Block's 4,000 Cut Wraps in Q2: 3 Diasporas, 3 Different Playbooks

Block announced its 40%+ workforce reduction on February 26, 2026, and per the Q1 10-Q the execution was "substantially complete by the end of the second quarter of fiscal 2026." That means the wave isn't coming. It's here, hitting the market this quarter, and most sourcers are still treating "ex-Block engineer" as one searchable label.

It isn't. The 4,000+ people coming out of Block split into three operationally distinct populations with different stacks, different cities, and different realistic acquirers. If you send the same outreach to a Cash App KMP engineer in St. Louis and a Square seller backend engineer in Atlanta, both will ignore you, and Stripe will hire them next week.

The actual cut, in numbers

The SEC 8-K filed February 26, 2026 lays it out cleanly. Block is reducing its workforce by more than 40%, from roughly 10,000 to "just under 6,000." The company guided $450M to $500M in restructuring charges. The Q1 10-Q came in at $495.3 million, near the top of the range, with execution wrapped by end of Q2.

4,000+

The stock jumped 25% on the announcement, which tells you investors read this as structural. There is no reversal coming. The people who are out are out.

One more number that matters for understanding who is in the cohort: between 2019 and 2022, Block's headcount more than tripled from 3,900 to 12,500. The cut is overwhelmingly a correction of pandemic-era over-hiring, not an AI restructuring. Dorsey's own internal memo conceded that during the pandemic he "incorrectly built two separate company structures for Square and Cash App," creating duplicated roles. The press framing of "AI pivot" is partial cover for the over-hire correction underneath.

That distinction matters for sourcing. The cohort skews 2 to 5 YOE, hired 2020 to 2022, mid-level rather than staff. If you go in looking for principal engineers displaced by AI, you will find some, but you will miss the actual shape of the available pool.

Diaspora #1: Cash App (Kotlin, KMP, St. Louis and SF)

Cash App is the crown jewel of this cohort and the one most sourcers will get wrong.

The org pioneered Kotlin Multiplatform in production. Per JetBrains' own case study, Cash App grew from 4 mobile engineers at launch in 2013 to 50 across iOS and Android serving 30 million MAUs, and gradually transitioned from shared JavaScript to KMP starting in 2018. The backend stack from public job specs: MySQL, DynamoDB, Aurora, Elasticache, S3, RDS, AWS, Kubernetes, Terraform, Go, and Kotlin.

The pedigree signal isn't a job title. It's open source. SQLDelight, Wire, Moshi, Misk, Okio, and Retrofit all originated inside Square and Cash App. If a candidate has commits to any of those repos, or shows up in the Kotlin Slack #cash-app or #multiplatform channels, or spoke at Droidcon, you are looking at a real Cash App engineer, not someone who was adjacent to the brand.

Geography is the second filter. Cash App is St. Louis-anchored with a strong SF presence. St. Louis is not a city most sourcing stacks index well, and that's the arbitrage. A LinkedIn boolean for "Cash App" + "Kotlin" + "San Francisco" misses half the org.

This is exactly the kind of query where a literal-string search tool gives up. Asking Refolk for "Kotlin Multiplatform engineers who worked on Cash App, including St. Louis, with commits to SQLDelight or Wire" returns a different and better list than the LinkedIn boolean equivalent, because the pedigree signal lives in GitHub, not in a title field.

Realistic acquirers: Stripe, Mercury, Ramp, Brex, Robinhood, Chime, Affirm. Klarna will quietly try to pick up Afterpay-adjacent people, which is its own irony given the acquisition history. If you're hiring against any of those names, you have maybe four weeks before the strongest profiles are gone.

Diaspora #2: Square seller (Ruby, Java, iOS, SF and Atlanta)

Square seller is the merchant and POS side, and it is a different engineering culture from Cash App. Ruby and Java heavy on the backend, iOS-heavy on mobile, with commerce and payments infra experience that translates directly to merchant-acquirer competitors.

Atlanta is the underappreciated city here. Refolk's index shows the US Kotlin engineer population concentrates in NY, the Bay, and Atlanta, and Atlanta is where Square seller built out a meaningful second hub. Most sourcing playbooks default to SF and miss the entire Atlanta cohort, which is both cheaper to hire and less picked over.

Dorsey's "two separate company structures" admission is also a sourcing tell here. Duplicated functions across Square and Cash App (platform infra, data platform, security, internal tools, design systems, recruiting eng) were almost certainly cut harder than product-line teams. Those platform people are stronger on average than the product engineers and more available, because nobody's recruiting them with a product-team narrative.

Target platform and infra titles from Block, not product engineers. The product engineers will get scooped fast by direct competitors. The platform engineers are mispriced and undertargeted.

Realistic acquirers: Adyen, Toast, Shopify, Lightspeed, Clover (Fiserv), SpotOn. Stripe will take both Cash App and Square seller people, but the Square seller cohort has a more natural fit at Adyen and Toast.

Diaspora #3: Bitkey and Proto (and a correction on TBD)

First, the correction. TBD, Block's decentralized-internet developer platform, was wound down in February 2025, not 2026. That diaspora has been in the market for over a year. Crypto-native shops have been picking through it since last spring. If anyone tells you the 2026 cut releases "TBD engineers," they're a year late.

What the 2026 cut likely does release, in thinner numbers, is more from the Proto team, which owns Bitkey and Block's bitcoin mining hardware initiative. Proto is the smallest of the three diasporas and by far the most specialized.

The Bitkey stack is rare. Block open-sourced Subzero, its HSM-based custodial solution, in 2020, and it secures Cash App customer balances and Block's own bitcoin treasury. Bitkey customer server keys live in AWS Nitro Enclaves with multi-engineer deployer-group approval and hardware-token auth. That is hardware wallet plus secure-enclave plus multisig cryptography experience, and globally it's a sub-100-person market.

Proto also runs bitcoin mining hardware. Chip design and FPGA talent inside Block did not exist anywhere else in the company, and exists at maybe a dozen companies worldwide.

Hardware wallet plus AWS Nitro Enclave plus multisig cryptography is a sub-100-person market globally. Move first or don't bother.

If even 20 to 40 Proto engineers were released, the inbound will be saturating within days. Realistic acquirers: Casa, Unchained Capital, Foundation Devices, Ledger, Trezor, Strike, River Financial, Fold, Coinbase Custody, Anchorage Digital. The Bitcoin Optech contributor list and the proto-at-block GitHub org are your two best starting points. Title search will fail you here because half these people don't have "Bitkey" on their LinkedIn at all.

Why "ex-Block engineer" is now a useless label

The three diasporas have almost no overlap in what they want to work on next. A Cash App Kotlin engineer applying to Casa would be a strange fit. A Bitkey cryptographer applying to Toast would be insulting. A Square seller Ruby engineer applying to River Financial would be a category error.

But every one of those moves is happening right now, because the sourcing market keeps treating Block as one company. It hasn't been one company operationally for years, and Dorsey said so out loud.

The practical implication: split your search. Build three separate sourcing lists, three separate outreach templates, three separate target-acquirer maps. The Cash App list keys on KMP, Kotlin OSS commits, and St. Louis. The Square seller list keys on Ruby/Java, payments infra, and Atlanta. The Bitkey list keys on Subzero, Nitro Enclaves, and Bitcoin Optech.

This is the moment where a plain-English sourcing layer earns its keep. Asking Refolk for "former Cash App mobile engineers with KMP experience, not in San Francisco" returns a clean list that a LinkedIn boolean can't, because the signal you actually care about (KMP plus St. Louis plus a Cash App org affiliation) isn't a structured field anywhere.

The four-week clock

Execution was substantially complete by end of Q2 FY2026, which means severance is paying out now, equity has crystallized, and the strongest profiles are taking calls this month. The over-hire correction means most of the cohort is mid-level, in-demand, and will land within 60 days.

Stripe and Adyen know this. Casa and Unchained know this. The question is whether your team is sending three different messages to three different lists, or one generic "we saw the news at Block" message that gets archived.

FAQ

When exactly did Block announce the layoffs?

February 26, 2026, via an 8-K. The plan was to reduce headcount by more than 40%, from roughly 10,000 to just under 6,000. Execution was "substantially complete by the end of the second quarter of fiscal 2026," which is why the diaspora is hitting the market in earnest right now rather than in March.

Is the 2026 cut an AI pivot or an over-hire correction?

Both, but mostly the latter. Dorsey publicly admitted that during the pandemic he built two parallel company structures for Square and Cash App, creating duplicated roles. Headcount more than tripled from 3,900 in 2019 to 12,500 in 2022. The AI framing is partially accurate but partially press cover. For sourcing purposes, assume the cohort skews 2 to 5 YOE mid-level, not displaced staff engineers.

Were TBD engineers part of the 2026 cut?

No. TBD was wound down in February 2025, a full year earlier. That diaspora has been on the market since last spring and crypto-native shops have already worked through it. The 2026 cut likely includes some additional Proto and Bitkey reductions, but TBD as an org was already gone.

What's the single highest-value sub-pool inside the cut?

Bitkey and Proto cryptography and embedded engineers, by a wide margin. The combination of HSM work on Subzero, AWS Nitro Enclaves, and multisig key management is a sub-100-person global talent pool. If you're a self-custody or bitcoin infrastructure company, source there first, this week, and ignore everything else until you've made offers.