Pool top-ups and refresher grants: keeping the pool from running dry

The pool is not a thing you create once. It is a balance you spend down, and the two questions that matter are how much to add, and when.

The number you must be able to state at any moment

  pool reserved
− options outstanding (granted, not yet exercised)
− shares already issued on exercise
+ forfeitures returned
──────────────────────────────
= unallocated pool

Using the company from part 2:

Item	Shares
Pool reserved	1,500,000
Options outstanding	600,000
Exercised into shares	40,000
Forfeitures returned	50,000
Unallocated	910,000

Note that exercised options leave the pool permanently — they have become shares. If you track only “granted”, your balance drifts wrong every time someone exercises.

Sizing a top-up from a plan

The same discipline as part 3: count the roles, not a percentage.

Next 12 months	Grant	Shares (of ~11.6M FD)
VP Product	1.0%	116,000
4 × engineers	0.25% each	116,000
2 × GTM	0.35% each	81,200
Refreshers, existing team	—	150,000
Buffer (~15%)	—	70,000
Required		533,200

With 910,000 unallocated, this plan is already covered — so at the next round the honest answer to “we want a 10% pool” is a hiring plan showing you need considerably less. That is a negotiation you can win with a spreadsheet, and it is worth real percentage points.

Refreshers: the problem they solve

An initial grant vests over four years and then stops (part 4). An employee in year four is earning no new equity while a new hire starts a fresh grant at a higher strike. Do nothing and your most experienced people quietly become your least incentivised.

Common approaches:

Annual refreshers. A smaller grant each year from year two or three, vesting over four years with no cliff. Smooths the cliff-edge at year four and makes equity a recurring part of compensation.

Evergreen top-up. Each year, grant enough to restore a full four-year forward vesting position. Predictable for employees, heavier on the pool.

Promotion and milestone grants. Tied to role change rather than the calendar. Less administrative load, but leaves long-tenured people in stable roles with nothing.

There is no universal right answer; the wrong answer is having no policy, because refreshers then become whoever complains loudest.

Grant size: use a framework, not instinct

Work from a target percentage by level and stage, converted to shares at the current fully diluted count, with the strike set by the latest 409A. A grant expressed as a percentage and converted to shares once, at grant, is defensible. A grant expressed as “about the same as the last engineer got” is how you end up with two people doing the same job holding wildly different equity.

Benchmarking data by role, stage and geography exists precisely because these ranges are wide and moving.

Do not over-reserve

Unallocated pool shares dilute existing holders and benefit no one until granted, and at the next round investors count the leftover pool as existing dilution — then ask for a top-up on top of it. Reserving 15% when the plan needs 6% costs you percentage points twice.

Housekeeping that prevents surprises

Process leavers promptly. Unvested options should be cancelled and returned on the termination date, not when someone remembers. An un-cancelled grant makes your unallocated balance a lie.

Track the exercise window. Vested options expire — commonly 90 days after termination. Expired options return to the pool; former employees should be told clearly before the window closes.

Watch authorised share headroom. A top-up that exceeds authorised-but-unissued shares needs a charter amendment, which needs approvals, which takes weeks (part 2).

Reconcile quarterly. Board-approved grants against grant letters against the cap table. Fifteen minutes, and it catches errors while they are still cheap to fix.

Next

Part 11 covers the two US tax mechanics that decide how much of an equity win an early employee or founder actually keeps: the 83(b) election and early exercise.