LinkedIn's 875 Cuts: Source the GBO PMs Before Microsoft Redeploys Them
LinkedIn just cut 875 people and closed its Graz office. Here is how to source the ex-Recruiter PMs and EU engineers before Microsoft reabsorbs them.
On May 13, 2026, at 7 a.m. Pacific, Daniel Shapero sent the memo. By 8 a.m., roughly 875 LinkedIn employees had a calendar invite for a notification meeting. By the end of the week, the entire Graz, Austria office was being wound down under the polite euphemism "underutilized office space." If you source technical talent for a living, the team that built the tools you use every day just got partially handed back to you.
This is the most meta sourcing event of Q2. The people who designed LinkedIn Recruiter, Talent Insights, and the Hiring Assistant agent stack are now updating the very profiles those products index. The window to reach them before Microsoft redeploys the engineers and before Stripe and Anthropic pick off the senior PMs is roughly two to six weeks.
What actually happened on May 13
Shapero, who took over as CEO last month after four years as COO under Ryan Roslansky, cut about 5% of LinkedIn's 17,500-person global workforce. The cuts hit the Global Business Organization, marketing, engineering, and product. Roslansky was promoted to EVP across LinkedIn and Microsoft Office, which is what created the seat Shapero now occupies and the strategic rewrite that drove the layoffs.
The framing in the memo matters. Shapero said LinkedIn is "scaling back investments in some areas including marketing campaigns, vendor spend, customer events and underutilized office space, so we can focus teams on priorities that have the broadest impact with the highest return on investment." A source told Reuters that LinkedIn was explicitly not citing AI as a reason. This is a reallocation, not a retreat.
That last point is the pitch you need to internalize. These are not ZIRP-era hires getting washed out of an unprofitable business. These are people who shipped products inside a division that just crossed $5B in quarterly revenue for the first time. The Hiring Assistant agentic product alone is at $450M ARR with named customers including AMD and Palo Alto Networks. That is the product Shapero is protecting while he cuts around it.
Why the GBO PM is the asymmetric hire, not the engineer
Here is the contrarian read most founders will miss in week one.
The engineers will move fast. Microsoft will redeploy a meaningful chunk internally because the parent company has been on its own headcount march (roughly 6,000 cut a year ago, another 9,000 in July, voluntary retirement buyouts for the first time in its 51-year history). The senior infra and ML platform people who survive that net will get pulled into Stripe, Anthropic, Databricks, and the usual suspects within two weeks. By June you are competing for them at a 25% premium.
The Global Business Organization product managers are different. These are the people who shipped Recruiter, Talent Insights, Sales Navigator monetization, and the Hiring Assistant agent loop. They sat in front of AMD's TA leadership and Palo Alto Networks' VP of Recruiting and negotiated multi-seat deals. They know the buyer, the objection map, the procurement cycle, and the integration partners. Almost no Series A or B startup can replicate that go-to-market knowledge in under 18 months of internal hiring.
The engineers leave with skills the market has named. The GBO PMs leave with relationships the market has not yet priced.
If you are building anything that sells to TA leaders, HR ops, or enterprise hiring teams, the GBO PM cohort is mispriced for the next 30 days. The challenge is finding them. LinkedIn's own people search is not going to surface a clean cut of "Recruiter PM who left in the May 13 cohort." Titles are noisy. The "Open to Work" badge is opt-in and lagging. And the people most worth talking to are the ones who have not updated their profile yet.
This is exactly the problem Refolk was built for. You can describe the person in plain English ("PM who worked on LinkedIn Recruiter or Talent Insights between 2021 and 2026, based in the Bay Area, Dublin, or Singapore") and get a ranked shortlist that pulls from LinkedIn, GitHub, and the open web rather than waiting on profile updates.
The Graz signal everyone is going to miss
Read the Shapero memo carefully and you will find exactly one public reference to the Graz closure: "underutilized office space." That is it. The entire EU engineering hub gets one phrase, buried mid-sentence.
Graz matters more than the headline number suggests. It is in Styria, a one-hour drive from the Slovenian border, two and a half hours to Vienna, three to Munich, three and a half to Ljubljana. Anyone with a linkedin.com email in Styria is now in play, and the relocation math is unusually friendly: you do not need to move them to hire them. Remote works, Vienna works, Munich works, Ljubljana works.
The Graz cohort also likely includes some of LinkedIn's EU data residency and GDPR engineering work. That is a rare, expensive specialty. Any EU-based company with a sourcing or data product needs to be scouting these profiles this week, not next month.
The local competition is real and named. TU Graz feeds into AVL List, Magna Steyr, Infineon Austria, and AT&S. Those four employers will absorb a meaningful percentage of the cohort if US and remote sourcers move slowly. AVL in particular has been hiring software engineers aggressively for its powertrain simulation stack. If you want the privacy engineers and the platform team before Infineon picks them up, you have roughly four to eight weeks.
What ex-LinkedIn talent actually built (and how to pitch them)
If you are going to reach out cold, know the products. The three surfaces whose teams were partially cut:
LinkedIn Recruiter. The product most of your competitors are paying $10K+ per seat for. The engineers and PMs here built the search ranking, the InMail deliverability stack, the project pipelines, and the integrations with Greenhouse, Lever, Ashby, and Workday. If you sell into TA, these people have already lost the arguments you are about to have.
Talent Insights. Market data, supply and demand reporting, the dashboard TA leaders show their CFO to justify headcount. Whoever shipped this knows what hiring leaders actually look at versus what they say they look at. That is gold for any HR analytics startup.
Hiring Assistant. The $450M ARR agentic product Shapero is protecting. People who worked on this and were cut anyway are the most interesting cohort, because they have agentic-product experience from the company that productized it earliest at scale. AMD and Palo Alto Networks are named reference customers. Anyone on the GTM side here has done enterprise agentic-AI deals that 95% of the market has not done yet.
The pitch writes itself: "You built profitable systems inside a business growing 12% YoY. The cut was about portfolio focus, not your work. We are building [X] and we want the person who shipped [specific product surface] running it here."
The broader Q2 pool this sits inside
LinkedIn's 875 is not happening in isolation. The same week:
Meta is cutting 8,000 on May 20. Cloudflare cut over 1,100 the same week as LinkedIn. Oracle has eliminated up to 30,000 positions, around 20% of its global workforce. Microsoft itself just offered voluntary retirement buyouts for the first time ever.
For a sourcer, this means two things. First, the LinkedIn-alumni cohort is competing for attention with five other large cohorts hitting the market in the same 14-day window. Generic "we are hiring" outreach is going to bounce. Second, the sourcing infrastructure you used in 2023 (Boolean strings, title filters, "Open to Work" sorts) is going to miss the people who matter, because the cohorts overlap and the profiles have not updated yet.
This is the second place Refolk earns its seat. When you describe the cohort in plain English ("ex-LinkedIn GBO PMs who shipped Recruiter or Talent Insights, currently based in the US, not yet at a new company") you skip the profile-update lag entirely. The index reads contribution graphs, conference talks, patent filings, blog posts, and other open-web signals that move before someone changes their LinkedIn headline.
The 14-day playbook
If you want to actually convert this window into hires, here is the compressed version.
Days 1 to 3. Build the cohort definition. Not by title, by product surface and tenure. "Worked on Hiring Assistant 2024 to 2026" is a better filter than "Senior Product Manager at LinkedIn." Use Refolk to pull a ranked list and cross-reference against the Graz geography for the EU subset.
Days 4 to 7. First-touch outreach with the "profitable systems, not displaced" framing. Skip the "sorry to hear" opener. These people are getting 50 of those. Lead with what they shipped and what you want them to ship next.
Days 8 to 14. Schedule the calls. The first wave of acceptances will come from people who are decisive. The second wave (and frankly the better wave for senior roles) will come from people taking two to three weeks to evaluate. Plan for both.
Day 15 onward. You are now competing with Microsoft internal redeploy offers, Stripe, Anthropic, and for the Graz cohort, AVL and Infineon. Your edge erodes from here.
What this tells us about LinkedIn as a sourcing surface
One last thought, because it would be strange not to say it. The 875 cuts and the Graz closure are the second public signal in three months that LinkedIn's product investment is consolidating around Hiring Assistant and agentic monetization rather than around the search, profile, and recruiter-tooling layer that sourcers actually use day-to-day. If you have been feeling like LinkedIn search has gotten worse for finding specific cohorts (the May HN thread on this was brutal), the team that would have fixed it just got smaller.
The pool of ex-LinkedIn engineers hiring well in Q2 includes the people who would have rebuilt that search. Sourcing LinkedIn alumni right now is not just opportunistic, it is a hedge against the platform itself becoming less useful for the work you do.
FAQ
How big is the actually addressable ex-LinkedIn pool from the May 13 cut?
About 875 people total, across Global Business Organization, marketing, engineering, and product, plus the entire Graz, Austria office. Realistically, after Microsoft internal redeploy and self-placement at obvious next-stops like Stripe, Anthropic, and Databricks, the externally sourceable pool is probably 400 to 550 over the next 60 days. The GBO PM subset is the most mispriced.
Why source ex-LinkedIn people specifically versus the Meta or Cloudflare cohort?
Because LinkedIn is growing 12% YoY and just crossed $5B in quarterly revenue. The cut is portfolio reallocation, not company-wide weakness, so the talent skews toward people who shipped profitable systems rather than people who got caught in a financial unwind. Meta's 8,000 and Oracle's 30,000 contain plenty of strong people too, but the signal-to-noise on LinkedIn alumni is higher right now.
What is the deadline before this window closes?
For senior engineers, roughly 14 days before Microsoft internal redeploy and external acqui-hires lock them up. For GBO PMs, four to six weeks because the buyer market for that profile is thinner and slower. For the Graz cohort, four to eight weeks before AVL List, Infineon Austria, Magna Steyr, and AT&S absorb most of them.
How do I find ex-LinkedIn employees who have not updated their profile yet?
Profile-update lag is the central problem. The people most worth reaching are the ones who have not yet flipped "Open to Work" because they are weighing offers internally or from network. The fix is to source by product surface and tenure rather than title, and to cross-reference open-web signals (conference talks, GitHub activity, patent filings, public writing) rather than waiting for the LinkedIn profile to catch up. That is the workflow Refolk runs by default.