Skip to content
Playbook
AdvancedHRPeopleOpsFounder

Equity refresh: designing the program that prevents the cliff

Without refresh, every employee hits a year-3 cliff where their new-hire grant runs out and the offer they could get elsewhere makes them leave.

10 min read
On this page
60-Second Summary
  • The 'unrefreshed' attrition cliff hits at year 3 of a 4-year vest. Plan for it.
  • Refresh ties to performance + band position; not flat 'cost of staying'.
  • Target unvested forward equity at all times (e.g. 1.5× annual grant equivalent vested per year forward).
  • Communicate the program so high performers know it exists; don't make it a 'squeaky-wheel' benefit.

Stock-option cliffs were designed for an era of 4-year stints. In 2026 most engineers consider their next move at month 30. If you don't have a refresh program, you're not retaining — you're hoping.

Why refresh

  1. Year 4 vest is 100% complete. Year 3 already feels '75% done'.
  2. Competing offers come with full new 4-year grants; your retention case has 12 months of vest left.
  3. Refresh smooths the cliff: every year, the employee has ~3-4 years of unvested ahead of them.
  4. Without refresh, you'll lose your best people first — they have the most market options.

Designing the program

ElementRecommended
EligibilityAll employees after 1 year
CadenceAnnual, aligned to comp cycle
SizeDifferentiated by performance + band position; bigger for high performers low in band
Vest4 years, monthly, no cliff (stacks on existing grants)
Promo top-upSeparate refresh when promoted to new level

The forward-vest math

Target unvested forward equity
  1. 1
    Step 1
    Compute per employee: unvested shares × current FMV.
  2. 2
    Step 2
    Target = 1.5× annual grant equivalent. If below target, refresh restores the buffer.
  3. 3
    Step 3
    Modulate by performance band: high performer = 100% target, solid = 70%, below = 0-30%.
  4. 4
    Step 4
    Roll up. Total refresh pool should be a planned line item, not a surprise to Finance.

Communicating it

The hidden-program failure

If refresh exists but isn't documented, it becomes a benefit only for people who threaten to leave. That's the worst possible distribution — you reward flight-risk signalling. Publish the criteria, the cadence, and how decisions are made.

Written by Pawan Joshi.Sources cited inline.
First published 16 Jun 2026See site changelog →