Orrick Legal Ninja Snapshots
19 minute read / 33 minute listen | October.01.2026
Prefer to listen?
Enjoy this text-to-speech recording of this article:
You've designed your equity plan, allocated awards, and communicated the opportunity to your team. But have you given the same attention to the cliff – the provision that says: "You don't get anything until you've been here long enough to prove you're serious"?
Frequently, founders and HR teams treat the cliff as boilerplate – a standard twelve-month waiting period that simply "comes with the ESOP." But in the wake of the landmark ruling of the German Federal Labour Court (Bundesarbeitsgericht – "BAG") of March 2025 and the ongoing talent wars in AI, deep tech, and other hotbeds, cliff design has quietly become one of the more strategically important – and often underestimated – levers in your equity toolkit.
In this Legal Ninja Snapshot, we'll decode the cliff from multiple angles: what it is, why it exists, the different flavors it comes in, how it can serve as a strategic hiring tool, and how cliff cultures differ between PE-backed and VC-backed companies. As always, we'll draw on practical data and real-world examples.
Data Sources
This Legal Ninja Snapshot draws in particular on two key sources: our own OLNS#8 – ESOPs, VSOPs & Co.: Employee Ownership for German Start-ups (2025-2026 Edition); and the Carta PE Executive Equity Report 2026 ("PE Report").
This Legal Ninja Snapshot
In this Legal Ninja Snapshot, we explore the often-overlooked cliff provision and why it deserves more attention than most founders give it. Here is what we cover:
…and Much More in Our OLNS#8
For comprehensive background information on employment participation programs, including detailed guidance on structuring them in their various forms under German law, we refer you to our OLNS#8 – ESOPs, VSOPs & Co.: Employee Ownership for German Start-ups.
For simplicity, this Legal Ninja Snapshot uses the term "ESOP" for all varieties of virtual or equity-based programs and the term "Awards" holistically to encompass the various forms of virtual/phantom equity, equity-based interests, options, and profit participation rights (PPRs) issued under an ESOP.
Let's dive in.
Let's start with the basics. A cliff is a minimum period an employee must remain with the company before any Awards vest or begin to vest. During the cliff period, 0 % vests. Once the cliff expires, the portion that would have vested during that time vests all at once. Think of it as a single unlock rather than a gradual accrual.
The market standard for ESOPs in German start-ups is a twelve-month cliff within a 48-month vesting period with linear monthly vesting thereafter. In plain English: for the first twelve months, the employee vests nothing. On the first anniversary, 25 % of the total grant vests at once. After that, an additional 1/48th (approximately 2.08 %) vests each month until the full grant is vested after four years.
According to Carta's PE Report, 96–98 % of all cliffs at both PE-backed and VC-backed companies are set at exactly twelve months. This uniformity isn't accidental – it reflects decades of market practice, investor expectations, and the kind of mathematical elegance that makes lawyers feel like they understand finance: 25 % vesting on the first anniversary of a four-year schedule.
But as the saying goes: "Tradition is just peer pressure from dead people." The fact that this structure made sense in the past doesn't mean it's the right fit for every company today.
The now-ubiquitous "four-year vesting with a one-year cliff" traces its lineage to the Silicon Valley venture capital ecosystem of the 1980s and early 1990s. Back then, early VC firms aimed to keep founders and early employees motivated for the crucial early years – long enough to weather product pivots, market entry, and scaling, but not so long as to tie talent down excessively.
But why four years? Venture investors in the 1980s typically expected holding periods of around five years from first meaningful equity investment to exit (IPO or acquisition) – and that is significantly less than what investors have to expect today when exit horizons are often north of eight years. But back then, four years seemed like a good compromise: long enough to cover the formative "value creation" window, but short enough to allow refresher grants or new incentive cycles. Once a few major law firms, VC funds, and company consultants adopted four years, it became a market norm.
And the one-year cliff? It was designed to protect against early departures. Founding teams and early-stage investors wanted a minimum evaluation period before anyone walked away with equity. If a hire didn't work out within the first year, no harm done – the Awards returned cleanly to the pool.
The cliff serves three core purposes:
In summary: the cliff is the simplest and most widely used mechanism for ensuring that equity ends up in the hands of those who actually build the business.
If the standard twelve-month cliff were the only game in town, this would be a very short Snapshot. But like most things in equity compensation, the devil is in the details – and the variations are more interesting (and more strategically significant) than most people think.
Several German companies have adopted 18- to 24-month cliffs for senior technical roles, particularly in AI and blockchain, where the argument often goes something like this: "In deep tech, where it takes 18 months just to understand the product architecture, a twelve-month cliff barely covers the learning curve." This may seem counterintuitive given that the Carta data shows only 70.2 % of management grants at PE-backed companies and 54.6 % at VC-backed companies include a cliff at all – but those figures reflect the negotiating power of senior hires, not the optimal incentive structure from the company's perspective.
And then there's Westwing. The e-commerce company, in its IPO prospectus, disclosed a 36-month cliff for its 2016 long-term incentive scheme targeting senior management. That's three full years before a single share vests – an unusually aggressive cliff that likely reflects the board's desire to lock in key executives through the critical post-IPO transition period.
The psychological impact of extended cliffs should not be underestimated. An employee approaching a 24-month cliff faces a much more significant financial decision when considering departure than one near a twelve-month cliff. If you've already invested 20 months and know that a substantial vesting event is just four months away, the "cost of leaving" calculation tilts heavily toward staying. Which, of course, is exactly the point.
German employment law's emphasis on proportionality means that cliff periods must be reasonable. ESOP provisions – especially for programs at the level of the employing entity – qualify as standard business terms (Allgemeine Geschäftsbedingungen – "AGB") and are therefore subject to the AGB control regime under secs. 307 et seq. BGB.
Many legal practitioners consider cliff periods of up to 24 months permissible, provided the overall vesting structure is proportionate and the cliff does not effectively deprive the employee of the value of their work contribution. However, there is no decisive BAG case law on the maximum permissible cliff length – meaning companies pushing beyond 24 months are operating in legal grey territory. The Westwing 36-month cliff, for instance, likely benefited from the fact that it targeted senior management, where courts may apply a more relaxed standard.
At the opposite end of the spectrum, there is a small but notable trend away from cliffs entirely. Sequoia's data from 647 companies shows that cliff usage for executive new hires dropped from approximately 85 % in 2023 to 78 % in 2026, and for non-executives from approximately 88 % to 81 % over the same period.
Maybe the most dramatic example is OpenAI, which according to its public announcements eliminated all vesting cliffs for new employees in late 2025, allowing shares to begin vesting immediately upon hire. In the context of the AI talent war – where top machine learning researchers can command seven-figure packages and switch employers at will – removing the cliff became a competitive necessity. The message: "We trust you from day one."
But let's be clear: what works for OpenAI in the overheated AI talent market doesn't necessarily translate to every company. Unless you're also sitting on billions in funding and competing for the dozen people who actually understand transformer architectures, the cliff remains a sensible feature of any ESOP.
After the BAG's landmark ruling in March 2025 eliminated the ability to forfeit vested Awards upon a voluntary departure, some companies have attempted a workaround: so-called "negative vesting". While the BAG has indicated that negative vesting may be permissible under certain conditions, the precise boundaries remain unsettled. More importantly, the negative signaling and detrimental incentive effects should not be underestimated, as we will see.
Here's how it works: the ESOP provides that upon the employee becoming a voluntary leaver, all vested Awards will either forfeit (i) gradually over a certain period of time or (ii) all at once after the expiration of a certain period of time following the employee's departure (occasionally referred to as "negative cliffs"). The key legal constraint is that the negative vesting must not occur faster than the original vesting – i.e., the forfeiture rate must be congruent with the vesting rate. So, for a four-year vesting schedule, the post-departure forfeiture period must also be at least four years.
The logic is that this doesn't violate the BAG's ruling because nothing is "taken away" immediately upon departure. Moreover, the BAG has recognized that a departing employee's prior contributions have a diminishing impact on subsequent exit proceeds – the longer the period between departure and exit, the less the employee's past work can be said to have influenced the outcome. Instead, the Awards slowly erode over time if the employee has left and no exit has materialized.
Here are two critical caveats: first, there are aspects of negative vesting schemes that remain untested in German courts. And second – foremost – while, while on its face the conceptual argument has some logic, this concept can make the ESOP significantly less attractive as it runs counter to an intuition that one of the co-author's five-year old son instinctively grasps "mine is mine – taking it is crime".
If Section II explored what cliffs are and presented some variations, this Section addresses what they can do. In the post-BAG-ruling world, cliff design has become more important.
As mentioned above, the BAG ruling leads to the unenforceability of the immediate forfeiture of already vested Awards upon a voluntary resignation. In plain English: you can no longer keep employees around by making them fear they'll lose their equity if they leave prior to the expiration of the vesting period. The "stick" is gone – you can only use the "carrot."
For many founders, this felt like losing a safety net. For others (and we count ourselves in this camp), it's actually a good thing – it forces companies to design ESOPs that genuinely motivate and reward employees and shifts the focus towards smart implementation, rather than relying on fear-based retention. But it also means that the tools you do have at your disposal need to be used more thoughtfully. And that should include cliff designs.
Here's the crucial distinction: an extended cliff delays vesting commencement, ensuring a minimum stay period before any ownership transfers. But since Awards haven't yet vested during the cliff, nothing is "taken away" from the employee – distinguishing cliffs from the treatment of a voluntary leaver as a bad leaver that the BAG struck down.
Consider the practical difference:
The psychological framing matters enormously. "You haven't earned it yet" is a very different message from "We're taking back what you earned."
Smart companies are combining cliff design with strategic initial grant sizing. Rather than making one large initial grant (the "set it and forget it" approach we cautioned against in our Legal Ninja Snapshot on Start-ups Staying Private Longer), companies can make a more conservative initial grant and plan for refresher/top-up grants after the cliff expires.
This hedges a specific risk: the scenario where a company allocates a generous equity package to a senior hire, only to see them leave shortly after the cliff expires – walking away with 25 % (or more) of a premium-sized grant. By starting smaller and topping up based on demonstrated performance, companies preserve pool flexibility while still offering competitive packages.
The cliff doesn't operate in isolation. The most effective retention strategies combine cliff design with complementary mechanisms:
Here's the flip side (because of course there's a flip side): extended cliffs can create barriers to talent acquisition. Shocking, we know – turns out that telling a candidate "you get nothing for two years" isn't universally beloved. Mid-level professionals with strong market positions may be unwilling to accept a 24-month cliff, particularly if they're foregoing the chance to keep vesting equity at their current employer. In a competitive hiring market, an aggressive cliff signals either strong conviction in long-term retention or a certain lack of trust. Spoiler alert: not every candidate will charitably assume the former.
The trade-off is real: a longer cliff provides better protection against early departures but narrows the talent pool (…we hear and understand the frustrated "but we need to hire fast"…). The right choice depends on the role, the market, and the company's competitive position. For a machine learning engineer in the current AI talent war, a 24-month cliff might be a deal-breaker. For a senior executive with a track record and a stake in the company's long-term success, it might be perfectly reasonable.
As always, context is king. Calibrate accordingly.
So far, we've treated the cliff as if all companies approached it the same way. They don't. One of the most interesting findings from Carta's PE Report is the significant divergence in cliff practices between PE-backed and VC-backed companies. The data covers grants starting vesting between 2022 and 2025, providing a robust window into current (US) market practice that holds learnings for the German ecosystem as well.
According to the PE Report, at PE-backed corporations, 87.1 % of employee grants have a cliff, and 70.2 % of management team grants have a cliff. When a cliff exists, 96.7 % (employees) and 96.6 % (management) are set at exactly one year, vesting 25 % of their allocation.
The most common executive vesting schedule at PE-backed corporations is four years, one-year cliff, monthly vesting – accounting for 42.8 % of all executive grants. The next most common: four years, one-year cliff, quarterly vesting (16.6 %). No-cliff structures come in at just 9.0 %.
For PE investors, consistency isn't just a preference – it's a governing principle. They apply the cliff broadly and uniformly across employee levels, treating equity compensation as a standardized component of the compensation architecture rather than a case-by-case negotiation.
VC-backed companies tell a different story. While 89.3 % of employee grants still have a cliff (slightly higher than PE), only 54.6 % of executive grants include one. That's a striking 15.6 percentage-point gap compared to PE (70.2 % vs. 54.6 %).
What explains this? A detailed analysis is beyond the scope of this Legal Ninja Snapshot, but the following considerations offer an initial explanatory approach. VC-backed companies are typically founder-led and negotiate executive packages individually. A late co-founder bringing a pre-existing relationship and deep industry expertise may negotiate away the cliff entirely. A hired CEO joining a late-stage start-up might demand immediate vesting to compensate for the equity they're leaving behind. In the VC world, the cliff is a negotiation variable; in the PE world, it's more like a policy.
Table 1: Cliff Prevalence – PE vs. VC Corporations
| PE-Backed Corporations | VC-Backed Corporations | |
|---|---|---|
| Employee grants with cliff | 87.1% | 89.3% |
| Management grants with cliff | 70.2% | 54.6% |
| When cliff exists: one-year cliff | 96.6–96.7% | 96.1–98.2% |
| When cliff exists: vests 25% | Yes (standard) | Yes (standard) |
Source: Carta PE Executive Equity Report 2026.
Table 2: Most Common Executive Vesting Schedules
| PE-Backed Corporations | VC-Backed Corporations | |
|---|---|---|
| four years, one-year cliff, monthly | 42.8% | 44.0% |
| four years, one-year cliff, quarterly | 16.6% | – |
| four years, no cliff, monthly | 6.4% | 19.8% |
| Immediate vesting, no cliff | 9.0% | 8.3% |
Source: Carta PE Executive Equity Report 2026.
Perhaps the starkest difference between PE and VC equity cultures lies in performance conditions. Approximately 61 % of initial management grants at PE-backed LLCs are tied to performance metrics – typically MOIC (Multiple on Invested Capital) or IRR (Internal Rate of Return). VC-backed companies, by contrast, remain overwhelmingly time-based, with performance vesting largely confined to executive-level arrangements.
At the risk of oversimplifying matters (again…), this difference reflects the fundamental nature of each investor type. PE firms are financial engineers who buy, optimize, and sell – if a metric can be measured and tied to compensation, they'll find a way to do it. VC firms are growth investors who bet on people and markets – and trust that four years of vesting will sort out the rest.
For German start-ups caught between these two worlds – or transitioning from VC darling to PE portfolio company – knowing who you're designing for matters.
So here's the bottom line: the cliff is not boilerplate. It's not a formality you copy-paste from something you found online or that your seed investor sends you. It's a strategic lever – and one that deserves some consideration when it comes to its interplay with other retention levers.
Here's our practical checklist for getting your cliff right:
In this context, please also refer to the following OLNS Snapshots: