If your first engineer joined on a standard four-year vesting schedule, their equity stops growing on their fourth anniversary. From that day, staying earns them salary and nothing else, while any competing offer comes with a fresh four-year grant. Plan a refresh grant around year two or three, not in the week they hand in notice. Size it on the role they hold now, not the role you hired them for, and consider starting it when the original grant ends so their equity never stops.
Many founders find out about this problem when the engineer who knows the whole system starts taking calls from recruiters. By then you are negotiating against a deadline. This post covers how to see it coming and what to do about it.
Why year four is the quiet risk
A typical early grant vests over four years with a one-year cliff. That structure does real work for retention in years one to three: every month the engineer stays, they earn more of a stake they believe in.
At month 48 the schedule ends. Carta, which runs cap tables for a large share of US startups, has written about the drop-off in retention around the end of the initial grant, and its data shows that close to 70 percent of employees who reach the end of their first grant receive a refresh (Carta on equity refreshes). The companies that keep their early engineers have usually planned for this. The ones that lose them usually have not.
For a startup, losing the first engineer at year four is expensive in a specific way. They are often the only person who knows why the system works as it does. Our post on what to do when your only engineer resigns covers the damage control, but the cheaper path is not getting there.
The three ways to structure a refresh
There is no single right structure. These are the three I see most, with the trade-offs.
Annual top-up grants
Each year, usually tied to the review cycle, you grant a smaller amount that vests over four years. Over time the engineer has several overlapping grants, so their vesting never drops off a cliff. This is the most common pattern at larger startups.
Trade-off: it needs discipline and a stable review process. At a company with three engineers and no formal reviews, annual grants tend to be skipped in a busy year and then forgotten.
One meaningful refresh at year two or three
You make one larger grant partway through the original schedule, sized to keep the engineer's total ownership in line with their current role. It vests over four years from the grant date.
Trade-off: simpler to run, but the grant overlaps the original one, so the engineer earns a lot in years three and four and then still faces a smaller cliff later.
A grant that starts when the original ends
Sometimes called a "back-loaded" or "boxcar" grant: you approve the refresh in year three, but vesting starts on the date the original grant finishes. The engineer knows their equity will keep growing without interruption, and you are not paying for overlap.
Trade-off: the engineer sees no immediate benefit, so it needs a clear explanation. For an early engineer who already believes in the company, it is often the cleanest option.
How to size it
Do not size the refresh by asking what feels generous. Use the same logic you would use to hire the role today.
- Write down the role the engineer actually performs now. In many startups the first engineer has become the de facto engineering lead, which is a different job from the one they were hired into. The titles for an early engineering team post helps you name it.
- Estimate what new-hire equity you would offer an outside candidate for that role today, at today's valuation. That is your reference point.
- Look at the unvested equity the engineer still holds. The refresh should bring their forward-looking stake close to what a new hire would get for the role over the next four years, adjusted for performance.
A worked example, with illustrative numbers. Your first engineer joined with 1.0 percent, now 30 months in. They now lead a team of three. You would offer an outside engineering lead around 0.6 percent at today's valuation. Their remaining unvested original grant is about 0.375 percent, ending in 18 months. A refresh of roughly 0.4 to 0.6 percent vesting over four years keeps their forward equity in the same range as a new hire, and they keep the ownership they already earned.
Remember the option exercise window too. If your plan gives departing employees 90 days to exercise, a large accumulated stake can be unaffordable to keep. The post on extending the exercise window explains why some companies lengthen it for long-tenured staff.
What to do before you grant anything
Talk to the engineer first
Ask what they want from the next two years. Some want more ownership; others want a role change, a sabbatical or more cash. Our post on keeping your first engineer argues that boredom, not money, is the most common reason they leave. A refresh that ignores that will not hold them.
Check the option pool
Refresh grants come from the same pool as new hires. If your hiring plan uses most of it, you need to plan a pool top-up with your board, ideally at the next financing. Investors expect this conversation; they do not like being surprised by it.
Get the paperwork right
Each grant needs board approval and a current fair market value for the strike price. In the US that usually means a recent 409A valuation. Do not promise a number in a Slack message before the board has approved it. Tax treatment is specific to the person and the instrument, so the engineer should talk to their own tax adviser.
Frequently asked questions
When should we first talk about a refresh?
Around the two-year mark, or at the first formal review after it. That leaves enough time to plan the pool and board approval without being reactive.
Should every engineer get a refresh?
Not automatically. Tie refreshes to role and performance, and say so. If you give everyone the same top-up, it stops meaning anything.
What if we cannot afford more dilution?
Then decide what else you can offer: a cash raise, a clearer path to a lead role, or a longer exercise window. Silence is the worst option, because the engineer will notice the date before you do.
Is a refresh enough to stop a counteroffer?
Rarely on its own. It works best as part of a plan the engineer helped shape. If you want help building that plan, including levels and equity bands for a small team, book a call or see what Fraction costs.