Google

The AI Turn: From Invisible to Central (2014–2025)

AI was nearly absent from Google's careers copy before 2014 (0.0% prevalence 1999–2013). A single mention of "deep learning" appeared in 2014 (2.6% prevalence), and the term remained sparse through 2017 (4.2%). But from 2019 onward, AI language exploded: 6.6% in 2019, 7.9% in 2020, 4.4% in 2022, 8.6% in 2023, 21.2% in 2024, and 23.8% in 2025. By 2024–2025, AI mentions appeared not just in mission/brand copy but in employee stories and process/logistics sections—signaling it as a core hiring and identity marker.

The framing has also shifted. Early mentions (2014–2015) were neutral or technical ("deep learning neural networks"). By 2019–2020, AI was paired with inclusion and responsible development ("Project Respect helps create more inclusive artificial intelligence algorithms"). By 2024–2025, "responsible AI" became explicit, with a dedicated role ("Head of Product - Responsible AI") featured in employee stories. This reflects Google's effort to position AI not as a neutral tool but as a domain where diversity and ethics matter.

Diversity, Equity, and Inclusion: From Absent to Structural (1999–2025)

DEI language was virtually absent from Google's careers copy in 1999–2016. Equal opportunity and merit-based hiring appeared in 1999–2000 ("not influenced by race, color, gender, sexual orientation..."), but these were framed as legal compliance, not values. The first explicit diversity language appeared in 2005 ("talented people from diverse backgrounds"), but remained aspirational and vague.

A sharp inflection occurred in 2017. That year, Google published its first diversity report and began naming specific demographic groups ("Blacks and Hispanics") and setting representation goals. By 2018, structural process language dominated: unconscious bias training, pay equity audits, leadership accountability, and intersectional data. The language shifted from "we celebrate diversity" to "we are taking a systemic approach to improve outcomes in workforce representation."

From 2020 onward, "belonging" became the organizing frame, emphasizing that everyone should feel they belong at Google and in tech. By 2022–2025, this extended to disability inclusion, LGBTQ+ equity, and military/veteran support. The tone moved from aspirational to operational: specific programs, metrics, and timelines (e.g., "double Black+ representation by 2025") replaced vague commitments.

Altruism and Purpose: Sustained but Reframed (1999–2025)

Google's altruism scores remained consistently high (z-scores 0.5–1.4 from 2004 onward), but the language evolved. Early framing (1999–2008) emphasized individual impact: "Your ideas can make a difference. Millions of people will use and appreciate your software." This was personal, aspirational, and tied to product impact.

From 2012 onward, altruism language broadened to include systemic change. By 2019–2022, Google's altruism copy increasingly focused on external impact—education programs, supplier diversity, nonprofit partnerships, and community investment. The 2021–2022 period saw the highest altruism scores (z-scores 1.18–1.37), driven by language about "building a world where belonging is the norm" and investing in underrepresented communities.

By 2024–2025, altruism language remained strong but became more specific to AI and accessibility: "AI has significant potential to help solve challenging global problems," and detailed work on inclusive camera technology, mammography screening, and online safety. The shift reflects Google's effort to position both its products and its hiring as serving broader social good.

Craft and Performance: Decline and Moderation (2000–2025)

Craft language (emphasis on quality, excellence, building great products) peaked in 2012 (z-score 1.21) and 2019 (z-score 1.31), then declined. By 2024, craft scores dropped to -0.70 (below peer average). Early craft language emphasized technical excellence ("world-class skills," "best technology") and the satisfaction of building meaningful products. By 2012, this expanded to include culture and process ("great, creative things are more likely to happen with the right company culture").

Performance language (focus on achievement, results, high standards) followed a similar arc: strong in 2005–2008, peaked in 2012 (z-score 1.21), and moderated thereafter. By 2024, performance scores fell to -2.43, the lowest in the dataset. This suggests Google's careers messaging has de-emphasized individual achievement and technical excellence relative to belonging, representation, and systemic change.

Meritocracy language remained relatively stable (z-scores -0.2 to 1.3) but showed no clear trend. Early language emphasized hiring "only the best" and basing "employment decisions exclusively on merit." By 2017–2019, meritocracy language persisted but was paired with acknowledgment of bias and systemic barriers, suggesting a shift from pure merit-based framing to equity-aware meritocracy.

Wellbeing and Perks: From Lifestyle to Systemic Support (2000–2025)

Wellbeing language was modest in early years (z-scores -0.2 to 0.6 in 2000–2008) but focused on lifestyle perks: "massage therapy, roller hockey, game room, gym, and all the snacks you can eat." By 2004, this expanded to on-site healthcare, childcare, and professional development.

From 2018 onward, wellbeing language shifted from perks to systemic support. Google introduced parental leave ("12 weeks fully paid leave"), caregiver leave (expanded to 14 weeks by 2022), and healthcare equity ("quality healthcare regardless of gender identity or sexual orientation"). Wellbeing scores rose sharply in 2018 (z-score 1.55), 2022 (z-score 1.50), and 2025 (z-score 1.90).

The 2024–2025 language emphasized mental health support for veterans, LGBTQ+ youth crisis services, and disability accommodations—moving from individual wellness to community and systemic wellbeing. This aligns with the broader shift toward belonging and equity.