Every founder budgets for the equity in an offer letter. Almost none of us budget for what happens two or three years later, when that same employee needs a second grant to keep believing the remaining shares are worth waiting for. I didn't either, until I sat down and actually modeled three years of refresh cycles against my cap table and found the real number was bigger, and more predictable, than I expected.
The cost nobody prices at the offer stage
When you hire someone, you price their grant once: a number of shares tied to a role, a level, and a fully diluted share count on that day. Then the company keeps raising, the team keeps growing, and two years later that same employee is roughly half vested with a shrinking incentive to stay. Most founders treat the refresh that follows as a surprise. It isn't. It's a predictable, modelable cost that belongs in your option pool planning from the start, the same way you'd model a second wave of hiring.
The math on a single refresh
Here's what one refresh actually looks like in shares and points, not just a percentage of an original grant. Say your company has 12 million fully diluted shares. A senior engineer hired at 0.4 percent gets 48,000 shares. Carta's compensation benchmarking puts a typical refresh at 25 to 50 percent of what a new hire in that role would get today, so at year two you issue a refresh of roughly 14,400 shares, 30 percent of the original grant. Against a 12 million share base, that single refresh dilutes everyone else by about 0.12 percentage points. Trivial in isolation. The problem is nobody issues just one.
Where the real number lives: the whole team, not one hire
Model a 15 to 20 person team where every employee past the 18-month mark gets an annual refresh at 25 to 30 percent of their original grant. That's the range the data backs, and it also lines up with what founders running this cycle for the first time report: one to two additional percentage points of dilution per year, compounding as the team grows and more employees cross the refresh threshold each cycle. Over three years, without a single new hire's own future refresh factored in, you're looking at three to six points of cumulative dilution that a pool sized only for new hires was never built to absorb.
Refresh cost versus replacement cost, the comparison that actually matters
The number that makes this easier to defend to a board isn't the dilution in isolation, it's the dilution against the alternative. Replacing a senior engineer who leaves because their equity ran dry typically costs 50 to 150 percent of their annual salary once you count recruiting fees, ramp time, and the productivity gap while the role sits open. A $180,000 engineer walking out the door is a $120,000 to $270,000 problem in hard and soft costs. The refresh that would have kept them, 14,400 shares in the example above, costs you a fraction of a percentage point of dilution and no cash. Run both numbers side by side before a board meeting and the refresh stops looking like a cost and starts looking like the cheaper option.
Budgeting it into your next option pool, not your next panic
This is the part founders skip because it happens at the wrong point in the fundraising process. Carta's pool data shows pre-seed to Series A companies allocate roughly 35 to 37 percent of their total option pool to refreshes rather than new hires, and that share climbs to 40 to 50 percent by Series B and beyond as more of the team crosses the refresh threshold each year. If you're negotiating a 15 percent pre-money option pool and haven't earmarked a third of it for refreshes, your real new-hire budget is closer to 10 percent than 15, and you'll find that out the hard way mid-cycle when the pool runs dry right as you need to move fast on retention.
The one-week move
Pull your cap table and list every employee who will cross 18 months of tenure in the next twelve months. For each one, take 25 to 30 percent of their original grant size as a placeholder refresh number, sum it across the list, and express the total as a percentage of your fully diluted share count. That single number, not a vague request to plan for equity refreshes, is what you bring to your board or your next term sheet negotiation. It turns a soft retention ask into a hard planning input, which is the only version of this conversation that gets a pool sized correctly the first time.
Frequently asked questions
How much dilution do equity refresh grants cost per year?
For a team where every employee past 18 months gets an annual refresh at 25 to 30 percent of their original grant, plan on roughly one to two percentage points of additional dilution per year at the seed to Series A stage, compounding as more employees cross the threshold.
Should refresh grants come out of a separate option pool?
No. They almost always draw from your existing pool rather than a new authorization, which is why pool sizing at your last round needs to account for refreshes and not just new hires.
How do I model refresh costs before a fundraise?
List every employee crossing 18 months of tenure in the next year, apply 25 to 30 percent of their original grant size, and sum the total as a percentage of your fully diluted share count. Bring that number, not an estimate, into your pool sizing conversation.
Is a refresh grant cheaper than replacing the employee?
Usually by a wide margin. Replacing a departed employee typically costs 50 to 150 percent of their annual salary in recruiting and ramp costs, versus a fraction of a percentage point of dilution for the refresh that would have retained them.
A refresh policy doesn't need a board debate every cycle once the model exists. Build the dilution number once, update it each round, and you'll never again be negotiating an option pool that's already too small before the ink is dry.