Optionality Inside the Job You Still Have

Maya's team went from nine people to four in two years.

She is one of the four, which she is grateful for and uneasy about, because the reason she survived is the reason the next round will need fewer people like her.

Most advice for her position is either denial or dramatics: hope it passes, or quit and start something. Both miss what she actually holds.

Maya's salary produces a surplus of about $27,200 a year. Over three years that is $81,600 of funding for her own conversion, with no investor, no equity given away, and no obligation to repay it.

A founder raising that amount would give up a slice of a company and answer to someone for its whole life. She has it already, and the only condition is that she keeps showing up.

The job is not the thing to escape. For two or three years it is the thing to use.

Your Salary Is the Cheapest Capital There Is

Every other source of money for a venture costs something permanent. Equity investors take ownership, lenders take a fixed claim and a repayment clock, and savings once spent are gone.

A salary arrives repeatedly, costs no ownership, and pays for the runway that lets you refuse bad terms later.

A job is also where the raw material sits: customers, processes, data, and a close view of what companies pay to have fixed.

The correct posture is not loyalty and not resentment. It is deliberate use.

What Optionality Means Here

An option is a right without an obligation, and its value comes from asymmetry: a small known cost against a large uncertain payoff.

Inside a job you can buy three: control rights over the automated systems you build, the moated parts of your role, and a negotiated share of what you create rather than a description of what you are called.

The premium is time, political capital, and the discomfort of asking for things people rarely ask for. The payoff is that when the role changes, you own something that does not change with it.

Negotiate for Ownership, Not Title

Titles are cheap to give and worth nothing on the day the function disappears. Ownership is expensive to give, which is why you must ask precisely.

Ask for one or more: equity or a profit share, a defined share of a measured saving, a licence to the methods you develop rather than a full assignment, named accountability on the output, or continued access to the customer relationship.

Suppose the supervised production pipeline Maya designed removed about $480,000 a year of outside agency and contractor spend. A $12,000 raise is 2.5 percent of that, paid once and absorbed into her base.

A 5 percent share of the verified saving is $24,000 in the first year, and it rises with the saving rather than with her title.

Some employers will say no, and a clean no still tells you where the upside is meant to go. Tax and legal treatment of equity, options and profit shares varies by country, so get advice locally first.

Own the Workflow

When a company automates a function, someone holds the controls: the specification, the quality standard, the evaluation set that decides whether output is acceptable, the vendor contracts, and the logins.

That person has control rights, the seventh of the seven scarcities, and the most available form of ownership for an employed reader.

Practical ownership means only you can run the thing. It is real, temporary, and decays the moment someone else documents the process.

Contractual ownership means a written agreement says what is yours: a licence, a revenue share, a right to use the method elsewhere. It survives reorganisations and your leaving.

Convert practical ownership into contractual ownership while it is still scarce, because scarcity is what you are negotiating with.

Build the Moated Parts on Purpose

Inside almost every job, some tasks resist automation for structural reasons rather than technical ones.

Someone has to be accountable for the decision, hold the difficult client, and make the call where the data is ambiguous and the consequence is real.

These parts are usually the least visible in a performance review. Spend your discretionary effort there.

Keep a short log of decisions you made, the reasoning, and what happened. That log is evidence of judgment, and evidence of judgment is what you sell when nobody is buying output.

Read the Contract Before You Build Anything

Employment and contractor agreements often contain intellectual property assignment clauses, invention clauses reaching work done outside hours, non-compete and non-solicitation terms, and moonlighting rules.

Enforceability varies enormously by jurisdiction: some places void non-competes almost entirely, others enforce them broadly. It is worth one paid hour with a lawyer where you live.

Three habits reduce risk in most places: build on your own time and equipment, keep your project in a different domain from your employer's business, and keep dated records of what you made and when.

If your employer has a process for approving outside work, use it. One line of written approval beats a year of assuming.

Stage the Exit

Nobody should jump. A jump converts an option into an obligation, and obligations force people to take bad work.

Stage it in four phases.

Test, zero to six months. Five hours a week and a budget under $500. The objective is not a product; it is three conversations with people who might pay and one small paid pilot, because paid is the only evidence that counts.

Parallel, six to eighteen months. Ten to fifteen hours a week delivering repeatedly to real customers, while the salary continues and the surplus machine runs. The objective is a repeatable process and a second and third customer who did not come from a friend.

Threshold. Two conditions, both required, both measured rather than felt.

Condition one: side income has covered your lean monthly base for three consecutive months, or side income plus runway covers the gap for at least eighteen months. Condition two: your runway is at its target.

Transition. Only when both conditions hold, and even then with the contract read, the accounts separated, and three months of work already booked.

Maya shows why the second version of condition one exists. Her lean base is $6,150, a large monthly number for a new service. If her side work reaches $2,750 a month, the gap is $3,400, and her $61,000 runway covers it for about eighteen months.

That is a defensible threshold. A hopeful feeling in month four is not.

The Option Also Decays

Staying is not free.

Three signals say the option is losing value: your contribution is fully documented and could be reproduced by someone junior, headcount in your function keeps falling while output rises, and you have stopped learning anything worth money elsewhere.

When two are true, move your timeline forward rather than waiting for the decision to be made for you.

What the Three Readers Do

Maya

Maya stops asking for a promotion and asks for three things instead.

A share of the measured saving from the production pipeline she built, proposed at 5 percent of verified reduction in external spend, which would be about $24,000 on $480,000. A licence to use the specification and evaluation framework she wrote for non-competing purposes. And her name as the accountable approver on the output that goes to clients.

She gets the third immediately, a version of the first as a bonus formula rather than a standing share, and a no on the second, which tells her where the company believes its value lives.

She keeps the salary, runs the surplus machine at $27,200 a year, and starts the test phase: five hours a week advising mid-sized firms on supervised content pipelines, with the threshold rule written down before she begins.

Tom

Tom has no employer, so his version is optionality inside the client relationships he still has.

His work is priced per word, the unit whose price collapsed. He proposes a different unit to two long-standing industrial clients: a monthly retainer of $750 each for reviewing, correcting and signing off automated translations of regulatory and equipment documentation.

That is $1,500 a month, contracted rather than per project, and it is the income that turns his deficit into the $1,129 monthly surplus in Lesson 13.

The negotiation is not about rates. It is about moving from producing text, which is abundant, to being the accountable reviewer, which is not.

He also asks permission to publish anonymised notes on the documentation errors he finds, the beginning of a reputation asset.

Leo

Leo has the least to negotiate with and the longest time horizon, which is a better position than it feels like.

He built the automated onboarding sequence his team depends on. He asks to be named its owner, to hold the administrator access, and to write the documentation himself, which costs his manager nothing and gives Leo control rights over a system the company relies on.

He also reads his contract, finds the outside work clause, and gets written approval for unrelated freelance projects before starting one.

Then he runs the numbers he already has: $880 a month of surplus, a $7,200 runway target, and a lean base of $2,410 his side work must eventually cover for three straight months. An unglamorous plan with dates on it is the only kind that survives.

Worksheet

1. Calculate your annual surplus and multiply by three. That is the funding your job provides.
2. List the automated or semi-automated systems in your role and name who holds the specification, the quality bar and the logins.
3. Write the one measured saving or revenue you can credibly claim credit for, in dollars, with the evidence.
4. Draft the specific ask: equity, profit share, share of saving, licence, accountability, or customer access. One sentence each.
5. Read your employment or contractor agreement and copy out the IP, invention, non-compete and outside-work clauses.
6. Book one paid hour with a lawyer in your country if any clause is ambiguous.
7. List the three tasks in your role that resist automation structurally, and redirect your discretionary effort there.
8. Write your threshold: your lean base, your runway target, and the two conditions that must both be true before you leave.
9. Set the test phase now: five hours a week, a budget under $500, three conversations and one paid pilot, with a date.
10. Check the three decay signals and write which of them are currently true.

Common Mistakes

Asking for a title

A title is a description that disappears with the function it describes.

Ask for a share, a licence, named accountability, or a relationship. Each survives a reorganisation.

Building the employer's system with no agreement

The architect who designs the automated workflow and signs a standard assignment clause gives away the most valuable thing they will make that decade.

Negotiate before you build, while your agreement is still needed. Afterwards you are asking for a favour.

Jumping on feeling rather than on threshold

Frustration is not a signal about market demand, and neither is a good month.

Write the two threshold conditions down while you are calm, and hold yourself to the numbers rather than the mood.

Ignoring the contract until it matters

People discover their invention clause during a dispute, the most expensive possible moment.

Read it this week, and confirm locally, because rules vary by country and by state or province.

Staying past the decay

The cost of staying rises as your contribution becomes documented and reproducible.

Two of the three decay signals is not a warning to consider. It is a date to bring forward.

The RW Finance Perspective

Long-term investors think in terms of optionality constantly, though they rarely use the word.

A business with a strong balance sheet, a durable advantage and cash coming in has choices: invest when competitors cannot, wait for a better price, or walk away from a bad deal. A business without those things accepts whatever terms it is offered.

That capacity to choose is why we examine financial strength, moats and management behaviour before we look at price.

The same structure applies to you. Runway is the balance sheet, the debt order is the obligation profile, the surplus machine is the cash generation, and optionality is what those three buy you.

Which is the answer to the question Part III opened with: how do you make a transition without going broke in the middle of it? You do not jump, and you do not wait. You build the runway first, you make sure no fixed obligation depends on your best year, you manufacture a surplus and automate it, and then you convert gradually while the old income is still paying, crossing only when two measured conditions are both true.

Stability is not the opposite of ambition here. It is what makes ambition affordable.

Part IV turns to the other side. You now have a position that can survive a transition; the next lessons are about what one person with AI and a clear problem can build, and why production has stopped being the constraint.

Key Takeaways

• A salary is the cheapest capital most people will ever have: it costs no ownership, requires no repayment, and arrives again next month.
• Titles disappear with the functions they describe, so negotiate for equity, a profit share, a share of a measured saving, a licence, or the customer relationship.
• A 5 percent share of a verified $480,000 saving is $24,000 and grows, while a $12,000 raise is 2.5 percent of the same value and is absorbed into your base.
• Control rights over an automated workflow are the most available ownership inside a job, and practical ownership should become a written agreement while it is still scarce.
• Read your IP, invention, non-compete and outside-work clauses before you build anything, and confirm enforceability locally.
• Stage the exit as test, parallel, threshold, and transition, and never jump, because jumping turns an option into an obligation.
• Cross the threshold only when side income covers your lean base for three consecutive months, or side income plus runway covers the gap for eighteen months, and runway is at target.
• You survive a transition by funding it from an income you still hold, not by escaping first and financing it later.