The Two-Offer Comparison Framework for Different Pay Shapes

You have two offers and they refuse to line up. One is a bigger base at a stable company; the other is a smaller base with a large equity grant and a scope jump. This guide is for candidates holding or expecting two offers with different pay shapes, and it gives you one repeatable judgement: convert everything to annualized, risk-adjusted, take-home dollars, then score the dimensions money cannot express on a weighted grid. Follow it and you will have a defensible decision instead of a coin flip.

Why two offers refuse to line up

The core problem is that the two offers are denominated in different currencies of value: guaranteed cash, probabilistic equity, scope, and stability. You cannot compare them until you translate everything into one unit, and the honest unit is annualized, risk-adjusted, after-tax dollars.

Base salary is the trap. It is the easiest variable to measure, so it dominates the decision even when it should not. But in a typical professional role, non-salary components add 20 to 50 percent to the value of a compensation package, and a private-company engineering offer skews far harder: 30 to 60 percent of total compensation can arrive as equity. If you compare bases, you are comparing the smallest honest slice of two very different pies.

The definition to anchor on: total comp equals base, plus the equity grant divided by its vesting years, plus bonus, plus benefits. Everything downstream is refining each of those terms until it reflects what actually lands in your account.

The offer is not a number. It is four different currencies pretending to be one.

The framework in one picture

The framework runs in two passes: a money pass that produces one dollar figure per offer, and a judgement pass that scores the dimensions dollars cannot hold. You do the money first to strip out the base-salary anchor, then the grid to catch what the spreadsheet misses.

There is one genuine order disagreement worth naming. Most sources start with the money math so you can see the real gap. A minority insist you write your three non-negotiables and three deal-breakers before looking at any numbers, so the base-salary figure never anchors your priorities. Both are defensible. My recommendation: write the non-negotiables first as a sealed envelope, then do the money, then open the envelope. That way the anchor never touches your priorities but the numbers still lead the comparison.

The money math: convert everything to annualized dollars

The rule is simple and it reorders offers constantly: divide any lump into the years you will actually hold it, and value every hidden component. A $20,000 signing bonus over a two-year stay is $10,000 per year, not $20,000.

Work through the full component list for each offer. The full picture includes target bonus, equity or stock options, retirement contributions including employer match, health insurance value, paid time off, professional development budgets, and any signing or relocation bonuses. Missing any one of these is how offers get mis-ranked.

The match deserves special attention because it is invisible on the base line and enormous in your account. A 6 percent match on a $100,000 salary is $6,000 per year forfeited by choosing the other job. A double-digit contribution on salary plus bonus can wipe out a five-figure base gap by itself, which is exactly why annualizing everything reorders offers so often.

Once both offers are annualized, normalize for where you will live. Run each gross through a free take-home calculator for the specific state, then add a realistic rent for the actual neighborhood, and compare what is left. A higher gross in a high-tax, high-rent metro can land below a lower gross elsewhere. This is a common source of misranking that has nothing to do with equity at all.

Risk-adjusting equity: the arithmetic, not the mood

The rule for equity is that face value lies, and the correction is arithmetic. Apply a 30 to 60 percent discount to private RSUs or options depending on stage, then multiply the projected after-tax value by your exit probability.

Quantify it directly. If there is a 25 percent chance of exit, multiply the projected after-tax value by 0.25; that is your expected value. Model the outcomes at conservative, base, and optimistic valuations, including dilution, taxes, and the vesting schedule. The optimistic case is a story; the conservative case is your floor.

Why discount at all? Because the base rates are brutal, and the discount is a rough proxy for them.

Table C - Startup downside base rates

With 75 percent of venture-backed firms never returning cash, an undiscounted grant overstates value by roughly the inverse of the exit probability. The 30 to 60 percent discount is not you being a pessimist; it is you pricing the base rate that Shikhar Ghosh's 2,000-company Harvard Business School dataset already measured. Note the split by funding model: a bootstrapped company's 58 percent five-year survival is nearly double the venture-backed 32 percent, so the funding story changes the discount you should apply.

Even the stable offer carries in-year risk. In one tracked year, 123,941 tech employees were laid off at 269 companies. A stable-company offer is safer, not safe, and that belongs in the stability score rather than in a false sense of a guaranteed number.

Read the equity fine print before it reads you

The equity terms decide whether a grant is an asset or a liability, and four terms do most of the damage: the cliff, the exercise window, the tax type, and the strike price. Get these on paper before you sign anything.

The standard structure is four-year vesting with a one-year cliff: 25 percent vests at the cliff, then monthly after. Cliffs are near-universally one year, at least 95 percent of the time. The cliff is binary. Leave on day 364 and you walk away with zero; leave on day 366 and you keep a year's worth. Check your expected tenure against that date.

The exercise window is the quiet killer. The default post-termination exercise period is 90 days; miss it and vested options are forfeited. As of one measurement, only about 20 percent of terminated grants extend that window beyond 90 days. With a high strike price and only 90 days to fund the purchase after leaving, you may forfeit equity you already earned. This is why you should likely avoid private companies with 90-day windows unless you are personally wealthy or senior enough to negotiate it out.

Taxes finish the picture. ISOs owe no regular income tax at exercise, but the spread becomes an AMT preference item, so an exercise can trigger a tax bill on paper gains you cannot yet sell. NSOs are ordinary income at exercise. For an ISO qualifying disposition you must hold two years from grant and one year from exercise, and a $100,000 annual vesting limit applies. If you take away one thing: a "vested" option is not free money, it is a future purchase with its own tax event.

Scope and stability: the dimensions dollars miss

Scope is scarcer than seniority, and market thinness is a real input to your stability score. Both are dimensions the money pass cannot see, and both are quantifiable from role-population data rather than guessed.

Start with scope. In Refolk's index of professional profiles, the higher the level, the thinner the supply, which means a scope-jump offer is worth more than a pay bump because the level itself is rare.

Table A - Role scarcity by level (United States)

A Staff title is 5.6 times rarer than Senior, and a Founding Engineer 7.4 times rarer than Staff. If one offer moves you up a level, you are acquiring something in genuinely short supply, and that is legitimate leverage in negotiation and a real weight in the grid.

Stability has a geographic component people forget. Your re-employment pool at the same level is not infinite, and it varies enormously by market.

Table B - Same title, two markets (Staff Software Engineer)

The UK Staff pool is 6.6 percent the size of the US one, roughly 15 times smaller. A UK candidate weighing a stable offer against a startup faces a much smaller re-employment pool at the same level if the startup fails. That thinner pool raises the real cost of the risky offer, so it belongs directly in the stability score. To see how the level you are weighing populates in your own market, Refolk can count current holders by title and geography before you set that weight.

Weight and total the matrix

The rule for the judgement pass is to score each offer across the factors you actually care about, weighted by your priorities and not the recruiter's pitch. Before scoring anything, write down your three non-negotiables and your three deal-breakers.

Build a weighted decision matrix. List the factors, assign each a weight that reflects how much it matters to you, score each offer 1 to 5 on each factor, multiply, and total. The factors that consistently earn a place: risk-adjusted total comp, manager, scope, stability, growth, and culture.

Factor                     Weight   Offer A (1-5)   Offer B (1-5)
Risk-adjusted total comp     30           _              _
Manager and team             20           _              _
Scope / level jump           15           _              _
Stability (incl. market)     15           _              _
Growth trajectory            10           _              _
Culture / values fit         10           _              _
------------------------------------------------------------------
Weighted total (sum W x S)  100          ___            ___

The matrix does two things a spreadsheet cannot. It forces you to name what you value before the numbers anchor you, and it makes the trade-offs visible so a scope jump can legitimately outweigh a modest cash gap. Do not skip the weights step; equal weighting quietly re-anchors you on whatever factor has the most sub-items.

How this goes wrong

Most bad offer decisions are not close calls; they are one of a short list of predictable errors. Here is the catalogue, what each looks like when it lies, and how to check it.

• Anchoring on base salary. Base is easiest to measure, so it dominates. Check: is the higher-base offer still ahead after equity, match, and cost of living are added? Often it is not.
• Face-value equity. A big grant number that assumes no dilution, no taxes, and a certain exit. Check: recompute with the 30 to 60 percent private discount and an exit probability applied.
• Ignoring the exercise window. A vested option is worthless if you cannot fund the exercise inside 90 days. Check: what is the strike price times your vested shares, and could you write that check on short notice?
• Cliff blindness. Leave on day 364 and you get zero. Check your expected tenure against the one-year cliff date before you value any equity.
• AMT surprise on ISOs. Treating an ISO exercise as tax-free. The bargain element becomes an AMT preference item even with no regular income tax due. Check: model the AMT on the exercise you actually plan to make.
• Signing-bonus mirage. Check the clawback clause. If you must repay the bonus after leaving within 12 months, you are effectively accepting lower-than-advertised pay for the first year.
• Optimism about the growth trajectory. Budget for the long-term worst case rather than assuming the growth keeps compounding. The optimistic exit is a story, not a plan.
• Faking leverage. Never invent an offer you do not have. The bluff falls apart the instant they call it, and it can cost you both jobs.

The procedure, end to end

Run these eight steps in order. Most take an hour or less; the diligence at the end takes a day or two of waiting. Done correctly, you finish with one weighted number per offer and a reconciled decision you can defend to yourself in six months.

The last step is where most people stop too early. Talk to one person who works at each company, cold messaged as a current employee. Ask what the matrix cannot show: what the manager is like on a bad week, whether the last round of layoffs hit their team, how equity refreshes actually work.

If the scorecard says one job but your gut says the other, that mismatch is data, not noise. It usually means you under-weighted a factor you actually care about. Go back to your weights, raise the one your gut is defending, and re-total. The disagreement is a weighting bug, not a tie.

To pressure-test the risky offer's downside before you decide, find people who have already lived it. A search for engineers who left the exact kind of company you are considering tells you what the exit actually looked like.

Before you sign: the final check

Run this list before you accept either offer. Each item is a place where a signed decision quietly went wrong for someone else.

Keeping the comparison current

The framework is evergreen; the inputs are not. Two numbers move under you and both are re-checkable. First, the private-equity discount tracks the exit environment, so re-derive it from current survival base rates rather than reusing an old figure. Second, the layoff picture and the market pool at your level change constantly, so re-count the re-employment pool for your title and geography when you are actually deciding rather than trusting a number from a prior search. The tables here are a snapshot from Refolk's index; the method for reading them survives, and Refolk can regenerate the counts for your exact title and market when you need them fresh. Keep the matrix template, throw away the numbers, and re-run it for every pair of offers.