← Back to insights

Lumen · Workforce Portrait

Meta

68,897 employees · +1% YoY · 12,016 hires last year · 16% attrition

One workforce, told in views. Each one built off a real piece of dataviz that deserves studying.

Five Slides on Meta

The same chart, told five ways. Click through to see what the "+1% YoY" press line actually hides.

59kMay'25JunJulAugSepOctNovDecJan'26FebMarAprMay

Source: LinkedIn TI · per-function monthly data · 2026-05-14

Title Race · 13 Months in 12 Seconds

The 10 biggest job titles at Meta, racing month by month over the last year. Bars overtake each other in real time. Watch ML Engineer climb past Production Engineer past Data Engineer past Engineering Manager.

MonthMay 2025
#1 titleSoftware Engineer
Top 10 total33,148
Biggest moverSoftware Engineer
Change vs May'25+0.0%
019k
1Software Engineer
18,591
2Sr Software Engineer
2,594
3Data Scientist
1,860
4Research Scientist
1,788
5Product Manager
1,752
6ML Engineer
1,588
7Engineering Manager
1,571
8Data Engineer
1,225
9Staff SW Engineer
1,212
10Production Engineer
967
May 2025
May 2026

Source: LinkedIn TI · per-title monthly headcount · 1.1 sec/month · 2026-05-14

How the Functions Actually Moved

Real 13-month trajectories for each of Meta's top 10 functions. Engineering grew +2,193 while HR collapsed -33%and Marketing/Arts cut hard. The "+1% YoY" total hides a radical reshuffle underneath.

59k0Engineering+7.2% · 32,711IT+4.7% · 4,816Operations+0.2% · 4,033Program/Proj Mgmt-6.2% · 3,488Product Mgmt-1.3% · 2,645Biz Dev-0.7% · 2,568Marketing-9.3% · 2,376Arts/Design-8.7% · 2,361Sales-1.3% · 1,460HR-33.1% · 1,313May'25JunJulAugSepOctNovDecJan'26FebMarAprMayMar: Arts cut -140

Source: LinkedIn TI · captured per-function via filter API · 13 months, all 10 top functions · 2026-05-14

The Skill Stack

Top 10 skills on Meta employee profiles. Bubbles pulse to that month's real count. Machine Learning keeps climbing through the year while older languages plateau.

MonthMay 2025
Skill-mentions164.1k
Hires (mentions)+5,032
Departures2,439
Net since May'25+0
ProgrammingData & AlgosML / AI
Python26.7kJava21.2kC++19.5kSQL19.1kAlgorithms15.8kJavaScript14.9kData Analysis14.2kMachine Learning12.7kSoftware Development13.2kC#6.7k
May 2025
May 2026

Source: LinkedIn TI · monthly skill counts via per-skill filter · 2026-05-14

05

Skill Cloud · 50 Skills, Sized + Colored

inspired by · d3-cloud · word cloud layout
d3-cloud · word cloud layout

Skill Cloud · The 50 Skills That Define Meta

Size = headcount with that skill, color = growth direction. Purple is fastest-growing (AI/ML cluster). Red is shrinking. The cloud is built with d3-cloud — collision packed with rotation.

Skills50
Hot growers (10%+)23
Shrinking2
LargestPython · 29.1k
FastestKubernetes +18%
15%+ growth 10–15% 5–10% 0–5% Shrinking

Source: LinkedIn TI · top 50 skills · 2026-05-14

Meta's Title Map

Top 10 titles, each a bubble. Sizes pulse to that month's real headcount. Watch ML Engineer and Production Engineer keep climbing while management roles ease back.

MonthMay 2025
Headcount (top 10)33,148
Hires this month+941
Departures448
Net since May'25+0
EngineeringML / ResearchDataProduct / Mgmt
SWE18,591Sr SWE2,594Research Sci1,788Data Sci1,860ML Eng1,588Eng Mgr1,571PM1,752Staff SWE1,212Data Eng1,225Prod Eng967
May 2025
May 2026

Source: LinkedIn TI · monthly headcount per title via filter · 2026-05-14

Meta on the Map

Top 10 metros plotted on a real geographic projection. Circle area = headcount, outline = growth direction. Greater Augusta, Bengaluru, and Taipei are the fastest-growing metros, marked in yellow.

Top metroSF Bay
Headcount22,667
Share of company33%
Fastest growing+24%
inAugusta Area
SF Bay22.7kSeattle Area9.3kNYC7.6kLondon4.7kLA1.9kAustin1.7kDublin1.5kSingapore1.4kDC Area1.1kBoston0.7kAugusta Area +24%Bengaluru Area +22%Taipei–Keelung Metropolitan area +17%

Source: LinkedIn TI · world TopoJSON via world-atlas · 2026-05-14

08
Flourish · extruded map

The Extruded Map · Headcount as 3D Bars

The world as a slowly rotating plane, with a 3D bar rising from each Meta metro. Bar height = headcount, color = growth. Drag to rotate, scroll to zoom. SF towers over everything; East Coast and London hold the line.

Tallest barSF Bay
SF headcount22,667
Hovered
Count
YoY

Source: LinkedIn TI · top 10 metros · world TopoJSON · 2026-05-14

Meta's World, In Three Dimensions

The top 10 metros as pins on a slowly rotating globe. Pin height = headcount, color = growth direction. Drag to rotate; scroll to zoom; hover for details. Powered by three.js + react-three-fiber.

Metros plotted10
Continents3
Largest pinSF Bay · 22.7k
Growing6 metros
Shrinking3 metros

Source: LinkedIn TI · top 10 metros · drag to rotate · scroll to zoom · 2026-05-14

Meta's World, By Headcount

Top 10 metros, drawn as a voronoi treemap sized by Meta headcount. Color = region. Cells with green outlines are growing, red are shrinking — width of the outline scales with the growth rate.

AmericasEMEAAPACgrowingshrinking
SF Bay22.7k 5% YoYSeattle Area9.3k 5% YoYNYC7.6k 4% YoYLondon4.7k 4% YoYLA1.9k 4% YoYAustin1.7k 6% YoYDublin1.5k 11% YoYSingapore1.4kDC Area1.1kBoston0.7k 2% YoY

Source: LinkedIn TI · top 10 metros · 2026-05-14

The Monthly Pulse

12 months of real flow into and out of Meta. Hires rise above the line, departures drop below. The yellow ribbon overlay is total headcount — peaked at 71,014 in Oct 2025, fell since.

71,014
peak headcount · Oct 2025
68,897 today (-2,117)
71,014 PEAKHIRESDEPARTURESMay'25JunJulAugSepOctNovDecJan'26FebMarAprMay

Source: LinkedIn TI · monthly chart values via Highcharts API · May 2026 is partial-month.

12

Daily Pulse · A Year in Squares

inspired by · Nivo calendar heatmap
Nivo calendar heatmap

Daily Pulse · A Year in Squares

The 12 months of hires, departures, and net change laid out as a calendar heatmap. Each square is a day, color is intensity. The heavy hiring tail of May–Sep 2025 and the Dec/Mar departure spikes pop right out.

ViewHires
12mo total+12,059
Peak month+1,728
Lightest month110

Source: LinkedIn TI monthly totals · daily distribution approximated (real data is monthly).

The Meta Talent Sankey

Last 12 months. Left: who Meta hires from. Right: where Meta's leavers go. Green flows in, red flows out. Watch the OpenAI strand: Meta's biggest single leak.

Amazon536Google432Microsoft355TikTok275Amazon Web Services (AWS)225Apple180Meta2,003Google335OpenAI229Microsoft125Apple118Amazon100LinkedIn65

Net +876 talent in 12mo · source: LinkedIn TI · 2026-05-14

13.5
linkedinideas.jamesraybould.me/talent-flows

Talent Flows · Four Views, Three Windows

The 24 biggest talent-flow relationships at Meta. Toggle between four visual modes and three time horizons. Watch how Google flips from +513 (24mo) to −69 (6mo) — relationship deteriorating in real time.

WindowLast 12 months
ModeOrbit
Net won (top 12)+1,605
Net lost (top 12)-616
Biggest singleAmazon (+436)

Companies orbit Meta. Closer to center = more talent moved. Color = direction.

Meta68,897Amazon+436Microsoft+230TikTok+221OpenAI-193AWS+190Rivos+160Anthropic-100Google+97Apple+62NVIDIA-57MSFT AI-53Facebook+52Intel+51Reddit-45Accenture+37Capital One+35Deloitte+34Roblox-30Netflix-26Stealth-26Waymo-26xAI-20Databricks-20Stealth AI-20

Source: LinkedIn TI · top 24 flow companies per window · 2026-05-14

13.7
slope chart · NYT-style

How the Flows Are Trending

Each line is one company. The Y axis is net talent flow with Meta. As we move right, the time window narrows (24mo → 12mo → 6mo). Lines that cross zero are relationships that flipped direction recently. Watch Google fall from a +513 net winner to a −69 net loser over the past 24 months.

Companies tracked32
Flipped relationships1
Biggest 24mo winnerAmazon
Biggest 6mo loserOpenAI
024 months12 months6 months(long-term)(annual)(recent)Rivos +159Amazon +140Amazon +826TikTok +72TikTok +440AWS +375Microsoft +517Anthropic -63Google -69Google +513OpenAI -131OpenAI -267

Source: LinkedIn TI · 6mo + 12mo + 24mo flow windows · 2026-05-14

14

Talent Flow Network · Force-directed

inspired by · Nivo network · force-directed
Nivo network · force-directed

Talent Flow Network · Force-directed

Each node is a company that exchanged talent with Meta over the last 12 months. Size = volume of flow. Color: green = Meta net won, red = Meta net lost. Closer to the center = stronger flow. Drag a node to explore.

Top inflowAmazon
Net won+1,412
Top outflowOpenAI
Net lost-193
Net total+876

Source: LinkedIn TI · 12-month flow companies · drag/hover to explore · 2026-05-14

Which Skills Are Rising at Meta?

Same 10 skills, ranked two ways. Left: by headcount today. Right: by 1-year growth rate. Big upward jumps = where the hiring engine is pointed.

Rank by countRank by 1y growth1. Python3. Python 22. Java6. Java 43. C++4. C++ 14. SQL7. SQL 35. Algorithms2. Algorithms 36. C5. C 17. JavaScript8. JavaScript 18. Data Analysis9. Data Analysis 19. Machine Learning1. Machine Learning 810. Software Development10. Software Development·

Source: LinkedIn TI · 10 skills, two rankings · 2026-05-14

Same Skills, Different Stories

Toggle ranking metric. Watch how the order rearranges itself — Python sits #1 by count, but ML/NumPy rise when you sort by growth.

  1. 1Python29,085count 29,085growth +9%hires 6,313
  2. 2Java22,766count 22,766growth +7%hires 4,305
  3. 3C++20,956count 20,956growth +8%hires 4,141
  4. 4SQL20,532count 20,532growth +7%hires 4,046
  5. 5Algorithms17,449count 17,449growth +10%hires 3,827
  6. 6C16,049count 16,049growth +8%hires 3,104
  7. 7JavaScript15,930count 15,930growth +7%hires 3,104
  8. 8Data Analysis14,787count 14,787growth +4%hires 2,942
  9. 9Machine Learning14,142count 14,142growth +11%hires 3,536
  10. 10Software Development13,643count 13,643growth +3%hires 2,240

Source: LinkedIn TI · 2026-05-14

Race to Headcount

10 metros, each a sweep around Meta. Sweep length = share of that metro's headcount. Color = 12-month growth.

San Francisco Bay Area22,667 · +5%Seattle Area9,314 · +5%New York City7,595 · +4%London4,663 · +4%Los Angeles1,894 · -4%Austin1,714 · -6%Dublin1,509 · +11%Singapore1,413 · +0%Washington DC-Baltimore Area1,088 · +0%Boston691 · -2%Meta68,897 total+1% YoY

Source: LinkedIn Talent Insights · 2026-05-14

When Each Function Peaked

Every function has a peak month. Sorted by share of headcount lost since peak. HR peaked back in May 2025— the month our window opens — and has fallen ever since. Engineering didn't peak until Feb 2026, riding out 8 of the 12 months of total-company decline.

May'25JunJulAugSepOctNovDecJan'26FebMarAprMayHRpeak 1,964 · now 1,313-33.1%-651 since peakMarketingpeak 2,660 · now 2,376-10.7%-284 since peakArts/Designpeak 2,616 · now 2,361-9.7%-255 since peakSalespeak 1,569 · now 1,460-6.9%-109 since peakProgram/Proj Mgmtpeak 3,723 · now 3,488-6.3%-235 since peakBiz Devpeak 2,644 · now 2,568-2.9%-76 since peakProduct Mgmtpeak 2,722 · now 2,645-2.8%-77 since peakOperationspeak 4,123 · now 4,033-2.2%-90 since peakITpeak 4,894 · now 4,816-1.6%-78 since peakEngineeringpeak 33,016 · now 32,711-0.9%-305 since peak

Source: LinkedIn TI · per-function monthly headcount filtered for each of the top 10 functions · 2026-05-14

The Top 10 Title Trajectories

Same 13-month window for each of Meta's top 10 titles. ML Engineers and Production Engineers are still climbing; most others peaked between Oct 2025 and Feb 2026 and have eased back since.

ML Engineer1,927 today
peak
+21.3% YoYstill climbing
Production Engineer1,131 today
peak
+17.0% YoYstill climbing
Staff SW Engineer1,244 today
peak
+2.6% YoYstill climbing
Data Scientist2,092 today
peak
+12.5% YoY-0.6% from peak
Data Engineer1,361 today
peak
+11.1% YoY-1.4% from peak
Research Scientist1,965 today
peak
+9.9% YoY-3.6% from peak
Software Engineer19,781 today
peak
+6.4% YoY-0.9% from peak
Sr Software Engineer2,760 today
peak
+6.4% YoY-1.4% from peak
Engineering Manager1,636 today
peak
+4.1% YoY-2.5% from peak
Product Manager1,744 today
peak
-0.5% YoY-2.7% from peak

Source: LinkedIn TI · Job Title filter applied per title · 2026-05-14

Where Meta Actually Bled

Each function's last 12 months: departures on the left, hires on the right. Width = absolute people. Sorted by net change so the biggest bleeders sit at the top. HR lost 651 net (33% of its body) and barely replaced any.

DEPARTURES (12mo)HIRES (12mo)HR1,9641,31375377-651net changeMarketing2,6192,376700441-243net changeProgram/Proj Mgmt3,7183,488575370-230net changeArts/Design2,5862,361558307-225net changeProduct Mgmt2,6792,645371297-34net changeSales1,4791,460404415-19net changeBiz Dev2,5852,568366276-17net changeOperations4,0264,033717702+7net changeIT4,6014,8167991,024+215net changeEngineering30,51832,7113,5556,089+2,193net change

Source: LinkedIn TI · per-function 12-month hires + departures via filter · 2026-05-14

Top Universities by Meta Headcount

Undergraduate alumni working at Meta today, with 1-year growth

  1. 1
    University of California, Berkeley+7% YoY · 382 hires last year
    1,946
  2. 2
    Carnegie Mellon University+5% YoY · 325 hires last year
    1,570
  3. 3
    Stanford University+2% YoY · 266 hires last year
    1,522
  4. 4
    Georgia Institute of Technology+12% YoY · 348 hires last year
    1,482
  5. 5
    University of Washington+3% YoY · 259 hires last year
    1,360
  6. 6
    University of Southern California+8% YoY · 263 hires last year
    1,300
  7. 7
    UCLA+8% YoY · 206 hires last year
    1,099
  8. 8
    University of Illinois Urbana-Champaign+8% YoY · 243 hires last year
    1,099
  9. 9
    Columbia University+7% YoY · 194 hires last year
    989
  10. 10
    UC San Diego+6% YoY · 195 hires last year
    943

Source: LinkedIn Talent Insights · scraped 2026-05-14