---
slug: total-compensation-architecture
type: pattern
summary: "The employer-side framework for designing and pricing a startup offer (salary band, option grant, benefits, and the cash-for-equity tradeoff) so it competes without matching a large company's cash."
created: 2026-05-26
updated: 2026-05-26
related:
  startup-equity-evaluation:
    relation: contrasts-with
    note: "This is the founder building the offer that the candidate-side framework decodes; the same grant, priced from one end of the table and read from the other."
  equity-compensation-types:
    relation: uses
    note: "Pricing a grant means choosing the instrument as deliberately as the size, because ISOs, NSOs, and RSUs carry different tax consequences for the hire and different administrative costs for the company."
  dilution:
    relation: produces
    note: "Every grant the founder issues comes out of the option pool and dilutes the existing cap table, so total-comp design is also dilution management."
  cap-table-hygiene:
    relation: informed-by
    note: "The option pool a founder sizes and spends through total-comp decisions is a cap-table line item, and disciplined grant-by-grant accounting is what keeps the pool from running dry before the hiring plan is done."
  hiring-sequence:

    relation: downstream-of
    note: "The sequencing decision names which role to fill and when; total-comp architecture prices the offer that closes it."
  talent-sourcing:

    relation: complements
    note: "Sourcing fills the top of the funnel with candidates; the compensation package is the offer the funnel carries to a yes."
  runway:
    relation: bounded-by
    note: "Cash compensation is the largest controllable burn line, so the salary band a founder can offer is set by the runway the company is willing to spend against the hire."
---

# Total Compensation Architecture

> **Pattern**
>
> A named solution to a recurring problem.

*Designing and pricing a complete startup offer — salary, equity, and benefits as one package — so it wins the hire the company needs without paying cash the company doesn't have.*

A founder needs a senior engineer and has $160,000 of annual cash to spend against a candidate a public company would pay $230,000. The instinct is to lead with the gap apology and an equity number large enough to feel like it closes it. That instinct misprices the offer in both directions: it gives away more of the company than the role warrants while still reading, to a candidate who knows how to value equity, as a below-market package dressed up with a percentage they can't size. The package is the company's most expensive recurring decision after the product itself, and most early founders assemble it by feel, one negotiation at a time, with no band, no grant ladder, and no account of what the option pool can afford. The result is offers that are simultaneously too generous and not competitive.

## Context

This decision sits on the employer side of the [talent-equity](talent-equity.md) part of the lifecycle, at the moment a founder has decided to make a hire and has to turn a role into a number. It follows the [hiring-sequence decision](hiring-sequence.md) that named which role to fill and when, and it runs alongside the [sourcing](talent-sourcing.md) that fills the funnel. It applies from the first non-founder hire through the early growth-stage team, and it binds hardest pre-Series-A, when cash is scarcest and the equity being spent is at its most valuable.

The package the founder builds here is the input to the [candidate's evaluation](startup-equity-evaluation.md): the same grant, viewed from the other end of the table. The founder prices it against a salary band and an option pool; the candidate decodes it back into expected value. An offer designed well from the employer side survives that decoding and still reads as fair, which is the only kind of offer that closes a candidate who can do the math.

## Problem

A founder must assemble an offer a specific candidate will accept over their alternatives, out of a cash budget set by [runway](runway.md) and an equity budget set by the [option pool](cap-table-hygiene.md), without overpaying on either axis. Underpay on cash and the candidate declines or leaves within a year; overpay and the next two hires don't fit the budget. Underpay on equity and the offer fails to compensate for the salary cut; overpay and the pool runs dry before the hiring plan is done, forcing a dilutive top-up the founder will regret at the next round. Both budgets are constrained, they trade against each other, and the candidate is reading the result with a framework the founder may not have used to build it. Done ad hoc, the package is inconsistent across hires, indefensible when two employees compare notes, and either too rich or too thin for what the role is worth.

## Forces

- **Cash budget versus equity budget.** Cash spends down runway immediately and visibly; equity spends down the option pool and dilutes everyone, but the cost is deferred and easy to underweight. A founder who protects cash by over-granting equity is trading a measured expense for an unmeasured one.
- **Competing for the hire versus protecting the next hires.** Every dollar and every basis point spent on this offer is unavailable for the rest of the plan. A package built to win one candidate at any cost breaks the budget for the team behind them.
- **Consistency versus negotiation.** A defensible structure prices roles by level and stage so two peers see comparable offers; ad hoc negotiation rewards whoever pushes hardest and produces pay bands that can't survive the day employees compare numbers.
- **The instrument's hidden cost to the hire.** The [form of the grant](equity-compensation-types.md) decides the candidate's tax bill and exercise economics, so a nominally generous grant in the wrong instrument can be worth far less to the hire than it costs the company. Pricing only the headline number misses this.
- **Competing on total value versus competing on cash.** A startup will rarely win a cash bidding war against a large company and shouldn't try. The package has to compete on a different axis — ownership, scope, the expected value of the equity — which only works against candidates who value those things, and only if the offer states them in terms the candidate can verify.

## Solution

**Price the offer as one package built from a salary band, an equity grant sized by a level-and-stage ladder, and benefits — spending against an explicit cash budget and an explicit option pool, and naming the equity in terms the candidate can verify.** The package is a structure the founder designs once and applies consistently, not a number invented per negotiation.

The four components, each priced deliberately:

1. **Salary band by role and level.** Set a band for the role using current stage-specific benchmark data, then position within it by seniority and how badly the role is needed. The standard reference is Carta's compensation data, which puts median early-stage new-hire engineering pay near $189,000 in 2025; bands shift by function, geography, and stage. The aim is a band the company can defend across hires, not a one-off number, and one the [runway](runway.md) can carry for the role's expected tenure.
2. **Equity grant sized by a level-and-stage ladder.** Size the grant by the hire's seniority and the company's stage, not by the salary gap. A common early-stage ladder runs from roughly 1–2% for a first key engineering or executive hire down toward 0.3% by the fifth or sixth employee and lower still as the company matures and each percentage point represents more value. Grant the same level the same percentage; let the ladder, not the negotiation, set it.
3. **The instrument, chosen for the hire's stage.** Choose the [instrument](equity-compensation-types.md) so the grant lands as favorably for the hire as it can: ISOs while the 409A is low and the recipient is an employee, NSOs for non-employees or grants past the $100,000 ISO line, RSUs once the share price has climbed past where options make sense. The instrument is a pricing decision, not a legal afterthought.
4. **Benefits and the non-cash case.** Health coverage, retirement access, and the increasingly expected flexible-work terms are table stakes that a candidate notices only when they're missing. Past the table stakes, the durable non-cash advantages a startup actually has are scope, ownership, and growth, and these belong in the offer conversation, stated plainly, not as a substitute for honest numbers.

Then communicate the package as a probability-honest whole. The single highest-trust move a founder can make is to give the candidate the inputs their [evaluation](startup-equity-evaluation.md) needs — the fully-diluted percentage, the strike, the preference stack, the realistic exit scenarios — rather than a percentage and a valuation-derived "value." A founder who hands over the numbers signals a clean cap table and a culture of candor, and a candidate who can verify the offer is far more likely to accept it than one asked to take a flattering number on faith.

> **💡 Size the pool against the plan, not the round**
>
> The option pool is sized at a financing round (the standard ask is 10–20% of post-money), and the temptation is to grant generously early while it looks abundant. Map every planned hire's grant against the pool before spending any of it. A pool that runs dry mid-plan forces a top-up at the next round, which dilutes the founders and existing team precisely when a clean cap table matters most for diligence. Disciplined grant-by-grant accounting against the hiring plan is what keeps the pool from becoming an emergency.

## How It Plays Out

Consider a seed-stage company making its fourth hire, a senior product manager. The founder has run two prior hires by feel: the first engineer got 1.2% after a hard negotiation, the second got 0.4% because she didn't push. Now the product manager is asking what's standard. With no ladder, the founder has no answer that survives the moment those three employees compare notes, and they will. The fix is to build the ladder retroactively: price the role at a band and a grant level, document the rationale, and accept that the second engineer's grant was low and may need a true-up. The cost of the missing structure is paid here, in an awkward correction, rather than avoided.

The cash-versus-equity tension shows up most sharply against a strong external offer. A founder wants a candidate weighing a $230,000 package at a public company and can offer $165,000 in cash. Closing the $65,000 annual gap with equity alone would require a grant far above the role's ladder level, blowing a hole in the pool. The package that actually competes does something different: it sets cash at the top of the defensible band, sizes equity at the role's ladder level, and then makes the real case on expected value and scope — what the grant is worth across the realistic exit distribution, and the ownership of a product surface the candidate would never get at the larger company. The founder who instead tries to win on a grant large enough to match the cash gap pays for it twice: once in dilution now, and again when the next two hires don't fit what's left of the pool.

The 2025 data sharpened one corner of this. Carta's reporting through 2025 and into early 2026 showed equity grants for AI-engineering roles rising materially against the broader market as demand for that talent outran supply. A founder hiring into that role is pricing against a moving band, and a ladder built on last year's numbers reads as below market to a candidate who knows the current ones. The discipline isn't to abandon the ladder; it's to date it and refresh the benchmark where the market is actually moving.

## Consequences

**Benefits.** A founder who builds the package as a structure prices every hire consistently, defends the offers when employees compare them, and spends both the cash and the option pool against an explicit plan rather than discovering mid-year that one is exhausted. The offers compete on the axis a startup can actually win — ownership and expected value — instead of a cash war it will lose. And handing the candidate verifiable numbers turns the offer into a trust signal, which closes the candidates worth closing and screens out the ones who'd have resented the equity later. The structure also makes the cap table legible: grants priced off a ladder and accounted against the pool are exactly the clean record [diligence](due-diligence.md) rewards.

**Liabilities.** Benchmarks are noisy, stage-dependent, and out of date the moment the market moves, so a ladder is a starting position that needs refreshing, not a fixed truth. A rigid ladder can also cost a genuinely exceptional hire who's worth breaking it for; the structure should inform the exception, not forbid it. Building the package well takes information a first-time founder may not have — current bands, pool math, instrument tradeoffs — and getting it wrong in either direction is expensive: too thin and the hire declines or churns, too rich and the plan breaks. The package competes on equity value, which means it only works with candidates who can and will value equity; against a candidate who only counts cash, the most carefully built startup offer still loses, and recognizing that early saves everyone the negotiation.

## Sources

- [Carta's State of Startup Compensation](https://carta.com/learn/) — the benchmark source for salary bands by role and stage, grant sizes by seniority, option-pool norms, and the 2025–2026 movement in equity grants for in-demand engineering roles that the pricing decisions here reference.
- Kruze Consulting's startup-compensation guidance — the accounting-firm practitioner reference on how early-stage companies actually set salary bands, size option pools against a hiring plan, and structure benefits under cash constraints.
- The US Internal Revenue Code's treatment of incentive and non-qualified stock options (IRC §422, including the $100,000 ISO limitation) — the statutory basis for the instrument choices that determine how a grant lands for the hire.
- Y Combinator's [equity and hiring guidance](https://www.ycombinator.com/library) — the canonical early-stage articulation of grant-level ladders for the first employees and the logic of competing on ownership rather than cash.

---

- [Next: Fractional Executives and Contractor Talent](fractional-contract-talent.md)
- [Previous: The Experience-and-Age Paradox](experience-age-paradox.md)