Startup Compensation Strategy: Building a Structure You Can Still Defend at Fifty People

How do I choose a position relative to the market I recruit from?

Decide what you are competing against before you decide what to pay. A company recruiting from large technology employers faces a different reference point than one recruiting from agencies or from other early-stage companies, even for identically titled roles. Write down which pool you are actually drawing from, because that pool sets the expectations you will meet at the offer stage.

State your position deliberately rather than discovering it. Some companies choose to lead on cash and grant less ownership, which suits candidates with immediate obligations and reduces dilution. Some choose the reverse, which conserves cash and concentrates the return in an uncertain future event. Some hold cash near the middle of their reference pool and differentiate on scope and autonomy instead. All three are defensible positions and the failure is having no position, which produces an outcome set by whoever negotiated hardest.

Recognize that the position has consequences for who says yes. Weighting heavily toward ownership systematically selects for candidates who can absorb risk, which correlates with financial circumstances rather than with capability. If you care about the range of people you can hire, that trade-off is worth making consciously rather than as a side effect of conserving cash.

What is the simplest level structure that will still work later?

Start with three or four levels per function and describe each in terms of scope and independence rather than years of experience. The useful distinctions are typically: delivers well-defined work with support, owns an area independently and handles ambiguity within it, and sets direction for an area and raises the capability of others. Those descriptions survive growth; a ladder built on years does not.

Attach a range to each level rather than a single figure, and define what movement within the range means. Without that definition, everyone is placed at the midpoint by default and there is no way to recognize someone performing beyond their level without promoting them into a scope they are not ready for. The range is the mechanism that lets performance and level be separate things.

Keep the whole thing on one page. Elaborate frameworks copied from large companies describe a coordination problem you do not have yet, and they cost more in maintenance than they return in clarity. You can add specificity later; you cannot easily remove a structure people have already been placed into.

How should I think about the cash and ownership mix?

Treat them as separate instruments with different purposes. Cash compensates for work done and is what most people budget their lives around. Ownership is a claim on a future outcome that is uncertain in both timing and value, and its actual worth to the holder depends heavily on the terms attached and on the tax rules where they live. Presenting them as interchangeable in an offer conversation misleads a candidate who has not been through an outcome before.

Be precise about what you are granting. Instrument type, vesting schedule, any cliff, what happens on departure, how long a departing holder has to act, and what happens in a sale or a financing all materially change the value of the same headline number. These mechanics vary by jurisdiction and by company structure, and their tax treatment can differ for the same instrument in two countries, so they should be drafted with counsel and explained accurately rather than approximately.

Some companies offer candidates a stated choice between more cash and more ownership within defined bounds. It respects that people's circumstances differ and it makes the trade-off explicit rather than implicit. It requires a published exchange rate and the discipline to hold it, or it becomes a negotiation channel that reproduces exactly the inconsistency you were trying to avoid.

How do I handle people in different locations?

There are three broad approaches and each has a real cost. Paying a single global rate regardless of location is simple and removes an entire category of dispute, and it means paying above local reference points in some markets, which some founders regard as a recruiting advantage and others as an inefficient use of cash. Adjusting by location tracks local reference points and creates ongoing questions about who decides the tiers and what happens when someone moves.

The third approach uses a small number of broad tiers rather than per-city adjustment, which reduces the administrative burden and the number of edge cases while still acknowledging large differences. Whichever you choose, write down the rule and the reasoning before you have to apply it under pressure, because the first time someone relocates you will be setting precedent whether you intend to or not.

Employment obligations, statutory benefits, and payroll costs vary independently of market pay rates, so the same package costs the company different amounts in different countries. Model the fully loaded figure per location rather than comparing headline salaries, and take local advice on what must be provided, since some elements are not optional and are not negotiable.

How much should I disclose internally about pay?

Disclosure exists on a spectrum rather than as a binary. Publishing the structure, meaning the levels, the ranges, and the philosophy, without publishing individual figures is a common middle position: people can see whether they are treated consistently without the discomfort of individual comparison. Publishing individual figures is a stronger commitment that requires the structure to be genuinely defensible first.

Note that this is not purely a matter of preference. Pay transparency requirements differ by jurisdiction, and in some places employers face obligations around disclosing ranges in job postings, responding to candidate or employee requests, or reporting on pay gaps. What applies to you depends on where you are and where your employees are, so confirm the requirements locally rather than assuming a policy choice is entirely yours.

Whatever level you choose, expect it to be tested. Somebody will eventually compare notes with a colleague, and the question you will be asked is not what they earn but why the difference exists. If you can answer with the level definitions and the range, the conversation is manageable. If the answer is that they negotiated harder eighteen months ago, you have a structural problem that disclosure merely revealed.

When and how should compensation be reviewed?

Set a fixed review point rather than responding to individual requests as they arrive. Ad hoc adjustments reward the people most comfortable asking, which is a poor proxy for contribution and reliably produces the inconsistency that becomes expensive later. An annual or twice-yearly cycle also lets you look at everyone together, which is the only way to see whether the structure is holding.

Separate the two questions the review answers. Has this person's level changed, meaning their scope and independence are now different, and is their position within the range still right given performance and how the market has moved? Conflating them means the only way to recognize good work is a promotion, which pushes people into scopes prematurely and creates a different problem a year later.

Additional ownership grants after the initial one are common as tenure extends and the original grant approaches full vesting, and companies handle them very differently: some on a schedule, some at promotion, some case by case. Whatever the approach, the same jurisdictional and tax considerations apply as with the initial grant, and consistency of rule matters more than generosity of any individual decision.

What compensation mistakes are hardest to reverse?

Paying an early hire well above the structure to close them. It is invisible until you build the structure, at which point that person sits outside it and every subsequent decision either compounds the inconsistency or requires a conversation nobody wants to have. If you must exceed a range to close someone, record the reason at the time and treat it as a known exception rather than a new precedent.

Inflating titles instead of compensation. A title costs nothing today and constrains everything later: it sets the person's external market expectation, it complicates hiring a more experienced person above them, and it makes any future correction feel like a demotion regardless of intent. Scope descriptions are the honest alternative and cost you nothing.

Making ownership promises verbally or in vague terms. A grant that was discussed but never documented, or documented with terms the recipient did not understand, is a dispute waiting for a liquidity event. Put every grant in writing, ensure the recipient has the mechanics explained accurately, and encourage them to take their own tax advice, because the treatment depends on their personal circumstances and where they are resident.

What to take away

• ▸Name the candidate pool you compete against before choosing where to sit against it.
• ▸Three or four levels described by scope and independence, each with a range, is enough for a long time.
• ▸Treat cash and ownership as separate instruments with different risk profiles, and explain the terms precisely.
• ▸Write the location rule before someone relocates, and model fully loaded cost per country rather than headline pay.
• ▸Publishing the structure without individual figures is a common and defensible middle position.
• ▸Record any exception to a range at the time, with its reason, so it does not silently become the new rule.