2024/2025 Crypto Compensation Report
Important disclaimers located at the end of the report.
Compensation is one of the most common topics we’re asked about by the Dragonfly portfolio (and beyond), and for good reason: Reliable industry data in crypto is scarce.
In 2023, we set out to change that, creating what we believe is the most comprehensive dataset and analysis in the space. Our first report surfaced valuable insights and sparked sharp questions from both portfolio and non-portfolio teams. Those conversations directly shaped the design of the 2024/2025 survey you’re reading now.
This dataset includes responses from 85 crypto companies (up from 49 last year and now including companies outside the Dragonfly portfolio), collected in late 2024 and Q1 2025. For the first time, it also includes roughly 3,400 separate, non-duplicated employee and candidate datapoints. We broadened the scope of our exploration, and where relevant, compared findings to our 2023 report to highlight year-over-year trends.
While the Dragonfly portfolio will have access to more granular slices of data, our aim with this report is to make industry-wide trends clear, accessible, and useful for anyone setting, negotiating, or trying to understand compensation and hiring benchmarks, whether you’re a founder, hiring manager, candidate, recruiter, or industry observer.
Key Highlights
- Most crypto companies were in growth mode, not hypergrowth mode
- Crypto hiring was global from day one. U.S.-only hiring was virtually non-existent
- Europe became the dominant international hub
- Salary and token compensation were down across almost every level and region
- Remote continued to reign, and companies weren’t planning to change that
- Crypto was hard to break into; less than 10% of roles were entry-level
- Engineering was king. Product roles started at senior. Design valued IC over leadership
- U.S. wages became the global standard for engineering leadership
Despite bullish bluechip token prices, a favorable U.S. administration, and overall positive sentiment, crypto hiring in ’25 has been cautious. Early swings included strong January job growth, a February tariff shock, and massive reactive job cuts in March, which left net hiring negative through H1.
Preliminary compensation reviews show no major correlation shifts (expected, since comp data moves more slowly than market conditions).
Compensation Landscape
Overall, we’d call crypto compensation in 2024 and early 2025 a down market, and practices still felt relatively immature compared to traditional sectors.
Salaries and token grants fell across nearly all levels. U.S. roles still led in cash compensation, while international teams narrowed the gap with larger equity and token packages. Equity shifted unevenly, especially with non-technical, non-executive roles. U.S. saw shrinking ranges (compression), while international cases sometimes reached 2–10× U.S. levels.
By stage, the expected pattern held: Early-stage companies offered lower salaries and more equity (often 2×), while later-stage teams reversed the balance. Tokens became less common overall, but remained meaningful in Go-to-Market (GTM), Product, and senior international roles.
Entry-level roles were hit hardest, with steep salary and token cuts, partly offset by higher equity. U.S. entry hires still earned more cash, but international peers typically received 2–3× more equity and larger token awards.
Mid-levels were squeezed, showing limited growth, while seniors fared better with smaller cuts, steadier equity, and tokens becoming increasingly concentrated at the top.
The biggest step-ups came at the Senior IC and Executive levels, creating a barbell effect that was most visible in Product and Engineering.
Compensation Benchmarks
Role-by-role compensation data for Software Engineering, Crypto Engineering, Developer Relations, Product Management, Design, Marketing, and Go-to-Market (i.e., Sales/BD/Partnerships) is below.
Dragonfly portfolio companies get additional splits by size, stage, funding, type, and location.
#Ranges & Benchmarking Tool
Key Takeaways
- International engineering executives out-earned U.S. peers for the first time ($530K–$780K total compensation), driven by token packages approaching ~3%
- For Crypto Engineers, the “certainty vs. upside” tradeoff was clearer than ever. U.S. engineers led on cash and total compensation at nearly every level, while international peers led on equity and token upside at the junior and mid-levels
- Product Management executives posted the highest salary of any role ($390K–$484K) and rivaled or exceeded Engineering in total compensation; international PM equity often ran ~2–10× U.S. levels
- Developer Relations became the most “borderless” function, with near-identical global compensation bands and small deltas at manager/executive tiers (U.S. led equity/total compensation; international led tokens)
- Design leadership wasn’t as valued as senior designers. U.S. Senior ICs (Principal/Senior) outearned Managers and even some Executives in total compensation
- Marketing was geographically split between cash and ownership. U.S. led salary and total compensation, while international equity ran ~3–10× U.S. levels
- Pay gaps in GTM total compensation narrowed at entry and manager levels. Senior ICs topped managers in total comp, and international leaders won on equity
Founder Compensation
In this report, we split company ownership into equity and tokens (versus grouping them together in last year’s report). For this reason, salary analysis is reported year-over-year, while equity and token data reflect 24/25 numbers.
Year over year, founders paid themselves more. Overall, more capital raised correlated with higher salaries and lower equity ownership (as expected).
U.S. founders generally outearned their international peers in terms of salary, equity, and tokens.
| STAGE | US AVG | US MEDIAN | INTL AVG | INTL MEDIAN |
|---|
Key Takeaways
- The average founder salary increased approximately 37% year-over-year, from $144,000 in 2023 to $197,000 in 24/25
- Seed-stage U.S. founders held the highest equity ownership of any group (32%)
- Token ownership was relatively consistent at Seed across geographies (9%), but varied significantly at later stages
- Rare outlier: At Series A, international founders earned higher salaries ($244,000) than U.S. founders ($198,000), while U.S. founders held more tokens (13% vs. 9%) and slightly more equity (20% vs. 19%)
- Rare outlier: At Series B, international founders reported the highest equity ownership (30%), but this was based on limited data points and should be interpreted with caution
Bonus & Variable-Pay Practices
Bonuses remained a selective but meaningful part of compensation.
Usage declined with scale: smaller and earlier-stage teams were most likely to experiment, mid-sized firms adopted selectively, and later-stage and infra-focused companies tapered them off in favor of long-term incentives.
Where they existed, bonuses tied to both company and individual performance (mixed) were the default. Company-only, individual-only, and role/level-based models were less common, and fixed bonuses were rare.
Key Takeaways
- Bonuses peak at the smallest company sizes (1–5 employees) and early stages (Seed/Series A), especially among companies with $5M–$19.9M in funding
- CeFi had the highest adoption at 71%; DeFi sat at 50%, heavily U.S.-driven and with a stronger tilt toward individual performance; Infra / L1 / L2 adoption was much lower (15–30%) and typically mixed
- The U.S. was more bonus-friendly, particularly for mid-sized teams
- Rare outlier: Fixed (non-performance) bonuses were uncommon
Token Compensation Analysis
Year over year, the way teams sized token grants shifted.
Many increasingly separated tokens from equity when calculating grants. At later stages, the norm was to peg grants to a Fair Market Value calculation (defined below), most often using a time-weighted average price (TWAP), although other methods still appeared at the edges.
As it came to vesting, most teams adhered to the familiar four-year schedule with a one-year cliff, although some experimented with hybrid models that blended time-based schedules with milestone-driven unlocks.
Our research on cap-table management showed most companies offered tokens from a dedicated employee token pool (essentially the analog to equity stock option pools). A handful of very large organizations still allocated grants against total supply, but that was increasingly the exception.
Some companies also introduced alternative token reward programs. Token bonuses and refreshers gained traction as ways to reward performance or reinforce retention, while experiments with staking, yield, or swap mechanisms were rare and mostly tied to product-specific use cases.
Adoption Status & Timing
In 24/25, we analyzed more teams with tokens than in 2023.
As you can imagine, very small teams were more uncertain about whether they might launch a token, while larger, better-funded, and later-stage teams were far more likely to have one live.
L1/L2 projects (where tokens are most likely to be native to the protocol) were naturally more likely to have a live token; DeFi projects fell in the middle; more compliance-heavy areas such as CeFi were the most token-averse; and consumer/NFT teams showed uncertainty.
International companies were slightly more likely to be live and less likely to be uncertain than U.S. peers, reflecting the heightened regulatory hesitation in the U.S. during 2024.
Key Takeaways
- The strongest near-term launch prospects were mid-sized (11–100) Series A/B companies with $5M–$40M in funding
- Live tokens scaled with size: 1–5 person teams had none (43% Pre-TGE, 29% undecided, 29% never launch), while >20-person teams were ~45% live
- Funding showed the same trend: <$5M companies were Pre-TGE (43%), and 41% of >40M companies had a live token
- Rare outlier: Gaming teams reported having no live tokens in 2024, down from 25% in 2023
Offer Composition (Cash, Tokens, Equity)
How companies structured their offers shifted with scale.
Very early teams leaned heavily on equity to keep flexibility with tokens. Mid-sized organizations typically transitioned to a blended model (equity + tokens), providing a risk-adjusted offer profile and additional liquidity. Most later-stage firms tilted back toward equity, once company value and token liquidity dynamics were more established.
Key Takeaways
- Nearly half of international teams (44%) offered only tokens. U.S. teams flipped that dynamic, with 39% being equity-only
- Very early companies relied heavily on equity (~71% at 1–5 employees), blended offers peaked at Series A, and things progressed steadily toward equity-only at later stages (Series B at 45%, Series C at 71%)
- Protocol-heavy areas like L1s, L2s, and DeFi naturally leaned into token-forward compensation, while infrastructure, consumer, and CeFi companies defaulted to equity
- Pre-TGE teams were nearly a third equity- and token-only each, and blended at 32%; while teams relied heavily on tokens once launched (47% token-only, 47% blended), leaving equity-only nearly absent.
- Rare outlier: A tiny minority (5%) offered only cash
Equity / Token Relationship (Proportionality)
Only a handful of teams had a proportional relationship between equity and tokens, where companies tied a future token grant to an individual’s equity stake.
It was the norm prior to TGE, but almost all companies severed the link as they matured. By late stage, equity and tokens were treated as entirely separate.
Key Takeaways
- 51% treated tokens and equity as separate compensation elements (no proportional link); by Series D+, this was universal
- Proportional models remained most common at Seed/pre-TGE (33%), but overall, usage declined as companies grew
- L2s were a near-finished case study: 90% had no proportional relationship, and none remained undecided
- Geography had no impact on approach
Token Calculation Methods
Companies use a variety of methods to calculate token allocations for employees, each with its own trade-offs. This report is not meant to be a definitive guide to the “best” methodology, nor does it attempt to cover every possible approach. (For a deeper dive into trade-offs, see this 2022 Dragonfly article.)
With that said, the most common and well-defined calculation methods we see are:
Fair-Market Value (described as “market price” in 2023 report)
Teams with an active token using this method start by determining the total dollar value they want to offer an employee. They then calculate the number of tokens to grant based on the token’s fair-market value (FMV) at the time of calculation, grant, or vesting. They may also use a time-weighted or volume-weighted average price (TWAP and VWAP, respectively).
We’ve seen a very small handful of pre-TGE teams use comparable token prices in the market (e.g., L1s citing other L1 token prices) as the pricing basis in their FMV calculation. A small subset also calculates against the token’s fully diluted value (FDV), using the price set by VCs who’ve secured rights to future tokens; because the VC’s token price is fixed, this can provide a reasonable reference price for compensation until the token becomes tradable.
Teams tend to favor this approach for its simplicity and ease of implementation, especially with rapidly growing headcount. We often remind teams, though, that token values can swing significantly with market volatility, creating uneven grants and outcomes across the token cap table, even for employees with similar levels and responsibilities. In this sense, volatility causes tokens to behave somewhat like public equity but without the same protective measures or stability (given equity’s stable valuation with a 409a).
Teams without a live token typically do not use this method.
Percentage of Tokens
With the percentage-of-tokens approach, employees are granted a fixed percentage of the total token pool, rather than a set dollar value or number of tokens. Companies define compensation bands (similar to equity bands in traditional startups) and map them to token ownership percentages. Each employee then receives a token grant that represents their share of the overall pool, adjusted for role, level, and token-specific nuances.
It can be argued that this method is the only one that explicitly accounts for market volatility while also helping mitigate pay inequity, minimize unnecessary dilution, and preserve employees’ asymmetric upside. However, it’s also the one with the most logistical overhead, and being effective with this approach requires diligent planning to ensure consistency and fairness across the team as headcount grows.
“Other” approaches may include annual grants, purely performance-bonus-based structures, employing a sliding scale between equity and tokens, and essentially just winging it.
Calculation Trends
In 24/25, companies shifted from percentage-based to FMV-based calculations.
Pre-TGE teams favored percentage-based grants due to headcount logistics and uncertainty around timing and valuation, with only a minority using FMV. Once tokens were live, however, nearly two-thirds adopted FMV, driven again by headcount considerations, along with market logistics and clearer historical pricing data.
Within FMV, teams most often used TWAP as the pricing basis (reference price), while earlier teams with evolving liquidity and infrastructure leaned on their token’s current price. Other approaches, such as using VC purchase price or VWAP as the pricing basis, or offering discounted employee pricing, remained niche and were mostly tied to pre-TGE or early-stage contexts.
Key Takeaways
- FMV became the leading approach (47%), up from 31% in 2023 (when percentage-based sizing led, overall, at 38%)
- Pre-TGE teams favored the percentage-based approach (42% vs. 21% FMV), while live-token teams shifted to FMV (63% vs. 22% percentage)
- TWAP was the dominant pricing basis (concentrated in later-stage teams), followed by the current market price at 31% (concentrated in earlier-stage teams)
- L1 leaned FMV at 60%, L2 was split evenly, DeFi was mixed (32% FMV), and infra was mostly non-FMV (20%)
- U.S. companies tilted toward FMV (55%), while international were more balanced (36% FMV vs. 36% percentage)
- Rare outliers: Niche FMV methods such as VC purchase price, discounted employee pricing, and VWAP remained uncommon (<15% of FMV adopters), largely seen in pre-TGE or early-stage contexts
Vesting Schedules & Start-Date Conventions
Pre-TGE companies (especially in DeFi and L1s) often began vesting at TGE to manage retention and circulating supply. Teams with live tokens typically started vesting on an employee’s start date (the most common approach, overall). These practices were largely consistent between U.S. and international teams, with minor differences mainly in pre-TGE groups.
Most companies used standard time-based vesting, usually a four-year plan with a one-year cliff.
A few teams explored vesting tied to performance, or a hybrid of performance and time. Anecdotally, teams were cautious to tie vesting to network-level KPIs (which are hard for individuals to control), instead focusing on auditable product, team, and individual milestones.
A small minority of companies indicated that vesting began after an additional lock-up period following the TGE.
Key Takeaways
- 61% of companies overall started vesting at the employee’s start date
- 59% of pre-TGE teams started vesting at TGE
- Vesting schedules were largely standardized, with 92% using time-based vesting, typically over a 4-year period with a 1-year cliff
- Rare outlier: Only a handful of companies deviated from standard time-based vesting: 6% used hybrid time-and-milestone models, and 2% relied entirely on performance-based schedules
Cap-Table Management
Most companies distributed tokens from a dedicated employee token pool (the token analog to an equity ESOP). Teams managing token pools by total supply appeared more often in larger organizations.
This is fundamentally a cap-table management choice, improving planning and governance, especially for teams using percentage-based grants and forecasting ranges across planned headcount.
Key Takeaways
- 68% of teams managed token compensation via employee token pools, whereas 32% used the total token supply
- Early teams overwhelmingly adopted employee token pools (Seed 82%, Series A 77%), but most moved to total supply as they grew
- Rare outliers: Regardless of company size, DeFi was the largest cohort using employee token pools (85%), while Infra was the largest user of total token supply (57%)
Alternative Token Programs
Roughly a third of companies, whether pre-TGE or live-token, offered or planned to offer alternative forms of token compensation outside standard grants.
Token bonuses and performance incentives were the most common mechanism, and they were often paired with token refreshers that served as “re-ups” after one’s cliff. Mid-sized U.S. teams led adoption with larger organizations starting to participate more cautiously.
Key Takeaways
- More than half of companies that offered refreshers also layered in bonuses, pairing short-term rewards with long-term retention
- Bonuses were commonly linked to performance, whereas refreshers were retention-oriented
- Adoption was strongest among U.S.-based, mid-sized Series A/B companies, particularly in L1 and L2 projects, with DeFi and Infra following
- Mid-sized firms showed the broadest experimentation, larger organizations engaged more conservatively, and almost none adopted the most novel approaches
- Rare outliers: Token swaps or conversions, employee token discounts, and staking/yield programs were all rare, and generally appeared in companies where the functionality was native to the product (e.g., staking in DeFi)
Geography & Remote Work
Hiring Footprint
Global-first was the default, with most teams hiring across borders from day one. Universally, access to talent and flexibility mattered more than geography.
U.S.-only hiring was rare, mostly appearing at Seed and Series A, and faded as companies grew. International-only hiring held steady across stages and sizes, usually reflecting cost-sensitive hiring strategy.
Key Takeaways
- Global was the default from day one, with 81% of companies hiring in both the U.S. and internationally
- Smaller teams often hired globally early. Companies with 21–100 employees and $20M–$40M in funding were the most consistent, even more so than larger companies
- Infrastructure teams were the most likely to hire globally at 80%
- Rare outlier: U.S.-only hiring was virtually non-existent at 6%, a niche strategy led by early consumer and DeFi teams who outgrew it with scale
International Employee Locations
Western Europe was the dominant international hiring hub.
Regional hiring shifts typically occurred at Series B+ as companies matured and required stronger local operations, leading to expansion into Asia, Canada, and Eastern Europe.
Key Takeaways
- Western Europe: 84% of companies between Series B and E (and about just the same amount with $40M+ funding) employed people there
- Eastern Europe: 63% of later-stage companies hired there, drawn by strong engineering pipelines and cost-effective talent
- Asia: Presence nearly doubled year over year (from 20% to 41%) to accommodate stronger adoption
- Canada: 38% of companies between Series B and E expanded into Canada, using its U.S. proximity, favorable regulation, and developer base as a hedge
- South America: Only 13% of companies between Series B and D expanded there
- Rare outlier: India (9%), Africa (4%), and Oceania (2%) remained under-tapped
Regulatory-Driven Hiring Shifts
From 2023 to 2024, regulation was one of the most discussed external factors shaping crypto talent strategy. In the U.S., heightened scrutiny around trading, custody, and protocol activity kept compliance considerations close to hiring decisions.
However, looking back at 2024 and early 2025, regulation had little impact on driving companies to hire outside the U.S. The adjustments that did occur were concentrated in larger, better-funded teams in Infra and DeFi (which are among the most likely to have or launch a token), as well as in CeFi (which is more heavily regulated).
Companies with both U.S. and international teams were the most likely to cite regulatory-driven moves, while U.S.-only companies largely didn’t react, with a small minority considering changes.
Key Takeaways
- Overall, only 14% of companies adjusted their hiring strategy due to regulatory pressure
- Nearly one in four companies with employees in both the U.S. and abroad said regulation directly influenced their decision to expand hiring internationally
- Among companies with only international employees, most were already global for other reasons, but 21% acknowledged that regulation played a role
Remote & Hybrid Work Policies
Crypto remained heavily remote, with most companies being fully distributed. Hybrid models were the next most common, blending remote work with mandated in-office time. Remote-first approaches (office optional) occupied a smaller middle ground, and very few companies operated fully in-office.
Philosophies were strong on both sides of the spectrum, and policies overall proved sticky: Almost all teams planned to maintain their current model.
Key Takeaways
- Over half of all companies were fully remote, over a quarter were hybrid, and remote-first were a smaller minority at 14%
- 94% of companies had no plans to change their policy
- U.S. teams leaned more remote (55%), while international peers leaned more hybrid (35%)
- Series A teams were relatively mixed in approach, but by Series B, remote dominated at 73%
- Rare outlier: Only 2% of companies were fully in-office
Cost of Living Adjustments
Cost-of-living adjustments (CoLA) remained steady, with little change from 2023.
Companies often skipped CoLA early on, with compensation practices still developing and a focus on securing core talent. Adoption was most common in the middle stages to manage budgets, before reverting to non-CoLA at later stages to attract top talent and deploy larger war chests more effectively.
U.S. and international adoption were nearly identical, underscoring that CoLA decisions were driven more by company than by region.
Most companies that adopted CoLA used structured methods, with only a few lacking a defined approach.
Key Takeaways
- Adoption held steady year-over-year (35% in 2023 vs. 38% in 24/25)
- CoLA adoption was rare at very small firms (14–20%), peaked at mid-sized companies (52%), and fell back at the largest firms (33%)
- U.S. and international adoption rates were both just under 40%
- Rare outlier: Public-stage companies universally applied CoLA (though this reflects a very small sample)
Organization & Hiring Trends
Team Composition & Level Mix
Crypto has long been engineering-dominant, and this held true across sizes, stages, and funding tiers from pre-seed through post-Series B. International teams leaned even more toward engineering, while product and marketing leadership was concentrated in the U.S.
In terms of growth over time, Seed teams were mostly engineers; Series A–B added senior PMs, designers, and GTM (with few managers); post‑Series B added structure in GTM and engineering.
Notably, entry-level hiring was scarce, constraining pipelines and diversity, and making it harder for new entrants to break into the industry (with product and marketing particularly tough). Executive hiring outside of engineering was limited.
Key Takeaways
- Engineering (software & crypto) made up about 67% of total headcount across sizes, stages, and funding tiers
- Entry-level roles represented only 10% of all positions
- Product was underbuilt, with more than half of PMs at the principal or executive level
- Marketing was also thin, accounting for only 7% of headcount, with slightly more coverage internationally
- GTM made up 11% of headcount, mostly at mid and senior levels
- Developer Relations was 3% of headcount, was mostly mid-level, and was compensated on par with other functions
- Design was skewed senior and lacked leadership, with about 44% at the principal level and fewer than 10% in manager or executive roles
- Orgs ran lean outside engineering: Marketing/Engineering 1:14 and Product/Engineering 1:13
- Rare outlier: In some international teams, product ratios reached as high as 20 engineers per PM
Headcount Planning Discipline
Headcount planning became more formal as companies scaled, with size the clearest predictor of discipline. Funding and stage followed the same pattern, with Series B serving as the inflection point, and larger raises being tied to more structure.
DeFi/CeFi tended to be more disciplined; AI/Consumer often lagged despite funding; and international companies ran more structured processes than U.S. peers and were less likely to have no process at all.
A counterintuitive wrinkle: nearly all sub-10-person teams lacked formal planning, even though many used percentage-based token compensation that demands headcount diligence. Many organizations stalled between being “somewhat” and “fully” formalized, pointing to a need for early direction.
Key Takeaways
- Only 12% of companies ran fully formalized plans, with most in “somewhat formal” or informal categories
- Series B was the turning point; >$40M raised companies were 2.4× more likely to plan than <$5M peers, but stalled at “somewhat” until Series D
- International companies were more structured than U.S. peers (62% vs. 49%)
- Rare outliers: We saw unexpected pockets of full headcount formalization in “Other” stage orgs (foundations, DAOs, bootstrapped) and small L2 teams
Hiring Momentum by Function
Most teams were in targeted growth mode, not hypergrowth. Overall, companies were selectively scaling core teams while keeping support lean, with many intentionally maintaining flat headcount and very few planning to reduce them overall.
Hiring momentum followed team composition as outlined above. Engineering was the clear priority, while Product and Marketing grew steadily but cautiously. GTM hiring gained momentum in scaling phases and in infrastructure and financial sectors, but slowed in consumer, gaming, and NFT companies. Design remained underbuilt, and Ops and HR/Recruiting stayed mostly flat, creating potential bottlenecks as technical teams grew faster than internal support.
Key Takeaways
- Engineering led growth, with about 78% of teams expanding and only 1% cutting. It was also the only function with a real entry-level pipeline
- GTM hiring rose with stage growth, with 57% increasing and 40% flat
- Product and Marketing were steady, with roughly half increasing and half flat
- Few teams scaled across the board; most expanded 2–4 functions while holding others flat, likely to streamline recruiting
- GTM scaled only when senior and exec talent was already in place, and companies rarely expanded GTM with entry-level roles
- Flat momentum often came with senior-heavy teams; once senior and exec layers were in place in Product, Design, or Engineering, those functions were more likely to hold steady and slow hiring
- Rare outlier: Cuts were rare at 3%, slightly higher in Design, Ops, and HR / Recruiting
Recruiting Funnel Metrics
Source of Hire
Candidates mainly came from proactive sourcing, followed by referrals, inbound applications, and agencies.
The smallest startups (1–5 employees) relied equally on referrals and inbound. By 11–20 employees, early-team referrals tapered and sourcing overtook inbound, where it would remain the primary channel into later company stages. As companies grew, referrals increased again as teams expanded their internal networks.
Agencies were always a minority, peaking in mid-sized, mid-stage firms and fading at both the smallest and largest companies.
Key Takeaways
- Seed companies leaned on sourcing (43%), Series A on referrals (32%), Series B balanced referrals (41%) and inbound (32%), and Public companies split evenly between sourcing and referrals (50/50)
- International companies leaned more heavily on inbound than the U.S. (35% and 19%, respectively)
- Rare outlier: Agency hires peaked in mid-stage firms (~20%) but were nearly absent in the smallest and largest companies
Time to Hire & Interview Length
Hiring speed tracked company maturity: early-stage startups moved fastest; mid-stage teams varied during rapid growth and while formalizing structure; and larger, later-stage companies slowed as processes became more robust.
Speed alone didn’t guarantee success. Agency-only and some teams with no recruiter (very often early and founder-led) moved fastest but struggled to close candidates, while later-stage companies and in-house teams offered the best balance of speed and conversion.
Interview length reflected the same trade-off between speed and effectiveness: no-recruiter/founder-led teams ran the shortest processes, in-house teams added more structured steps that improved results, and agency-led searches dragged without the same internal access.
›Recruiting model breakdownExpand
Key Takeaways
- Average hiring took 3.8 weeks, with about 4 interviews
- Early-stage teams hired fastest at 3.3–3.6 weeks (~4 interviews). Mid-stage averaged 3.8–4.3 weeks (4–5 interviews). Larger orgs slowed to 4.5–6.3 weeks with 5–6 interviews
- Agency-only was fastest (3 weeks) but converted poorly (50%)
- No-recruiter/founder-led averaged 3.5 weeks with a 64% accept rate
- In-house recruiters took 3.9 weeks, yet achieved the best acceptance (73%)
- Rare outlier: Founders of small teams were fast with referrals, specifically, often running sub-2-week closes, but took significantly longer with other channels
Pass-through Rates
Many teams expect their funnel to tighten as interviews progress, starting lenient and ending more selective. In practice, pass-through rates remained fairly steady across stages, so the funnel didn’t consistently narrow as teams may have assumed. Relatedly, being tough early didn’t guarantee stronger results later in the funnel.
Smaller startups were less efficient when they and/or their processes lacked clarity or credibility. Larger, later-stage teams achieved stronger pass-through rates, especially from the final round to offer, thanks to well-defined practices and brand recognition.
Adding nuance to what we previously wrote on source of hire/candidate channels, sourced candidates experienced the highest pass-through rates, followed by those who came in via inbound channels, referrals, and then agencies (in that order).
| STAGE | PRESCREEN → 1ST | 1ST → FINAL | FINAL → OFFER |
|---|---|---|---|
| In-house | 42% | 41% | 58% |
| Hybrid | 44% | 44% | 41% |
| No Recruiter | 35% | 33% | 44% |
| Agency | 48% | 38% | 23% |
| Average | 41% | 40% | 51% |
In-house recruiting teams ran the strongest process end-to-end. Agencies were the weakest, generating volume early but breaking down later in the funnel. Hybrid setups performed better than having an external agency, and founder-led teams (without a recruiter) were more stringent at the top of the funnel and saw stronger conversions in later stages. Still, neither were as effective as in-house teams overall.
#Pass-through Rate Calculator
Key Takeaways
- Overall average pass-through rates: 41% prescreen, 40% first→final, 51% final→offer
- Larger teams (>100 employees) had higher pass-through at every stage (46% → 49% → 59%)
- Smaller teams (<20 employees) showed lower pass-through (34–36% prescreen, 28–36% first→final, 34–46% final→offer)
- Longer hiring cycles (6–8 weeks) showed the highest final-to-offer pass-through at 57%
- Sourcing was the strongest channel, with the highest pass-through at each stage (47% → 46% → 53%)
- Geography was neutral: U.S. (41% → 40% → 49%) and international (41% → 40% → 53%) companies performed almost identically
Offer Accept Rates
Offer outcomes were highly polarized: while a meaningful share of companies closed nearly every offer, another segment struggled to convert at all, with few in the middle.
Conversion advantages tracked with size, stage, and capital: later-stage, better-funded teams generally received more offers, even when their processes took longer, while the smallest, least-funded startups moved quickly but often lost candidates.
L1, L2, and AI teams converted talent more effectively, while Infra and NFTs teams lagged. Geography made little difference, with U.S. and international companies seeing similar outcomes.
Key Takeaways
- 40% of companies closed near 90%, 20% sat closer to 20%, and a few landed in the middle
- About two-thirds of offers were accepted on average, overall
- Smaller teams of 1–5 employees averaged 54% acceptance, while those with over 100 closed above 83%
- Companies with <$5M in funding closed under 50%, compared with 73% for those >$40M
- Acceptance lagged at Seed (~52%), peaked at Series A (~80%), dipped at Series B (71%), and rebounded to 80–90% in later stages
- L1s, L2s, and AI led at around 80%, while Infra (~57%) and NFTs (~50%) lagged
- Rare outlier: A few very small teams closed below 30% despite competitive pay, reflecting candidate risk appetite alignment, or perception and stability concerns
Offer Decline Reasons
Declines were primarily driven by offers that fell below expectations and stronger competing offers (compensation, overall).
Earlier teams lost more on compensation, mid-stage teams struggled with role clarity and process, and later-stage teams faced challenges with culture and/or role fit.
Key Takeaways
- Compensation drove 83% of declines, with 44% from below-expectation offers, and 39% from stronger competing offers
- U.S. teams lost more on below-expectation offers (52%), while international teams lost more on competing offers (48%) and culture/clarity
- Rare outlier: Only 17% of declines were tied to role, culture, process, remote preference, or location
Attrition & Retention Drivers
Attrition was moderate overall but became more volatile as companies scaled. Smaller and international teams generally retained talent better, supported by a stronger connection to company mission. Most departures stemmed from role misalignment or employees being poached.
At Series A and beyond, growing pains surfaced around culture and lacking growth opportunities. As companies further matured and secured greater funding, cultural shifts became more pronounced, and departures were increasingly tied to fit rather than compensation.
Compensation-related exits were less frequent overall, but when they occurred, they drove much higher attrition, showing pay issues had an outsized impact.
Key takeaways
- Average attrition was 7.8%, with volatility increasing alongside scale, funding, and stage maturity
- Attrition spiked at 21–50 FTEs (4% → 9.2%), ramped up at Series A (8.5%), and climbed with scale, hitting 10.6% at 100+ employees and double digits beyond $40M raised
- The strongest retention came from smaller teams (11–20 employees at 3.5% attrition), international companies (6.0% vs. 8.9% U.S.), and sub-$5M startups (2.5%)
- Job mismatches and external poaching (candidates receiving higher-paying offers) accounted for >40% of attrition
- Hires closed in 2–6 weeks show 7% attrition vs 12% when cycles exceed 6 weeks
- Rare outlier: L1s (15% attrition) carry extra pressure, potentially given token exposure and capital intensity
Recruiting Challenges
The supply of crypto talent continues to trail demand.
Everyone faces top-of-the-funnel problems (finding qualified candidates), but small teams feel it most. The issues broaden with mid-sized and later-stage companies to include more around compensation mismatches, slower processes, and competitive pressure. Once companies surpass $40M raised or 100 employees, compensation concerns nearly equal top-of-the-funnel problems, making employer brand, offer design, and narrative clarity increasingly critical.
Key takeaways
- Finding qualified candidates was the top constraint across stage, size, funding, sector, and geography
- Beyond $40M raised or 100+ employees, pay mismatches became the second-biggest hurdle
- Slow processes were most common in >100 FTE companies, highlighting the need for streamlining before scale
- Rare outlier: Infrastructure companies report the broadest spread of challenges: 63% cite talent scarcity, 27% compensation, 5% competition, and another 5% remote/global hiring (the only sector to mention this)
Talent Org Models
Most companies relied on in-house recruiters, introducing agencies to augment pipeline when needed.
Just as with pass-through rates, recruiting team structure had a direct impact on acceptance rates. In-house teams were the strongest and most consistent. Hybrid models (in-house recruiter with external agency support) provided flexibility but didn’t always improve results. Founder-led hiring (no recruiter) proved workable early on, but became difficult to sustain as companies grew. Agency-only approaches were rare and typically the weakest setup, carrying risks without strong internal ownership.
Key Takeaways
- Most companies relied on in-house recruiting (60%) and had a recruiter by Series A
- In-house teams delivered the strongest acceptance rates (73%), outperforming founder-led (no recruiter) teams (64%), hybrid models (59%), and agency-only setups (50%). Founder-led approaches proved workable early but unsustainable at scale
- Hybrid models grew with maturity, reaching 33% at Series B/C companies with >$40M
- Small, underfunded teams (<$5M) mostly relied on founder-led or ad-hoc hiring (60%, no recruiter)
- Agency-only recruiting appeared only at Seed and Series A, and all later-stage companies had in-house teams
- Rare outlier: CeFi companies were the most agency-reliant, with 17% using external agencies only
Appendices
Demographics and Methodology
Roles: "Crypto engineering" refers to protocol/blockchain engineers; "Go‑to‑Market (GTM)" includes Sales, Partnerships, and Business Development; Developer Relations and Marketing are separate from last year’s GTM
Compensation: Total compensation is self‑reported by respondents and not derived as Salary + Equity + Tokens; for GTM roles, variable compensation includes a dollar amount and a percentage of salary
Methodology: Companies selected preset salary, equity, token, variable compensation, and total compensation bands by role and level; we report average lower and upper bounds; values combine a broader dataset of roughly 3,400 separate, non-duplicated employee and candidate datapoints using weighted averages; founder compensation uses averages/medians from free‑form responses
Founder Ownership: Percentages reflect total founder ownership (not per‑founder)
International: Companies headquartered outside the U.S
Timing: “24/25” = late 2024 + Q1 ’25; approximately 60% of companies and 30% of candidates submitted data in Q1 ’25
Type Categories: "Other" includes foundations, DAOs, and bootstrapped projects
Percentages: Values are normalized to 100% after filtering; rounding may shift categories by ±1 pp
Visualizations: Unless noted, visuals reflect 24/25 data and averages
Low Sample: Less than five responses; indicated in tooltips (hover on desktop; tap on mobile)
Special Thanks
First and foremost, a major special thanks to Felix France of FCC / Blockcomp for his contributions to survey design, data collection, and analysis.
We’re also grateful for the thoughtful feedback and support from friends and colleagues across the ecosystem, including CJ Wilson; Richard Rodairos; Haseeb Qureshi; Lindsay Lin; Tom Schmidt; Phillip Bodine; Matt Moore; Jess Furr; Bryan Edelman; Casey Taylor; Rob Hadick; Annica Benning; and Dragonfly founders, recruiters, and hiring managers.
Thanks, too, to our friends at Missing Link, Pyxis, Lawrence Harvey, and Blueprint.
Lastly, thank you to our readers and the broader community for their patience while we built this report, and everyone who responded and provided data.
Terms of Use & Disclosures
THIS REPORT IS FOR INFORMATIONAL PURPOSES ONLY.
Any content included in this report is for informational purposes only and not an offer, solicitation, or recommendation to buy or sell any securities or pursue any particular investment strategy, and does not constitute the provision of investment, business, or legal advice. Dragonfly Digital Management, LLC (together with its affiliates, "Dragonfly") does not provide advice of any kind through this report, its website, or social media posts. Information in this report or any posts may not be used or relied upon in evaluating the merits of any investment.
This report should not be considered financial, investment, tax or legal advice. You should consult your own advisers as to business, financial, tax, legal, or other related matters concerning any investment.
DISCLAIMER: THERE ARE NO WARRANTIES, EXPRESSED OR IMPLIED, AS TO ACCURACY, COMPLETENESS, OR RESULTS OBTAINED FROM ANY INFORMATION PROVIDED HEREIN OR ON THE MATERIAL PROVIDED. Certain information contained herein (including forward-looking statements, blockchain data, economic and market information) has been obtained from published sources and/or prepared by third parties and in certain cases has not been updated through the date hereof. While such sources are believed to be reliable, neither Dragonfly, its respective affiliates and employees, or our information providers warrant or guarantee the timeliness, accuracy or completeness of this information, and shall not be liable for any errors or inaccuracies, regardless of cause, or the lack of timeliness of, or for any delay or interruption in the transmission thereof to the user. The information set forth herein does not purport to be timely or complete and is subject to change without notice.
Dragonfly makes no representations about the enduring accuracy of this information, or its appropriateness for a given scenario.
Review our Terms of Use and Disclosures for important details about Dragonfly.