# The Startup Runway Read, One Seed Employer From Form D to a Raise-By Month

*Turn any US venture-backed employer's Form D filings and headcount into a low, mid, and high months-left range and a raise-by month you can check against the founder.*

- Canonical URL: https://www.refolk.ai/candidates/guides/startup-runway-read-form-d
- Pillar: Reading the market
- Format: Teardown
- Published: 2026-11-01
- Last reviewed: 2026-11-01
- Reading time: 20 min
- Keywords: how to estimate startup runway before joining, SEC Form D look up company funding, how to tell if a startup is running out of money, check startup funding before accepting offer, questions to ask startup about runway, is this startup financially stable

## Key takeaways

- For any one offering, the cash raised is the amount sold on its latest Form D amendment, because amendments restate cumulative totals and the SEC itself counts only the increments.
- More than half of roughly 40,000 venture rounds studied at Lehigh had no Form D, so a missing filing is the base case and tells you almost nothing on its own.
- Published benchmarks put all-in burn at about $8K to $12K per employee per month, and derived figures stay near that band from seed to Series B, which makes headcount the most portable public input for a burn estimate.
- Carta's seed-to-Series A median has run between 616 and 774 days, and only 20% of seed companies on Carta reached Series A within 24 months, so seed cash has to last longer than many founders plan for.
- Set the raise-by month 9 to 12 months before your estimated cash-out month, because a normal raise takes 3 to 6 months of calendar time.
- Refolk's index holds 1,613 US profiles titled Fractional CFO against 247 US CFO profiles with startup in the headline, which suggests a founder's runway figure may come from part-time finance help, so ask when it was last updated.

This guide is for job seekers about to spend real interview time on a US venture-backed startup who want their own estimate of its cash before they hear the founder's. I call it the runway read: a desk estimate, built from SEC Form D filings, a headcount, and a cost per head, that ends in a low, mid, and high months-left range and a dated raise-by month. One composite seed company is carried through every step, including the three mistakes that most often inflate the number.

## What does a runway read estimate, and what can it not prove?

A runway read estimates how many months of cash a startup has left. It uses three public inputs: cash raised according to Form D, a headcount, and a loaded cost per head. The result is a range, not a fact. It cannot see the bank balance, revenue, or unfiled rounds, so its job is to tell you whether the founder's runway answer is plausible. It does not replace that answer.

Four terms carry the whole method:

- **Runway**: the number of months until cash reaches zero at the current net burn.
- **Gross burn**: total monthly spending, before any revenue.
- **Net burn**: gross burn minus monthly revenue. Runway is computed on net burn.
- **Form D**: the notice a company files with the SEC after selling securities under Regulation D. It is due within 15 calendar days of the first sale.

What a runway read can show:

- Whether the founder's runway figure fits the money the company has publicly sold and the team it is paying.
- Roughly when the company has to start its next raise.
- Which questions would close the gap between your number and theirs.

What it cannot show:

- The current bank balance, or revenue you cannot see.
- Rounds that never produced a Form D, which is more than half of venture rounds in the largest study available.
- Venture debt arranged without a filing.
- Whether the next raise will succeed.

## Which public inputs feed the estimate?

Three inputs feed the estimate: the Form D amount sold per offering with its date of first sale, a headcount range, and an all-in cost per head per month. Each one can mislead in a known way, so each gets a check.

### Form D fields that matter

Form D asks for three amounts in Item 13: the total offering amount (the target), the amount sold (what was actually raised), and the total remaining to be sold. The offering amount may be reported as "indefinite". The **date of first sale** is the date the first investor is irrevocably committed to invest. A company may file before it has sold anything.

Amendments (form type D/A) carry cumulative totals. When the SEC aggregates Regulation D data, it counts only the increment each amendment adds. The form has security-type boxes (Equity, Debt, Option, Warrant or Other Right to Acquire) but no field for a round label like "Seed" or "Series A". I could not establish how companies classify SAFEs on the form, so do not infer the instrument from the box alone.

To find filings, EDGAR Full-Text Search lets you search by keyword, ticker, company name, CIK number, and reporter's last name. It covers filings submitted electronically since 2001. The **CIK** is the SEC's permanent ID number for a filer. It stays the same when the name changes, so it is the key you should track.

### Headcount

Count current staff from a profile index or the company's own pages. Record three more numbers: the 6-month change, open roles, and how many profiles are contractors, offshore, or stale. Headcount proves the size of the payroll. It lies when the profiles include people who left without updating, contractors billed at different rates, or offshore staff whose cost is nowhere near a US salary.

### Cost per head

Two methods are in circulation, and they do not agree.

- **All-in per head**: one source puts burn at $8K to $12K per employee per month, all-in, and says this stays stable across stages.
- **Salary-based**: loaded cost is typically 1.25x to 1.4x base salary (a rule of thumb is 1.3x). Payroll is 60% to 80% of total burn, so divide loaded payroll by that share to get all-in burn.

The stage benchmarks below show why per-head cost is the most portable input. The two sources disagree on total burn by about 2x, but divide burn by team size and both land near the same band.

| Stage | Fiscallion burn (team) | Centsight burn (team) | Burn per head |
|---|---|---|---|
| Seed | ~$75K (5-10) | $100K-$250K (8-20) | ~$10.0K / ~$12.5K |
| Series A | ~$250K (15-40) | $300K-$700K (25-60) | ~$9.1K / ~$11.8K |
| Series B | ~$600K (40-80+) | $600K-$1.5M (60-150) | ~$10.0K / ~$10.0K |

Burn per head is derived as median burn divided by team-size midpoint, Fiscallion first and Centsight second. Both sources are vendor content with no published method, so treat the table as a range, not a measurement.

#### From payroll to runway

1. **Runway** - Cash raised since last round, plus any debt, divided by net burn
2. **Net burn** - Gross burn minus revenue, if revenue can be estimated
3. **Gross burn** - Loaded payroll divided by payroll's 60-80% share, or $8K-$12K per head all-in
4. **Loaded payroll** - Base salaries times 1.25x to 1.4x
5. **Headcount** - Current staff, adjusted for stale, contractor, and offshore profiles

*Each layer adds a cost or a source of cash that the layer above it leaves out, and skipping any layer moves the months-left figure.*

## The runway read, step by step

The procedure has eight steps and takes about two and a half hours for one company, most of it on the filing ledger and the cross-check. Every step ends in a concrete artifact, so you can stop at any point and still know what you have.

#### The runway read

1. **Find the entity on EDGAR** - Search by legal name, then CIK, checking old names and holding entities. Done is one CIK and a list of every D and D/A filing (about 10 minutes).
2. **Build the filing ledger** - Record date filed, date of first sale, amount offered, amount sold, and security type. Done is one row per offering with amendments folded in (about 30 minutes).
3. **Back out the cash raised** - Use the amount sold on each offering's latest amendment, never the sum of filings. Done is cash in per offering, with debt flagged (about 15 minutes).
4. **Cross-check the funding story** - Match the ledger against press releases, investor pages, founder claims, and WARN notices if staff is 100 or more. Done is every claimed round matched or marked unfiled (about 30 minutes).
5. **Count headcount as a range** - Count current staff, the 6-month change, and open roles, then strip stale, contractor, and offshore distortions. Done is a low, mid, and high headcount (about 20 minutes).
6. **Estimate gross and net burn** - Multiply headcounts by $8K, $10K, and $12K per head per month, run the salary method as a stress check, and subtract revenue if known. Done is a low, mid, and high burn (about 15 minutes).
7. **Compute months left** - Subtract burn times months since date of first sale from cash raised since the last round, then divide by burn. Done is a low, mid, and high range, with and without debt.
8. **Set a raise-by month and test the founder** - Subtract 9 to 12 months from each cash-out month and convert to calendar months. Done is a dated window to put beside the founder's answer.

Some practitioners run this in reverse: they take the founder's answer first and test it against public data. That works if you already have the call booked. I prefer building the number first so the founder's figure cannot anchor you.

## Teardown: one seed company from EDGAR to a months-left range

Company S is a composite US seed-stage software company built to hit every fork in the method. Its dollar figures and headcounts are illustrative and were chosen to make the arithmetic visible. The rules and benchmarks applied to it are the sourced ones above. Months are relative: M0 is the seed round's date of first sale, and today is M14.

### Finding the entity

The first search used the product brand name and returned nothing. That was the first wrong turn: Company S's brand is not its legal name. The legal name from the careers page footer and the offer-letter template did return a filer, but under an earlier corporate name. The CIK pulled up four filings: three Form Ds and one D/A.

### The filing ledger

| Offering | First sale | Offered | Sold (latest) | Type |
|---|---|---|---|---|
| Pre-seed | M-17 | $1.0M | $1.0M | Other |
| Seed (D, then D/A at M5) | M0 | $6.0M | $5.5M | Equity |
| Mid-cycle | M9 | Indefinite | $1.5M | Debt |

The seed offering has two filings: the original Form D showing $3.2M sold, and a D/A at M5 showing $5.5M sold. They take up one row because they describe one offering.

### Wrong turn: summing every filing

Adding every amount sold gives $1.0M + $3.2M + $5.5M + $1.5M = $11.2M. Counting just the seed's two filings as separate money gives $8.7M for the seed alone. Both double-count. The D/A restates the cumulative total, so the seed raised $5.5M, not $8.7M.

> **Rule:** One offering, one number
>
> For each offering, cash raised is the amount sold on its latest amendment. Never add an original Form D to its D/A, because the SEC's own aggregation counts only the increments.

### Wrong turn: the target and the filing date

Two smaller misreads sat in the same rows. Reading the seed's $6.0M offering amount as cash raised overstates the seed by $0.5M. The debt row has no target at all ("indefinite"), so only the amount sold means anything there.

The larger error was dating the seed from the D/A at M5 instead of the date of first sale at M0. At the mid-case burn below, that one change cuts the estimated spend by five months of burn and inflates runway from 18.4 months to 23.4. Always measure elapsed time from the date of first sale.

### Cross-checking the story

The seed press release said "$6M seed". The filing says $5.5M sold. The press figure probably matches the target, but $0.5M may also have closed later without a new amendment. I logged it as an open question rather than a contradiction.

The M9 debt row matched no press release. It could be a convertible bridge from existing investors or venture debt. Either way it is cash in, so I kept it as a separate line and built a with-debt version of every number. WARN notices were irrelevant: federal WARN applies only to employers with 100 or more employees, and Company S is far smaller.

### Counting heads

The profile count gave 22 current people. Two had joined other companies without updating, which left 20. Four of those 20 were offshore contractors. The 6-month change was +3, and there were 2 open roles.

- **Low, 16**: US core staff, with offshore contractors at a rough quarter weight.
- **Mid, 17**: US core staff, with offshore contractors at about half weight.
- **High, 19**: closer to full weight, as if the open roles fill soon.

Fork: the team grew by 3 in the last six months, so using today's headcount for all 14 months overstates past spending. I kept the conservative version. In the mid case, the correction is worth about 1.4 months.

### Burn and months left

Revenue could not be estimated from public pages, so net burn equals gross burn here. That makes the range conservative. Cash in since the last round is the seed's $5.5M, with the $1.5M debt added in the second column. I set aside whatever was left of the pre-seed money, which also makes the estimate conservative.

| Case | Monthly burn | Months left | With debt |
|---|---|---|---|
| Low (16 x $8K) | $128K | 29.0 | 40.7 |
| Mid (17 x $10K) | $170K | 18.4 | 27.2 |
| High (19 x $12K) | $228K | 10.1 | 16.7 |

Sanity check against the stage table: $170K and 17 people sit inside Centsight's seed band and above Fiscallion's. Stress check with the salary method: one published example puts a $140K engineer at about $187K a year loaded (1.34x), or about $15.6K a month. Divided by an 80% payroll share, that is about $19.5K per head all-in, roughly double the benchmark. If Company S is mostly senior US engineers, the true number sits at or beyond the high case. That is a question to ask, not a number to publish.

### Raise-by month

Without debt, cash-out lands around M43 (low), M32 (mid), and M24 (high). Subtracting 9 to 12 months gives raise-by windows of M31 to M34, M20 to M23, and M12 to M15. Today is M14, so the high case says the raise should already have started. Converted to calendar months, the mid window falls six to nine months after the read date.

The founder later said "about 30 months of runway". That matches only the low case, or the mid case with the debt counted. The answer is plausible, but only if burn is near $128K a month or the M9 debt is in the number. The founder call then had three specific questions instead of one vague one. Company S cleared the read, so the next hour went to the application. [Refolk](/candidates) tailored the resume to that posting and scored the fit, which is where the time saved on the desk work belongs.

> A Form D tells you what a company sold, never what it has left.

## How does a runway read go wrong?

Most bad reads come from one of eight mistakes, and five of them make the company look richer than it is. Each one below lists what it does to the estimate and how to catch it.

1. **Summing amendments.** A D/A restates the cumulative amount sold, so adding it to the original double-counts the money. Check: the SEC itself uses only the increments.
2. **Reading the offering amount as cash raised.** The offering amount is a target, and an "indefinite" offering has no target at all. Check: use Item 13(b), the amount sold.
3. **Treating the filing date as the close date.** A company can file before any sale, and the first sale can be up to 15 days before the filing. Check: measure elapsed months from the date of first sale.
4. **Taking "no Form D" to mean "no funding".** A Lehigh study of 40,000 venture rounds found more than half had no Form D, and a company can raise under Section 4(a)(2) without one. Check: compare with press releases and investor portfolio pages.
5. **Using the wrong headcount.** Profile counts include contractors, offshore staff, and stale profiles, and applying a 1.25x to 1.4x multiplier to US salaries overstates offshore cost. Check: look at locations and recent joiners.
6. **Leaving out revenue and debt.** Gross burn is not net burn. Venture debt can extend runway without any Form D, and debt securities may appear under the form's Debt box. Check: ask about both directly.
7. **Trusting thin benchmarks.** Many burn tables come from vendor blogs with no stated method. A fintech-only study shows median burn rising 3.2x from seed to Series A, from $128K to $412K a month, so sector matters too. Check: present benchmarks as a range and name the source.
8. **Relying on WARN for small startups.** Federal WARN covers employers with 100 or more employees and requires 60 days' notice. A team under 100 triggers nothing, so a missing notice tells you nothing. Some states set lower thresholds, such as New Hampshire at 75 employees.

> **Watch out:** Absence is the base case
>
> No Form D is not no funding. Early-stage high-tech and biotech firms are the most likely to skip the filing, so a gap means something only when a company that filed every earlier round suddenly stops.

Enforcement does not fix the gap. The SEC's first-ever enforcement action aimed at failure to file Form D carried penalties of $60,000 to $195,000, which tells you how rarely the rule has been pressed. Filing costs nothing, because the SEC charges no fee for a Form D or an amendment, so skipping it is a choice the company made, not a cost it avoided. Late filing rates are not publicly established. If a date of first sale sits well before the filing date, note it and move on.

One more limit: nobody has publicly measured which signals reliably come before a shutdown. Hiring freezes, no backfills, bridge notes, and filing gaps are all things to ask about. None of them is a validated predictor, and a guide that tells you otherwise is overclaiming.

## How long does seed cash have to last?

Longer than the typical plan. Carta's seed-to-Series A median has run from 616 to 774 days, about 20 to 26 months, while a raise still takes 3 to 6 months of calendar time. That means many seed companies hit their raise-by month before Series A arrives.

| Measure | Value | Source |
|---|---|---|
| Seed-to-A median, earliest of three readings | 774 days | Carta, via pitchwise.se |
| Seed-to-A median, middle reading | 616 days | Carta |
| Seed-to-A median, latest reading (9,843 US rounds) | 1.9 years | Carta |
| Active raise duration | 3-6 months | CFO Advisors |
| Start raising at | 9-12 months left | CFO Advisors |

**20% - Seed companies on Carta that reached Series A within 24 months**

Down from 40% at the last fundraising peak, per Carta's Peter Walker; re-check Carta's published round-timing data before relying on the current value.

Apply this to your own read. Take the months elapsed since the seed's date of first sale and compare them with the current Carta median. In the teardown, Company S is at M14 against a median of roughly 20 to 23 months, and its mid-case raise-by window is M20 to M23. Those line up, which means the company should be preparing now, not after you join.

Sources disagree on the lead time. Most say start raising with 9 to 12 months left, while Angel Investors Network says 6 to 9. VC Boom ties it to stage: a Series A company burning $300K or more a month should start at 14 months. Another source says a Series A raise alone takes 4 to 9 months. I use 9 to 12 for seed and early Series A, and 14 for anything burning over $300K a month. I could not find a current Series A to Series B median, so for later-stage employers, lean on the lead time and leave out the gap benchmark.

## How do you test the founder's runway answer?

Put the founder's number next to your range and ask which case it matches. If it matches your mid case, that is a good sign. If it matches only your low case, or needs the debt to work, ask about those assumptions. If it exceeds all three, there is either cash you cannot see or a model that is out of date.

#### Reading the founder's answer

Horizontal axis runs from Founder's number inside your range to Founder's number above your range. Vertical axis runs from Thin filing trail to Complete filing trail.

| Quadrant | What it means |
| --- | --- |
| Trail complete, number fits | Proceed and ask one question to confirm net burn |
| Trail complete, number exceeds | Press on revenue, unfiled debt, and when the model was last updated |
| Trail thin, number fits | Ask for round sizes directly; your range rests on headcount alone |
| Trail thin, number exceeds | Treat the answer as unverified until the founder names cash in and burn |

*Where the founder's number lands against your range, and how complete the filing trail is, decides whether you probe, proceed, or press.*

These are the questions to ask a startup about runway. Each one targets a specific assumption in your estimate.

**Founder runway questions**

```
When you say <N> months of runway, is that on today's net burn or on the hiring plan?
Roughly what is monthly net burn today, and how much revenue offsets it?
Does that figure include the <debt / note> that closed around <month of first sale>?
When do you plan to start the next raise, and what milestone does it depend on?
Who maintains the runway model, and when was it last updated?
Is there any venture debt or credit line that has not been announced?
```

*Fill the angle brackets from your ledger and estimate before the call; ask them in this order.*

The last two questions matter more than they look. Refolk's index points to a large pool of part-time finance help behind startup numbers.

**1,613 - US profiles titled Fractional CFO in Refolk's index**

Against 247 US Chief Financial Officer profiles with "startup" in the headline, a narrow lower-bound count.

| Segment in Refolk's index | Count | Ratio (derived) |
|---|---|---|
| Fractional CFO, US | 1,613 | 4.2x the UK count |
| Fractional CFO, UK | 381 | 1.0x |
| CFO with "startup" in headline, US | 247 | Fractional is 6.5x (not like-for-like) |

In a 25-profile sample of the US fractional set, 13 held exactly the title "Fractional CFO", with matched employers including Slip Robotics, GreenSight, and Aurora Therapeutics. My inference, not a measurement: many startups get their runway figure from a fractional model that may not be updated every month. A founder quoting a stale model is not lying, but the number may lag the current team size. That is why the question about when the model was last updated is in the template.

> **Tip:** Ask the question you can check
>
> "What's your runway?" invites a rounded answer. "Does that include the note from <month>?" can be checked against your ledger, and it tells the founder you did the reading.

## Keeping the read current

A runway read goes stale the moment the company hires, raises, or cuts. Re-run steps 2, 5, and 7 whenever a new Form D or D/A appears, when headcount moves by more than two or three people, or when an offer arrives. Each re-run takes well under an hour because the ledger already exists.

Keep the worksheet below for each employer you are seriously considering. Dating it lets you compare later reads with earlier ones.

**Runway read worksheet**

```
RUNWAY READ: <legal name>   CIK: <number>   Read date: <month>
LEDGER (one row per offering, amendments folded in)
Offering | First sale | Latest filing | Offered | Sold (latest) | Type | Note
CASH IN SINCE LAST ROUND: $____ equity   + $____ debt (flagged)
HEADCOUNT: raw __  stale __  offshore/contractor __  open roles __  6-mo change __
HEADCOUNT RANGE: low __  mid __  high __
COST PER HEAD / MONTH: low $8K  mid $10K  high $12K   stress: salary x 1.3 / 0.8
GROSS BURN: low ____  mid ____  high ____   REVENUE (if known): ____
NET BURN: low ____  mid ____  high ____
MONTHS SINCE DATE OF FIRST SALE: __
MONTHS LEFT = (cash in - net burn x months elapsed) / net burn
  low __  mid __  high __   with debt: low __  mid __  high __
CASH-OUT MONTH: low ____  mid ____  high ____
RAISE-BY WINDOW (cash-out minus 9 to 12): low ____  mid ____  high ____
FOUNDER'S ANSWER: __ months   Matches: low / mid / high / none
OPEN QUESTIONS:
```

*One copy per employer; keep every dated version so you can see the trend.*

#### Before you trust the read

- [ ] The CIK is confirmed, and old names or holding entities have been searched.
- [ ] Each offering appears once, valued at the amount sold on its latest amendment.
- [ ] Elapsed months run from the date of first sale, not the filing date.
- [ ] Every round the company claims is matched to a filing or marked unfiled.
- [ ] Debt-type filings are flagged and carried in a separate with-debt column.
- [ ] Headcount is a low, mid, and high range with stale and offshore profiles adjusted.
- [ ] Burn is shown as a range with its benchmark source named.
- [ ] The raise-by window is written as calendar months, not relative ones.
- [ ] The founder's answer is placed against a specific case, with the questions that would close the gap.

When the read clears and you decide to apply, the work shifts from the company's finances to your own fit. Refolk writes the resume from your history, tailors it to that posting, and scores how well you match, so the time this read saved goes into the applications that deserve it.

## Frequently asked questions

### Can I look up how much a startup has raised on SEC EDGAR?

Yes, if it filed a Form D. Search EDGAR by company name or CIK number, open each Form D, and read the amount sold, not the total offering amount, which is only the target. Use the amount sold from the latest amendment for each offering, because amendments restate the running total. EDGAR full-text search covers electronic filings back to 2001, but many venture rounds never appear there at all.

### What does it mean if a startup has no Form D filing?

On its own it means very little. A Lehigh study of about 40,000 venture rounds found more than half had no Form D, and early-stage tech and biotech companies skip it most. Companies can also raise under Section 4(a)(2), which does not require the form. A gap only means something when a company that used to file every round stops filing. Check press releases and investor portfolio pages instead.

### How many months of runway should a startup have when I join?

Enough that the next raise has not already become urgent. Most sources say a company should start raising with 9 to 12 months of cash left, because a raise takes 3 to 6 months. If your mid-case estimate puts the company inside that window today, the next year of your job depends on a raise going well. That is not automatically a no, but ask about it directly before you sign.

### How do I estimate a startup's burn rate from headcount?

Multiply headcount by an all-in monthly cost per head. Published benchmarks put that at roughly $8K to $12K per employee per month across stages, so use $8K, $10K, and $12K for low, mid, and high. As a stress check, take a typical base salary, multiply by 1.25 to 1.4 for loaded cost, and divide by payroll's 60% to 80% share of burn. For a senior US engineering team, the stress check can run about twice the benchmark.

### What does indefinite mean as the total offering amount on a Form D?

It means the company did not state a target for that offering. The form allows the total offering amount to be reported as indefinite, so you cannot compare sold against offered for that row. Ignore the offering field and use the amount sold, which is still reported. Indefinite offerings often sit alongside debt or open-ended instruments, so check the security type box and treat the cash as its own ledger row.

### Is a bridge note or debt filing a sign a startup is in trouble?

Not on its own, and nobody has publicly measured how often it comes before a shutdown. A debt-type filing between priced rounds adds cash and extends runway, so include it in a with-debt version of your estimate. What it does tell you is that the company needed or wanted money before its next priced round. Ask what the note is for and whether the runway number you were given includes it.

---

*From the Refolk guide library. I revise these guides rather than replacing them, so the current version is always at https://www.refolk.ai/candidates/guides/startup-runway-read-form-d*