funnel number cannot repair a vague customer or undefined value event.

Carta’s April 2026 interviews with pre-seed investors illustrate how investor approaches differ: some companies at this stage have no revenue or MVP, while team, market, plans, speed to market, and signs of customer value can all matter. Treat such observations as market context, not thresholds or promises that a particular investor will fund you.

Start with an evidence map, not a benchmark list

Create one row for each major claim in the fundraising story:

Evidence quality should be explicit:

• Observed: product event, transaction, completed workflow, or documented outcome.
• Committed: payment, signed agreement, real data, scheduled implementation, or internal introduction.
• Reported: a customer describes recent behavior or consequence.
• Hypothetical: survey preference, future intention, compliment, or waitlist signup without more commitment.

Hypothetical evidence is not useless, but label it honestly.

Metrics by business model

The right metric depends on how value is created and how often the customer should receive it.

Define every metric with a numerator, denominator, eligibility rule, time window, source, owner, and exclusions. Otherwise a chart can improve merely because the definition changed.

Problem and demand evidence before product scale

Track the customer-learning funnel:

Relevant interview rate = completed interviews with the target participant ÷ qualified people contacted

Recent-problem rate = relevant interviewees describing a qualifying recent event ÷ relevant interviews

Commitment rate = qualified prospects making the defined commitment ÷ qualified prospects offered the test

The raw count and context matter more than the percentage in a small sample. Record who was included, why they qualified, what was offered, and what counted as commitment.

Useful artifacts include:

• Dated notes reconstructing a recent workflow.
• Evidence of current spending or resource use.
• A paid pilot or signed scope.
• Customer-provided real inputs.
• An introduction to the user, security owner, or budget holder.
• A second use without extraordinary founder persuasion.

Letters of intent vary widely. An unsigned expression of interest, a non-binding letter, and a contracted paid pilot are not equivalent. Describe the exact commitment and conditions.

Activation: define the first value event

Activation is not account creation. It is the earliest observable event that shows the customer received a meaningful slice of the promised value.

Examples:

• A finance lead uploads a valid file and resolves the first flagged exception.
• A manager invites a teammate and completes the first weekly planning cycle.
• A developer sends a successful production-like request and receives the required result.
• A marketplace buyer completes a qualifying transaction with a supplier.

Formula:

Activation rate = new eligible customers reaching the defined value event within the activation window ÷ new eligible customers

Worked example:

In June, 18 clinics began an onboarding pilot. Two were excluded under a prewritten rule because their required integration was not supported. Of the 16 eligible clinics, 11 uploaded a valid recall list and sent at least one patient-approved reminder within 14 days.

Activation rate = 11 ÷ 16 = 68.75%

That number is not “good” or “bad” without context. The useful follow-up is why five eligible clinics failed, whether the activation event predicts ongoing value, and whether the exclusions were legitimate rather than cosmetic.

Retention and repeat behavior

Retention must match the natural frequency of the job.

For a weekly workflow:

Week-4 eligible-account retention = accounts completing the core workflow in week 4 ÷ activated accounts eligible to use it in week 4

For a quarterly workflow, weekly retention would be meaningless. You might instead measure whether the next required cycle was completed and whether the customer renewed.

Use cohorts rather than combining every user across time. A simple table is enough:

Do not claim improvement from this illustration without examining sample size, cohort composition, reminders, product changes, and exclusions. The table shows the method, not a benchmark.

Also record why customers return or leave. A retention ratio can reveal a symptom; customer evidence explains the mechanism.

Revenue quality and unit economics

At pre-seed, revenue may be small, lumpy, service-assisted, discounted, or non-recurring. Report it with enough detail to understand its quality.

Useful definitions:

Monthly recurring revenue (MRR) = normalized recurring subscription revenue active for the month

Gross margin = (revenue − direct costs required to deliver that revenue) ÷ revenue

Average contract value = total contracted value for included contracts ÷ number of included contracts

Revenue concentration = revenue from the largest customer or group ÷ total revenue for the period

State what you exclude. One-time setup fees are not MRR. A non-cancelable annual contract, a month-to-month subscription, and an informal pilot should not be described as the same revenue quality.

Illustrative gross-margin calculation

A productized reporting service earns €8,000 in June. Direct analyst labor is €2,400, customer-specific data costs are €600, and payment fees are €200.

Gross profit = €8,000 − €2,400 − €600 − €200 = €4,800

Gross margin = €4,800 ÷ €8,000 = 60%

Whether that supports an investable software story depends on how delivery changes with automation, the customer value, pricing, and the rest of the model. Do not remove founder labor from direct cost merely to make the economics look better; show both current and clearly labeled target-state assumptions.

Acquisition evidence without pretending it is scalable

Founder-led sales is normal in many pre-seed companies. Track it honestly:

Segment results by source and batch. A founder’s former colleagues, a paid campaign, an association partnership, and cold outreach are different acquisition evidence.

Avoid presenting customer acquisition cost (CAC) as mature if founder selling time, free implementation, failed experiments, or discounts are omitted. A transparent early view can show:

Fully loaded experimental CAC = attributable sales and marketing cash cost + estimated direct acquisition labor cost ÷ new customers from that experiment

Label estimates and explain which costs are excluded. The point is to understand the motion, not manufacture comparability with a scaled company.

Market evidence

Top-down market-size slides are not operating metrics. Pair them with a bottom-up entry model:

Reachable annual market = number of target accounts in the initial segment × plausible annual contract value

Every input needs a source or test. “Plausible contract value” should connect to actual customer budgets, alternatives, pilots, or quotes—not a percentage of a broad industry total.

Show expansion as a sequence:

1. Initial user, trigger, and job.
2. Number of reachable accounts with that profile.
3. Adjacent users or workflows supported by evidence.
4. Product or distribution changes required to expand.

This makes a large vision more credible because the first wedge is concrete.

Execution and learning velocity

Shipping counts do not matter unless they reduce a meaningful risk. Maintain an experiment ledger:

Useful execution indicators include:

• Time from identified risk to customer-facing test.
• Percentage of planned experiments that reach a decision.
• Time from customer input to delivered result.
• Number of repeated manual exceptions.
• Decisions reversed because of evidence.

Do not reward the team for maximizing experiments. A thoughtful test that stops a weak direction can create more value than ten inconclusive releases.

Runway and use of funds

Fundraising metrics should connect capital to a de-risking plan.

Net burn = cash operating outflows − cash operating inflows for the period

Simple runway = unrestricted cash available for operations ÷ expected monthly net burn

When burn changes materially, use a monthly cash forecast rather than dividing by one historical month. Include hiring timing, taxes, debt, vendor commitments, fundraising cost, and contingency assumptions.

Build a milestone model:

Use financial modeling to make the assumptions explicit. Capital should fund a defined change in evidence, not simply “18 months of growth.”

A one-page pre-seed metrics dashboard

Keep the primary dashboard small:

Customer and problem

• Target customer and qualifying trigger.
• Relevant interviews and recent-problem observations.
• Paid pilots, signed agreements, or other exact commitments.

Product value

• Eligible new customers.
• Activated customers and activation definition.
• Time to first value.
• Top three activation failures.

Repeat and revenue

• Eligible cohort retention or repeat rate.
• Revenue split by recurring, one-time, and service-assisted components.
• Gross margin using stated direct costs.
• Customer concentration and refund/cancellation reasons.

Acquisition

• Qualified funnel by source.
• Median or distribution of sales-cycle time where the sample supports it.
• Experimental acquisition cost with included and excluded costs.

Cash and execution

• Cash, expected burn, and forecast runway.
• Current milestone, riskiest assumption, and last experiment decision.

Link every number to a source report, ledger, contract, or product query. Date the definitions. If a definition changes, preserve the old series or clearly annotate the break.

Build the investor packet around evidence

The metrics should connect to the fundraising materials:

• Deck: the concise claim and most decision-relevant evidence.
• Metric appendix: definitions, cohort tables, source notes, and limitations.
• Data room: governing documents, cap table, financial records, material contracts, intellectual property records, compliance documents, and the underlying reports appropriate to diligence.
• Model: hiring, revenue, burn, cash, and milestone assumptions.
• Narrative: why the team is suited to the problem and what was learned from contradictory evidence.

Use business pitch examples to improve the structure without copying another company’s claims. Review the distinction between pre-seed and seed, and use the pre-seed funding guide to map the process. Advice about raising a seed round may be useful later, but do not present seed-stage metrics as a universal pre-seed gate.

Securities-law caution

Raising money usually involves offering or selling securities. The U.S. Securities and Exchange Commission’s private-company guidance, updated in April 2026, explains that every offer and sale of securities must be registered with the SEC or qualify for an exemption, including private and friends-and-family transactions. The SEC also lists common startup securities, including stock, notes, LLC interests, and SAFEs.

Do not choose an instrument, contact investors, advertise an offering, accept funds, or make disclosure claims based only on a blog post or template. Work with qualified securities counsel and tax professionals for the jurisdictions and investors involved. Keep metrics accurate, consistently defined, supportable, and appropriately qualified.

Questions investors may ask about the numbers

Prepare clear answers to:

1. What exactly counts as an active, activated, retained, paid, or contracted customer?
2. Which customers are related parties, discounted, service-assisted, or not yet live?
3. What changed in the product, segment, price, or definition during the charted period?
4. Why is the chosen retention interval appropriate to the workflow?
5. Which growth came from founder relationships versus a repeatable source?
6. What do customers do instead, and why do they switch?
7. What result would disprove the current thesis?
8. What milestone will the round fund, and what happens if it takes twice as long?

A credible “we do not know yet; here is the test” is better than false precision.

What to prove before raising

You do not need to prove every part of the company. You should be able to show:

• A specific customer and consequential problem.
• The strongest available commitment or usage evidence, labeled accurately.
• A value event and a reason it matters.
• Early repeat behavior at the correct interval, or a plan to test it.
• Honest revenue quality and delivery economics where revenue exists.
• A plausible first path to more customers.
• Consistent metric definitions and source records.
• A milestone-based use of funds and downside-aware cash plan.
• The material risks that remain.

Pre-seed metrics are not a password that unlocks capital. They are a disciplined account of what the company knows, what it does not know, and why the next investment could change that boundary.