You’re two or three years into your startup. You have hired a great team and want to retain them. It’s time to consider refreshing their stock options to motivate them to stay longer. How many options should you grant to each employee?

Startups should pay key people market rate to retain them. Otherwise, they may leave the business, lured by the promise of greater compensation elsewhere. Let’s walk through an example.

Assume Anna is a key employee at a startup RedCircle which wants to keep her. Anna received a grant of 100,000 options on her start date. These options have a standard vesting schedule: a four-year vest with a one year cliff.

This means the first 25,000 options vest after Anna has remained with the company for one year. Thereafter, the remaining 75,000 options vest monthly: 2083 options per month. The table below shows her annual vesting schedule through the first six years.

At the beginning of year 3, half of Anna’s stock options has vested. If she were to leave and find exactly the same role at a younger competitor, BlueSquare, she would receive 100,000 options in the second company. (Note: I assume the stock options in both cases are of equivalent value.)¹ At her current company, she has 50,000 options yet to be vested.

If Anna acted as a pure capitalist and considered only share count, she should leave her current startup. She stands to gain more by taking the second startup’s offer: 150k options vs 100k options.

Here we have to introduce an important wrinkle. Over the past two years, RedCircle has increased in value. It has grown and raised capital at a higher valuation. So 75,000 RedCircle options are equal in value to 100,000 BlueSquare options.²

What grant size should RedCircle issue to Anna at Year 3 to provide her a market-value grant? From Year 3 to Year 6, Anna should vest 75,000 RedCircle options to equal the BlueSquare offer. From her first grant, Anna has 50,000 vesting in years 3 and 4.

To present a market offer, RedCircle should grant Anna an additional 25,000 options, which vest over 4 years. This is 6,250 options per year. With this grant, Anna will have 75,000 options vesting over the next four years.

Early in its life, RedCircle should create a refresh grant policy for members of the team. The refresh policy above is one mechanism to do this. By granting more options at the end of the second year equal to or greater than market comparables, the startup will ensure its compensation packages are competitive.

There are alternatives to this approach. Notably, former VC and current Wealthfront CEO Andy Rachleff advocates annual refresh grants equal to 25% of the market grant, vesting over 4 years. I’ve run the math and the difference between the two policies is less than 5% in the scenarios I tested.

The most important thing is to create a refresh grant/retention grant policy early in a company’s life. The next most important thing is to be able to articulate to employees the rationale of the grant sizes. With either rubric, the message is the same: the employee is doing a terrific job and the startup would like to mark their compensation to market periodically. And the outcome is higher retention and greater employee satisfaction.

Thanks to my partner Alex Bard for inspiring this post.

[1] Option math is complicated. It involves predicting the value of a business into the future and considering the strike price of the option grant. Economists have been awarded Nobel Prizes for work on the topic. I’ve greatly simplified this example by considering share count only.

[2] Again I’m simplifying. It’s nearly impossible to compare the value of options in two very early stage companies. But as companies grow, scale and raise capital, the value of a share in that company increases, because the risk of the option decreases. Factors like strike prices and taxes are other important considerations.