Cap Table Decoded
Cap Table DecodedHow to Read a Startup Cap Table as an Incoming Engineer

How to Read a Startup Cap Table as an Incoming Engineer

Understand ownership percentages and liquidation preferences before accepting your equity offer.

· 9 min read

Reading a startup cap table is the most important step an incoming engineer can take to understand an equity offer. The share count is not the answer. Your ownership percentage, future dilution, and place in the payout order can turn the grant into meaningful money or leave you with nothing.

Why engineers — not just investors — need to read a cap table

A startup sells for a substantial sum, its investors get their money back, and its engineers receive nothing for their shares. That outcome can come straight from the ownership terms agreed to before those engineers joined.

Yet an offer usually puts a share count and a strike price in front of you. Those numbers tell you how many shares you can buy and at what price. They do not tell you how much of the company those shares represent or who gets paid before you. The cap table makes those questions visible. Reading it is equity self-defense.

What a cap table actually contains and how it is structured

Start at the bottom, with the share totals. A cap table records who owns the company, which securities they hold, and their ownership percentages. The basic, or undiluted, percentage counts only the shares that are now outstanding. The fully diluted percentage counts more shares, including options and other shares that could be issued. That second total is the starting point for understanding your offer.

Next, find the employee option pool and its breakdown. The table should show which grants have already been made and which shares are set aside for future grants, and whether the options granted are vested, exercised, or unexercised. The distinction between holding an option and owning a share connects the ownership record to the decisions you will eventually face about buying your shares.

Then read the individual rows and security columns. Rows identify founders, employees, advisors, angels, and venture funds. The columns show common stock, preferred stock issued in funding rounds, options, and convertible instruments such as SAFEs and convertible notes in separate groups. Founders and employees who exercise options usually have common stock, while investors have preferred. The terms for those classes show how each is paid at an exit, including preference multiples and whether preferred shares participate in the proceeds that are left.

Why your share count means almost nothing without the fully diluted denominator

A raw share count is a marketing number until you know what to divide it by. For a simple hypothetical, a grant in a company with a smaller fully diluted share total represents a larger ownership stake. The same grant against a larger total represents a smaller stake. The offer can look identical on its face. The ownership is not.

Your starting calculation is the grant’s share count divided by the fully diluted total. Carta’s formula includes outstanding common shares, preferred stock as converted to common, options, warrants, and unissued option pool shares. That is broader than the issued-share total. Using only currently outstanding shares can make your percentage look larger than the share structure supports.

Convertibles require a separate question because the displayed denominator may exclude them. Carta states that SAFEs and convertible notes are not counted in its fully diluted share count until they convert. For your own model, however, they are future dilution that cannot be ignored. Ask for your fully diluted ownership percentage and confirm whether outstanding convertibles are reflected in that calculation. If they are not, ask for a scenario showing their conversion rather than treating the displayed percentage as the whole picture.

Notes add another source of uncertainty. Unlike SAFEs, convertible notes accrue interest and have a maturity date. A note coming due before the next priced round can lead to conversion, a repayment demand, or renegotiation that creates additional shares. The amount raised through these instruments can change the fraction you end up owning.

How each funding round compounds dilution on your original grant

The denominator keeps moving after you accept the offer. Each priced funding round adds shares for new investors, so those who already own shares have lower ownership percentages. Your original grant can stay the same size while your stake shrinks. Across several rounds, those reductions compound.

A priced round often requires the company to set aside more shares for employee grants before the investment closes. That pre-money expansion dilutes existing holders, including engineers with earlier grants. The pool serves future hiring, but its cost falls on the ownership already in place. New investor shares then add another layer of dilution.

Ask what your stake would look like after the next two rounds with an option pool refresh each time, and model that before accepting. A founding engineer’s meaningful pre-seed stake can become a materially smaller fraction by Series C. At some leading AI companies, several large rounds since 2024 have made it possible for that stake to shrink within a single four-year vesting window. That pace is not universal, though, as the median time between rounds across the broader AI startup population has stretched to its longest in over a decade.

How liquidation preferences determine whether common stock pays out at all

Ownership percentage tells you the size of your slice, but not when you get served. Preferred stock gives investors a claim ahead of common shareholders in an acquisition or wind-down. Those claims form the liquidation preference stack. If the proceeds run out while paying it, common gets nothing.

That is how venture deals usually work, even when a company is doing well. A company can sell for an amount equal to the investors’ preference claims and leave no proceeds for employees’ common shares. The preference multiple helps set the size of those claims. How much capital has been raised helps you understand the deal, but you still need to read the terms.

Participating investors take their preference first, then share proportionally in the remaining proceeds with common holders. They receive money through both parts of the calculation. Non-participating investors instead choose between taking their preference and converting to common to share proportionally. They cannot do both, which leaves a very different outcome for employees at the same exit price.

How to translate the cap table into a personal exit scenario

To turn those terms into an offer decision, run a payout waterfall at three exit values. Use one below the preference stack, one just above it, and one well above it. Keep the assumptions explicit. Your current fully diluted percentage is useful, but an exit model needs the percentage remaining after the dilution you are assuming.

Take a simple example with substantial total preference claims. In this example, investors own a substantial share of the company on a fully diluted basis, and your grant represents a small stake included within the common holders’ share. Treat the grant as fully vested and exercised for this illustration. These calculations show what you get before your exercise cost and taxes. Start with participating preferred so that each stage of the waterfall is visible.

At an exit below the preference stack, the proceeds do not cover the investors’ preference claims. All of the money goes toward the investors’ claims. Common shareholders get nothing, so your payout is zero. Multiplying the headline sale price by your ownership percentage would give you an amount you never receive.

At an exit just above the preference stack, a small remainder is left after the preference is paid. Participating investors then take their ownership-based share of what is left, leaving the rest for common holders. Your grant represents a larger percentage of the common holders’ pool than of the whole company. You therefore receive that larger percentage of the amount left for common holders. The denominator changes when the calculation moves from all shareholders to common shareholders alone.

At an exit well above the preference stack, the same participating structure leaves a substantial remainder after preferences. Investors receive half of that remainder, and common holders share the rest. Your share of the common pool now produces a meaningful payout because there is enough value beyond the preference stack to reach your grant.

At the exit just above the preference stack, the investors in this example would take their preference, with conversion yielding less. Common holders would split the remainder, and you would receive more than under the participating structure. At the exit well above the preference stack, those investors would instead convert and receive their ownership-based share of the proceeds. Your grant would then receive its fully diluted percentage of the total sale proceeds.

This is why the calculation cannot stop at subtracting preferences and multiplying the balance by your headline percentage. The remaining ownership pool depends on who takes a preference, who participates, and who converts. What you get also depends on vesting, exercise costs, and taxes.

Finally, compare those possible payouts with the cash pay you would give up. Founding engineer salaries typically sit below market, so foregone salary is part of the price of the bet. For AI/ML engineers, the higher pay offered at large technology companies since 2024 can make that opportunity cost especially high. Consider whether you would invest that cash equivalent directly in this company at its current valuation. An expected-value view weighs the possible outcomes by how likely you think they are, rather than treating the largest payout as the grant’s present value.

Four equity terms beyond grant size that change the actual value of an offer

For negotiation, bring four terms beyond grant size into the conversation. Start with the post-termination exercise window. A standard 90-day window can force you to find the exercise cash and cover the tax bill shortly after leaving. If you cannot exercise within the window, you lose the options regardless of their potential value. Ask before signing whether the company offers a multi-year window, as an increasing number of employee-friendly startups do.

Next, ask about double-trigger acceleration. This protects unvested equity when the company is acquired and you are then terminated. An acquirer can terminate you after closing, leaving you without those unvested shares. This protection is increasingly common and directly negotiable.

Early exercise is the third term to discuss. It lets you buy shares before they vest and file an 83(b) election, which could shift future gains from ordinary income treatment to long-term capital gains treatment. It can be especially valuable early on, when the strike price is low and the spread at exercise is minimal. Timing is part of the decision. If you can use early exercise, you need to act on the exercise and election soon after joining.

The fourth discussion is about anti-dilution protection and whose shares receive it. Preferred investors can have contractual protection in a down round, with their conversion rate adjusted to give them more shares. Engineers holding common have no equivalent protection under that structure. Ask how those terms would affect your stake in a down-round scenario. Unlike a longer exercise window, this is not a benefit to presume is available simply because investors have it.

What to ask for and what a company's willingness to share signals about the org

Ask for the total fully diluted share count, your ownership percentage, and whether outstanding SAFEs or convertible notes are reflected in it. Ask for the preference structure, including participation terms and multiples, and how much has been raised to date. These are the inputs needed to understand the offer. Asking for them is not aggressive.

The response also tells you something about the organization. A company that shares the numbers, explains the preference stack plainly, and can model dilution on request demonstrates operational maturity and respect for your judgment. A company that refuses to share any of this leaves you unable to price compensation it wants you to accept. Its lack of understanding of its own structure or wish to keep you from understanding it is a disqualifying signal.

Raise the questions early, especially in a fast-moving hiring process. Waiting until the offer stage gives a misunderstanding about ownership more time to shape your expectations. A well-run startup should regard these questions as evidence of a careful prospective hire. In your next recruiter or founder conversation, ask them to walk through your fully diluted ownership percentage and explain how the preference stack would affect your payout.

Filed underOffer Valuation

More in Offer Valuation