Refolk
StandardOffers and negotiation

The Offer Total Comp Standard, Each Pay Line Graded Count, Haircut, or Exclude

Turn any written offer into a year-one and a steady-state total comp figure, each line tied to a letter clause, that a second grader reproduces within rounding.

24 min readLast reviewed November 15, 2026Read as Markdown

Key takeaways

  • On a back-loaded 5/15/40/40 vesting schedule, year-one equity is 0.20x the grant-divided-by-four figure, so a $400,000 grant pays $20,000 in year one, not $100,000.
  • Target bonus counts at full target only in the steady-state figure; in year one it is cut by the letter's proration fraction and drops to $0 when the start date passes a hire cutoff such as Dropbox's November 1.
  • Private-company options carry $0 paper value at grant on the 409A basis, because the strike price must be at or above the common stock's fair market value, so the preferred-price spread belongs on a memo line.
  • Sign-on repayment is not always prorated: Dropbox's $100,000 is owed in full on any exit in the first year, and AMD's $250,000 is owed in full until 13 full months before stepping down 8.33% per month.
  • Any pay line without a written amount in a clause, including refreshers and retirement match, is graded Exclude and contributes $0 to both figures.
  • A total comp figure is done only when a second person, grading the same letter from scratch, reproduces both the year-one and steady-state numbers within rounding and every dollar of gap to the recruiter's headline is explained.

This standard is for anyone holding a written offer, or two, who needs one year-one and one steady-state total compensation figure they can trust. It grades every pay line as Count, Haircut, or Exclude, each grade tied to a clause in the letter. That way your own math, the recruiter's headline, and a friend checking the work all land on the same dollars. You can use it as written, or as the shared definition for anyone helping you compare offers.

What a reproducible total comp figure is

A reproducible total comp figure is a pair of numbers, year one and steady state, where every dollar traces to a written clause and a second person grading the same letter lands within rounding. If a second grader cannot reproduce it, it is an opinion about the offer, not a measurement of it.

A standard is needed because the public formulas disagree. Some pages add the full four-year grant to year one, some count the maximum bonus, and some fold in health premiums. Levels.fyi, the most widely cited public dataset, defines total compensation as base salary plus annual bonus plus the annualized value of stock grants. Sign-on bonuses count only in the year they are paid, benefits are left out, and stock is divided evenly across the years it takes to vest. Its own materials then conflict. Its founder wrote that the displayed bonus is the annual recurring bonus summed with one quarter of the sign-on. Its data guide says the total can sit slightly above the sum because of relocation or profit sharing. If the most-cited source uses two conventions, two people valuing the same letter will produce two numbers.

This standard fixes the conventions:

  • Year one is what the letter pays for the first twelve months of employment: base, the first bonus cycle at its prorated amount, the sign-on, and the first year of vesting.
  • Steady state is a typical later year: base, target bonus, and the grant divided by its vest years. No sign-on.
  • Count means the full written amount enters the figure.
  • Haircut means the written amount enters reduced by the letter's own formula: a proration fraction, a vesting percentage, or an annualization.
  • Exclude means $0 in the figure. You can still record the line as a memo so nothing is lost, but it does not change the total.

The whole method depends on tying each line to a clause, and the employer has people whose job is to write those clauses. Compensation is a staffed function inside most employers of size.

3,017
People titled Compensation Analyst in the United States, in Refolk's index
The top-10 employers for this title each hold one person, so the function is spread across employers rather than concentrated in a few.

In Refolk's index of professional profiles, the top-10 employer list for US Compensation Analysts shows one person per company, including Okta, Syngenta and Robert Half. No single employer dominates. The narrower specialty is rare. Only 8 people in the US are titled Equity Compensation, and they work mostly at plan administrators and banks: Fidelity Investments (2), UBS, Goldman Sachs and ADP. My reading: share-level details usually live in plan documents and administrator records, not in the recruiter's summary email. Ask for the documents.

Grading rules for each pay line

Each pay line gets one grade for year one and one for steady state, and each grade rests on a named clause. The two tables below are the standard. The list after them says what each criterion proves and how it can mislead.

Cash lines

LineYear oneSteady stateClause it rests on
Base salaryCountCountSalary clause, annualized
Target bonusHaircutCount at targetProration, cutoff, payout date
Sign-onCount in year paidExcludeSign-on agreement
Other cashCount if writtenExcludeClause stating the amount

Equity and benefit lines

LineYear oneSteady stateClause it rests on
RSUsHaircut to year-1 %Haircut to grant ÷ vest yearsGrant notice
Options (private)Exclude, $0Exclude, memo spreadStrike and latest 409A
RefreshersExclude unless writtenExclude unless writtenClause stating an amount
Retirement matchExcludeExclude, memoBenefits stay out

What each criterion proves, and what it looks like when it misleads:

  • Base salary, Count. Proves guaranteed cash at the stated rate. It misleads when the letter states a monthly or hourly figure and someone annualizes with the wrong multiplier. Cite the clause and show the arithmetic.
  • Target bonus, Haircut in year one. Proves what the first cycle can pay given your start date. It misleads when target is entered at 100% for a start date past the plan's cutoff, or when the payout falls after the date you would be gone.
  • Target bonus, Count at target in steady state. Proves the plan's stated expectation, not a promise. It misleads when someone uses the maximum instead. Never count the maximum bonus.
  • Sign-on, Count in year paid. Proves one-time cash. It misleads when it is spread across four years to raise the steady-state number, or when the repayment terms make it effectively a loan you are likely to pay back.
  • RSUs, Haircut. Proves value at the grant's dollar figure on the letter's own schedule. It misleads when grant ÷ 4 is used for year one on a back-loaded schedule, or when the dollar figure is treated as if the share count were already fixed.
  • Options, Exclude. Proves you are not inventing value at grant. The spread to the latest preferred price is useful information, but it goes on a memo line, not into the total.
  • Refreshers and match, Exclude. Proves the total holds only written commitments. A recruiter's "typical refresher" is a hope, not a term.

How to grade the target bonus and sign-on in year one

Year-one bonus is target multiplied by the letter's proration fraction, or $0 if the start date falls after a hire cutoff. Where the letter requires it, you collect it only if you are still employed on the payout date. Sign-on counts in full in the year paid, with its repayment curve recorded beside it.

Target bonus: the clauses that change year one

Four filed offer letters with bonus clauses show the range of terms:

  • Dropbox: no bonus is earned unless you are still employed on the payment date. The bonus is prorated if you are employed less than a year, and a start on or after November 1 means no bonus that year.
  • Electronic Arts: you must be employed by January 15 of the fiscal year. The bonus is prorated by actual months, and you must be actively employed when bonuses are paid.
  • A filed letter (SEC CIK 896842): payment by April 15 of the following fiscal year, provided you are actively employed on the payment date.
  • Starbucks: prorated from your eligibility date through the end of the fiscal year.

Across those 4 letters, proration appears in 4, the employed-on-payment-date rule in 3, a hire cutoff in 2, and a stated payout month in 1. The sample is small and not random, and how often each term appears across all offer letters is not publicly known. Use it as a list of terms to look for, not as a frequency.

Year-one bonus from the letter's own clauses

  1. Hire cutoff
    Compare the start date to any cutoff; past it, year one is $0
  2. Proration basis
    Apply the letter's fraction, by months or from eligibility date to fiscal year end
  3. Payout date
    Record when the first cycle pays, such as April 15 of the next fiscal year
  4. Employed-on-payment rule
    If required, the bonus pays only if you are still there that day
  5. Record
    Year one equals target times fraction; steady state equals target
Year-one bonus is target times the letter's proration fraction, and only if you clear the hire cutoff and are employed on the payout date.

The payment-date rule is the term people miss. With an April 15 payout, someone who leaves in March collects $0, even after earning a full prorated share. A bonus earned in year one can be paid in year two. Record the payout date on the worksheet. If your planned tenure is uncertain, that date is the one that matters.

An illustration from the filed letters: Dropbox paired a $500,000 base with a 100% target bonus. Steady state counts $500,000 of bonus. Year one depends entirely on the start date: a start on or after November 1 makes it $0, and an earlier start gets the prorated fraction.

Sign-on: count it, then map what you would owe

Sign-on terms are often in a separate agreement, apart from the offer letter, so confirm you have it. Filed letters show four repayment structures:

  • Full repayment within 12 months, any reason: Dropbox. Exelixis uses the same window but applies it only to voluntary exits.
  • Prorated over 730 days in year two: Electronic Arts.
  • 100% then a monthly step-down: AMD requires 100% repayment before 13 full months, then 8.33% less per full month. Partial months do not count.
  • Daily proration per installment, voluntary exit only: Starbucks. Year-one repayment is ((365 minus days employed) ÷ 365) × $2,500,000.

The table below applies those clauses to two exit points.

Letter and amountOwed at month 6Owed at month 18Trigger
Dropbox, $100,000100%0%Any reason
AMD, $250,000100%About 50% ($125k)Termination
EA, $200,000Not verifiedAbout 25% (183/730)Employment ends
Exelixis, $10,000100%0%Voluntary exit

The columns are computed from the filed clauses. The EA year-one repayment clause was not captured in the source, so its month-6 figure is not verified. Read your own letter for that window. AMD counts only full months, so recompute at your exact exit month instead of trusting the round number.

Two design points follow from these clauses. First, the trigger wording decides who carries the risk. "Any reason" at Dropbox puts layoff risk on you, while the voluntary-only triggers at Exelixis and Starbucks do not. Second, back-loaded vesting and the sign-on are often one design. On a 5/15/40/40 schedule, year-one equity is 0.20x straight-line, and a sign-on is a common way to fill that gap. Grade the two lines together. Graded separately, they can push the year-one figure wrong in either direction.

Sign-on repayment risk by trigger and curve

All-or-nothing repaymentProrated repayment
All-or-nothing, voluntary (Exelixis)
Count it; the full amount is at risk only if you choose to leave inside 12 months
All-or-nothing, any reason (Dropbox)
Count it, flag the layoff exposure, and ask for voluntary-only wording
Prorated, voluntary (Starbucks)
Count it; lowest risk to you, and the amount owed shrinks daily
Prorated, any end of employment (EA)
Count it; record the amount owed at months 6, 12 and 18
Voluntary exit onlyAny reason or any termination
The trigger decides who carries layoff risk and the curve decides how fast that risk shrinks, so read both before trusting the sign-on line.

How to value RSUs and options in an offer letter

Value RSUs at the letter's dollar figure on the letter's own vesting schedule. Year one is the year-1 percentage times the grant, and steady state is the grant divided by its vest years. Value private-company options at $0 at grant, and keep the spread to the preferred price as a memo line.

Share count: the dollar figure is not the share count

Most letters state a dollar value and convert it to shares using a pricing window: the set of trading days whose average price is divided into the dollar figure. Filed letters use at least five windows:

  • AMD: the average closing price over the 30 trading days ending on the grant date, with the grant made on the 15th of the month after the start date. The filed AMD letter valued the grant at $1,125,000.
  • CIK 1366561: a 30-day period ending on the last day of the month before your start date, rounded down.
  • DexCom: a 30-trading-day period ending five trading days before the grant date, rounded up.
  • RingCentral: the trading days in a named month. One letter sized 63,343 RSUs on the April 2025 average close.
  • A letter quoted on Blind: a 30-trading-day period ending on the Tuesday before the grant date.

So two letters with the same dollar figure can deliver different share counts. A dollar-denominated grant leaves the price risk between offer and grant with you. NASPP illustrates a stock falling from $34 to $23 in one month: after the stock recovered, the later-granted employee's grant was worth $14,756 against about $10,000, roughly 48% more. The reverse also happens. In one anecdotal case, a ZS candidate saw the stock rise from $250 to $350 before joining, which they described as effectively reducing the stock they would get. No public benchmark shows how far prices typically move between offer and grant, so do not assume a size. Record the window, the grant date, and the rounding direction, and compute the share count once the price is known.

For the year-one and steady-state dollars, this standard uses the letter's dollar figure. The share count is a memo line that explains why the value at vest will differ.

Vesting shape: year one is the grant notice's percentage

How the grant vests changes year one far more than most people expect. The table applies four published schedules to an illustrative $400,000 grant. The dollar columns are derived.

ScheduleYear-one $Versus grant ÷ 4Year-four $
Even 25/25/25/25100,0001.00x100,000
Amazon 5/15/40/4020,0000.20x160,000
ByteDance 15/25/25/3560,0000.60x140,000
Google 33/33/22/12132,0001.32x48,000

These schedules come from secondary sources, not filed letters. Levels.fyi notes that ByteDance sometimes issues 15/25/25/35. A benefits guide documents Amazon's 5/15/40/40, and Forbes reported Google's move to 33/33/22/12. The filed exhibits I found used even schedules: Starbucks ratable at one quarter a year, and Exelixis with a one-quarter cliff then 1/48 monthly. Your grant notice overrides all of these.

Steady state uses grant ÷ vest years even on an uneven schedule. That matches the Levels.fyi convention and keeps offers on one basis. Put the year-by-year shape on the worksheet as a memo, because for the same $400,000 grant, year four pays $160,000 on the Amazon shape and $48,000 on the Google shape.

Options: $0 at grant on the 409A basis

A 409A valuation is an independent appraisal of a private company's common stock that sets the floor for option strike prices. Under 26 U.S.C. 409A, options must be granted with an exercise price at or above the common stock's fair market value, so their intrinsic value at grant is $0. Grade options Exclude in both figures.

A recruiter's options figure often implies the preferred price. Preferred stock is the class investors buy, and it is typically priced above common. One published example: a 409A shows common at 25 cents while Series A preferred trades at $1.50. Published ratios of common to preferred vary widely by source: 10 to 30% of the preferred valuation, 20 to 40% at seed stage, and 60 to 80% at Seed or Series A. What basis a given recruiter assumes is not publicly known, so ask. Record (preferred price minus strike) × option count as a memo line labeled "preferred-basis spread, not cash".

The grading procedure, step by step

The procedure takes one offer letter to two graded figures and a second-grader sign-off in about a working day. Most of that is waiting for documents. The time estimates are working estimates from practice, not documented benchmarks.

Grade one offer letter

  1. Gather every governing document
    Collect the offer letter, grant notice, bonus plan, and sign-on agreement, which is often a separate document. Done when every pay line maps to a clause; allow about a day if the recruiter has to send items.
  2. Count base salary
    Enter the annualized base in both columns and cite the salary clause. Takes minutes.
  3. Haircut the target bonus for year one
    Year one is target times the proration fraction, or $0 past a hire cutoff; steady state is target, never maximum. Record the payout date and any employed-on-payment-date rule.
  4. Count the sign-on in the year paid
    Year one only, excluded from steady state. Record the trigger wording and the amount owed at months 6, 12 and 18.
  5. Grade RSUs from the grant notice
    Record the pricing window, grant date, rounding, and share count. Year one is the year-1 percentage times the grant; steady state is the grant divided by vest years.
  6. Carry options at $0 on the 409A basis
    Use $0 intrinsic value at grant because the strike is set at or above fair market value. Put the preferred-basis spread on a memo line.
  7. Exclude every line with no written amount
    Refreshers, retirement match, and other cash get $0 unless a clause states an amount; benefits stay out of both figures.
  8. Have a second person regrade from scratch
    Hand over the letter, the documents, and this standard, not your worksheet. Done when both figures match within rounding and the gap to the recruiter's headline is explained line by line.

Step 1 is where most of the time goes. Ask for everything missing in one email so the answers arrive in writing and in one place.

Document request to the recruiter
Hi [recruiter name],

Thanks again for the offer. Before I respond, could you send the documents that govern each pay line, so I am reading the same terms you are?

1. Bonus: the plan or clause showing the proration basis, any hire cutoff date, the payout date, and whether I must be employed on the payout date.
2. Equity: the grant notice or plan summary showing the grant date, the pricing window used to convert the dollar value into shares, the rounding direction, and the vesting percentage for each year.
3. Sign-on: the sign-on agreement if it is separate, including the repayment trigger and how the amount owed changes month by month.
4. Options (if any): the strike price, and the date and per-share value of the latest 409A valuation.
5. Anything else in the total you quoted, such as refreshers, retirement match, or relocation, with the clause that states the amount.

Email is perfect so I have everything in one place. Thank you,
[your name]

Delete any numbered item your offer does not include; keep the request in writing.

How this goes wrong: failure modes and false positives

Most bad total comp figures fail in one of seven ways. The most dangerous ones produce a number that looks right because it matches the recruiter's. Each failure below has a giveaway and a check.

  1. Grant ÷ 4 on a back-loaded schedule. On the 5/15/40/40 schedule this overstates year-one equity 5x: $100,000 instead of $20,000 on a $400,000 grant. Giveaway: your year-one figure matches the recruiter's annualized number exactly. Check: use the year-1 percentage from the grant notice.
  2. Bonus counted at full target in year one. Wrong when the start date falls after a cutoff (Dropbox: November 1; EA: January 15) or the payout lands after a planned exit. Giveaway: the year-one bonus equals the steady-state bonus. Check: compare the start date to the cutoff, and the payment date to your expected tenure.
  3. Treating the RSU dollar figure as a share count. Windows range from the month before start to five trading days before grant, so letters with identical dollar values can deliver different share counts. Giveaway: a share count on the worksheet with no window or price beside it. Check: compute shares from the letter's own window and note the rounding direction.
  4. Options valued at the preferred price. This creates value that does not exist at grant, because the strike is set at the 409A fair market value. Giveaway: any dollar value for options in either total. Check: ask for the strike and the latest 409A in writing.
  5. Assuming sign-on clawback is always prorated. Dropbox's first year and AMD's first 13 months are all-or-nothing. Giveaway: a worksheet that says "prorated" without a repayment curve. Check: map the amount owed at months 6, 12 and 18.
  6. Mixing two Levels.fyi conventions. One counts sign-on in the year paid, the other annualizes it at one quarter. Giveaway: the steady-state figure includes a slice of the sign-on. Check: state the convention at the top of the worksheet.
  7. Counting refreshers or retirement match with no written amount. Giveaway: a line whose clause column is blank or says "verbal". Check: exclude it unless a clause gives an amount.

False positives: agreement that proves less than it seems

  • Your number matches the recruiter's. That proves you used the same convention, not the right one. Failure 1 is the usual cause.
  • Two graders agree. That proves the number is reproducible, not that it is correct. Two people who both divide by four will agree perfectly. This is why the second grader works from the letter and this standard, never from your worksheet.
  • A benchmark agrees. A dataset median uses that dataset's convention. Matching it tells you nothing unless you have checked the convention line by line.
A number that matches the recruiter's proves you used their convention, not that either of you read the letter.

Limits of the evidence behind this standard

Some of the evidence is thin, and the standard says so:

  • The bonus-term counts come from four filed letters, not a random sample.
  • The uneven vesting schedules come from secondary sources, not filed letters.
  • The EA month-6 repayment figure was not verified.
  • No public benchmark exists for price moves between offer and grant.
  • The valuation basis recruiters assume for options is not publicly known.

Where the standard rests on one of these, the fix is the same: read your own clause, and ask in writing when the clause is silent.

Second-grader check before the number goes anywhere

Before either figure goes into a comparison, a counter, or a conversation with a recruiter, a second person grades the same letter from scratch and must land within rounding on both figures. Then you explain every dollar of gap to the recruiter's headline, line by line.

Give the second grader the letter, the supporting documents, and this standard, not your worksheet. Allow about 30 minutes. When the two of you disagree, the cause is almost always a single line. Settle it by reading that clause together, not by averaging the two numbers. Averaging hides exactly the error the check exists to catch.

Total comp grading worksheet
OFFER: [company], letter dated [date], start date [date]
CONVENTION: sign-on counted in year paid; steady state = base + target bonus + grant ÷ vest years

LINE | Y1 GRADE | Y1 $ | SS GRADE | SS $ | CLAUSE
Base salary | Count | | Count | |
Target bonus | Haircut (target x fraction) | | Count at target | |
Sign-on | Count | | Exclude | 0 |
Other cash | Count if written | | Exclude | 0 |
RSUs | Haircut (year-1 %) | | Haircut (grant / vest years) | |
Options | Exclude | 0 | Exclude | 0 |
Refreshers | Exclude unless written | | Exclude unless written | |
Retirement match | Exclude | 0 | Exclude | 0 |

MEMO
Bonus payout date / employed-on-payment rule:
RSU pricing window / grant date / rounding / share count:
Vesting by year (%):
Sign-on owed at month 6 / 12 / 18, and trigger:
Options: strike / latest 409A value / preferred-basis spread (not cash):

TOTALS: Year one $______  Steady state $______
RECRUITER HEADLINE: $______  GAP EXPLAINED BY LINE:
SECOND GRADER: [name]  Y1 $______  SS $______  Within rounding: Y / N

One worksheet per offer; fill the clause column before any dollar column.

Before the figures leave your desk

  • Every line on the worksheet cites a clause or a named document in the clause column.
  • The year-one bonus uses the letter's proration fraction, and the start date has been checked against any hire cutoff.
  • The bonus payout date and any employed-on-payment-date rule are recorded as memo lines.
  • No line uses the maximum bonus.
  • Year-one RSU value uses the grant notice's year-1 vesting percentage, not grant ÷ vest years, unless the schedule is even.
  • The RSU pricing window, grant date, and rounding direction are recorded, and the share count is computed or marked unknown.
  • Options are carried at $0 in both totals, with the strike and latest 409A requested in writing.
  • The sign-on appears in year one only, with the amount owed at months 6, 12 and 18 and the trigger wording recorded.
  • Refreshers, retirement match, health benefits, and every line without a written amount contribute $0.
  • The sign-on convention is stated at the top of the worksheet.
  • A second grader working from the letter alone reproduced both figures within rounding.
  • Every dollar of difference from the recruiter's headline is explained by a named line.

Keeping the standard current, including UK offers

The standard stays current if you re-read the governing documents for each new offer instead of relying on what you remember about a company, because vesting schedules, bonus terms, and dataset definitions all change. UK readers should also expect different vocabulary for the same function.

Check these again for every offer:

  • Vesting schedules. Google's reported move to 33/33/22/12 shows a large employer can change its schedule. Take the percentages from this offer's grant notice, never from a past offer or a forum post.
  • Bonus and sign-on terms. Cutoffs, payout dates, and repayment curves differ between letters from the same employer. Grade the letter in hand.
  • Pricing windows. Read the window and the rounding direction in each letter. They are not industry constants.
  • Dataset definitions. Before comparing to any benchmark, re-read that dataset's definition page and note its sign-on and benefits treatment.

UK offers come with a vocabulary gap. In Refolk's index, the people who define pay lines go by different titles on each side of the Atlantic.

Title, countryPeople in Refolk's index
Compensation Analyst, US3,017
Compensation Analyst, UK45
Reward Analyst, UK183
Equity Compensation, US8
4.1x
UK Reward Analysts per UK Compensation Analyst, in Refolk's index
UK employers mostly call this function "reward", so UK readers searching "total comp" will mostly find US conventions.

The US has 67x the UK count of Compensation Analysts. In the UK, the people writing these clauses work under "Reward" titles at employers such as HSBC, Railpen, Ageas UK, Deloitte and Sainsbury's. The grading criteria still apply, because they rest on clauses rather than vocabulary. The 409A rule, though, is US tax law. For options granted outside the US, ask in writing what sets the strike price before you grade the line.

What to do next: graded figures are an input, not a decision. Pushing back on a weak line is a separate job, the written counter, and it goes better when you can cite the clause and the dollar gap. If the stronger offer still falls short, keep your other applications moving while you negotiate. Refolk tailors your resume and cover letter to each new posting, so you can keep applying without rewriting from scratch, and a competing offer has a better chance of arriving in time to matter. Refolk's fit score also helps you decide which open processes are worth speeding up when you need a second letter to compare.

Questions job seekers ask

Should I count target bonus in total comp?

Yes, but it counts at full target only in the steady-state figure. In year one, multiply target by the letter's proration fraction, and use $0 if your start date falls after a hire cutoff such as Dropbox's November 1 or EA's January 15. Never count the maximum bonus. Also record the payout date, because letters that require you to be employed on the payment date push the first bonus to whenever that date falls.

How do I value RSUs in an offer letter?

Use the letter's dollar figure on the letter's own vesting schedule. Year one is the year-1 vesting percentage times the grant, taken from the grant notice. Steady state is the grant divided by its vest years. Record the pricing window and rounding direction separately, because the share count is fixed only when the window closes, and share price moves between offer and grant change how many shares you actually receive.

Why does the recruiter's total comp number seem higher than mine?

Usually because the recruiter's number uses a different convention on one or two lines. Common causes are dividing a back-loaded grant evenly by four, counting full target bonus in year one, spreading the sign-on across years, valuing options at the preferred price, or including an unwritten refresher. Grade each line against its clause and write down the dollar difference per line. When every dollar of the gap is explained, you know which assumptions to ask about.

Does a sign-on bonus count in total compensation?

It counts in the year it is paid and is excluded from steady state. This follows the Levels.fyi definition page, although another Levels.fyi source describes spreading sign-on at one quarter per year, so say on your worksheet which convention you used. Next to the sign-on, record the repayment terms, since some letters want the full amount back on any exit in the first year.

How should I value stock options at a private company?

Use $0 in both total comp figures at grant. Under the 409A rule, the strike price must be at or above the common stock's fair market value, so options have no intrinsic value on the day they are granted. Ask in writing for the strike price and the latest 409A value. Record the spread to the latest preferred price as a memo line labeled not cash. Published ratios of common to preferred value vary too widely to assume one.

What is the difference between year one and annual total compensation?

Year one is what the letter actually pays in your first twelve months: base, the prorated first bonus, the sign-on, and the first year's vesting. Annual or steady-state comp is a typical later year: base, full target bonus, and the grant divided by its vest years, with no sign-on. The two figures can differ sharply. On a 5/15/40/40 schedule, year-one equity is one fifth of the steady-state figure.

Put this to work

Paste your career in once. Every application after that is written for you.

Drop a resume or a LinkedIn URL. I rank the live openings against it, rewrite the resume and write a cover letter for the best of them, and fill in the employer's form when you press the button. You read, you decide what goes out.

  1. 01Drop your resume

    A PDF or a LinkedIn URL. About a minute, once.

  2. 02I rank the openings

    Every weekday morning, the live catalog scored against your history. Up to 20 worth your time, not two hundred links.

  3. 03Each one is written up

    Resume rewritten for the posting, a cover letter, a fit score. Press send, or let me fill in the form.

  • New matches ranked and written before you are up.
  • Every bullet stays inside what your history supports. Nothing invented.
  • Queued, submitted, interviewing, offer: one screen, not a spreadsheet.

Free, no card. Up to 25 applications sent a week, and nothing is sent until you say so.

Read next