Average Severance Package by Industry (2024–2025 Data)

The gap between the best and worst severance deals is larger than most people realize. Here's where your industry sits — and what that means for your negotiation.

If you asked ten people what a "normal" severance package looks like, you'd get ten different answers — and most of them would be wrong for your situation. That's because the range isn't just wide, it's industry-specific in ways that make cross-sector comparisons nearly meaningless. A twelve-week severance in retail is exceptional. In media, it's below average. Understanding where your industry sits is the first step to knowing whether your offer deserves a signature or a counter.

The data in this post comes primarily from the Challenger, Gray & Christmas 2025 Severance & Salary Benchmarking Report (8,000+ respondents across approximately 840 companies, using 2024 plan data), cross-referenced with the LHH 2024 Severance & Separation Benchmark Study — the largest ongoing primary research efforts on U.S. severance practices. Where figures differ, the more conservative estimate is used.

The Numbers: Industry-by-Industry Averages

The first thing to notice: the top of this table isn't where most people work. Chemicals and media are generous outliers; manufacturing, tech, and retail represent a much larger share of layoff volume. The widely reported tech layoff numbers from 2023–2026 happened against a backdrop of packages that average just 11.2 weeks — lower than healthcare, finance, or aerospace, despite the media attention suggesting otherwise.

The second thing to notice: tech's 11.2 week average obscures a bifurcation. Mega-cap companies (Meta, Google, Microsoft) have published formulas that produce 20–30+ weeks for long-tenured employees. Smaller tech companies — Series B startups, mid-market SaaS — are closer to 4–8 weeks, pulling the average down. Which side of that divide your offer falls on matters a lot.

What Changed Between 2023 and 2024

The headline finding from Challenger's 2025 report is striking: average severance rose 24% from 2023 to 2024, jumping from 15.6 to 19.3 weeks across all industries. LHH's longitudinal data shows an even longer trend — a 72% cumulative increase since 2020.

That's not a blip. It reflects several intersecting forces.

Formalization of severance policies. In 2017, approximately 30% of companies had no written severance policy — meaning employees were at the full discretion of management. By 2024, that figure had dropped to 3–4%. Written policies tend to produce more consistent — and generally more generous — outcomes than ad hoc decisions.

Tightening of executive accountability. Institutional investors and proxy advisors pushed hard throughout the early 2020s for documented, board-approved severance frameworks at public companies. That pressure trickled down to broader employee populations.

Labor market dynamics post-2020. The tight labor market of 2021–2022 created expectations that companies maintained even as hiring slowed. Employees who watched peers receive generous packages during voluntary exits in 2022–2023 had reference points when evaluating involuntary separation offers in 2024.

One caveat worth repeating: the self-reported nature of these surveys means the data skews toward companies that gave severance at all. Hourly workers, employees terminated for cause, and small employers that don't participate in compensation surveys are systematically underrepresented. Treat industry averages as competitive benchmarks for formally employed salaried workers — not universal baselines.

How the Big Tech Companies Actually Compare

Because tech layoffs have been so visible since 2022, the specific formulas that major companies have disclosed publicly are worth examining directly.

Meta uses 16 weeks of base pay plus two weeks per year of service, with no cap. A seven-year Meta employee receives 30 weeks. That's the most generous published formula among mega-cap tech companies.

Google matches the 16-week base, adds two weeks per year, and layers in 16 weeks of equity vesting acceleration. Total financial impact for a long-tenured Googler is substantial — though Google has increasingly shifted toward Voluntary Exit Programs rather than involuntary RIFs, which changes the math.

Microsoft used a 12 weeks plus two weeks per year formula in its May 2025 round — slightly below Meta and Google but still solidly above the tech industry average.

Amazon uses a one-to-two weeks per year formula with a cap of 20–26 weeks, depending on level. For long-tenured employees, that cap compresses the benefit significantly compared to peers.

Salesforce provides approximately five months minimum, which is competitive even against Meta and Google in the short-tenure range. Stripe's formula starts at 14 weeks minimum with tenure additions.

How Tenure Changes Everything

The industry averages above cover all tenures blended together. In practice, how long you've been at the company affects your benchmark more than almost any other single variable.

At 1–3 years of service, a typical individual contributor should expect 4–8 weeks with one to three months of COBRA and basic outplacement support. Below four weeks with no benefits extension is below market across every industry except retail.

At 3–5 years, the typical range is 6–15 weeks. Offers of 20–26 weeks exist at the generous end — the Meta and Google formulas produce around 22–26 weeks for five-year employees.

At 5–10 years, the benchmark moves to 12–26 weeks with three to six months of COBRA coverage. Below 12 weeks for a six-plus-year employee signals a formula that's below standard.

At 10 or more years, typical offers run 20–40 weeks, with generous packages reaching 40–60+ weeks at uncapped hybrid formulas. Six months of COBRA is standard; twelve months is common at this level.

Level adjustments: These ranges apply to individual contributors. Add 25–50% for managers, and multiply by 2–4x for executives. A VP with 8 years of service who receives the same offer as an engineer with 8 years of service at the same company has a very different negotiating position.

A Note on What the Data Doesn't Capture

Three things the industry benchmarks won't tell you:

Benefits quality. Two packages can both read "12 weeks" on paper and differ by $15,000 in actual value based on whether COBRA is paid directly (non-taxable) or as a cash stipend (taxable), and whether outplacement is a meaningful program or a login to a career website.

Equity treatment. If you have unvested stock, it's not in this table. The financial impact of whether your employer vests you through the next quarterly date, accelerates six months, or forfeits everything on day one can dwarf the cash severance amount entirely. No industry benchmark covers this.

Clawback and restrictive covenant scope. A 20-week offer with a broad two-year non-compete that prevents you from working in your field is materially less valuable than a 15-week offer with no restrictions. The headline weeks figure tells you nothing about what you're giving up.

How to Use This Data in a Negotiation

Industry benchmarks are most useful as a starting point for a written counter offer. "My offer is X weeks; the 2024 industry average for [sector] is Y weeks for someone with my tenure" is a factual, defensible statement that moves the conversation away from emotion and toward market data — which is where employers are most comfortable negotiating.

A few things that make benchmark data land better in a real conversation. First, be specific about tenure. An industry average that includes two-year and fifteen-year employees is less useful than a tenure-adjusted benchmark. Second, don't lead with the most generous outlier (Meta's formula) as your baseline — it will read as unrealistic and undermine your credibility. Lead with the industry average; note the generous end as context. Third, focus the counter on the items where the gap is clearest. If your cash is close to benchmark but your COBRA is thin, that's where to focus energy.

The goal isn't to get to the top of the range. It's to close a meaningful gap with data, in writing, before you sign something you can't unsign.