Gen Z is the demographic cohort born between 1997 and 2012, following the Millennials and preceding Generation Alpha. Its members turn 14 to 29 during 2026. The label names the first group to meet the internet, the smartphone and the social feed as ambient conditions rather than arrivals. Its boundaries are fixed by neither statute nor statistical convention. They were drawn by a research institution, adopted by the press, then hardened into a planning unit by an industry that buys media against age.

That last step turns the term from sociology into an ad tech problem. No demand-side platform sells a Gen Z segment, so buyers have to express the cohort in whatever age fields a platform offers, and those fields were never built to carry birth-year boundaries.

How the cohort reaches a bid request

Google Ads, Search Ads 360 and Display & Video 360 expose one fixed age ladder: unknown, 18 to 24, 25 to 34, 35 to 44, 45 to 54, 55 to 64, and 65 and over. Meta, Microsoft Advertising and Amazon Ads use comparable bands. A cohort defined by birth year has to be mapped onto those buckets, and the mapping moves every year.

The 2026 arithmetic shows the problem. Members born between 2002 and 2008 are 18 to 24 this year, so that bracket is entirely Gen Z. The 25 to 34 bracket covers birth years 1992 to 2001, mixing five Gen Z birth years with five Millennial ones. Members born from 2009 to 2012 are minors, so four of the cohort's sixteen birth years sit outside the ladder altogether: according to Google's Search Ads 360 documentation, users under 18 are reported in the unknown age group. A buyer selecting both brackets captures most of Gen Z, a slice of the Millennials, and none of the youngest quarter.

Underneath the interface, the signal has thinned. Early OpenRTB versions carried a user object with yob, a four-digit year of birth, alongside gender. Both were deprecated in OpenRTB 2.6, and Google's published OpenRTB proto marks each field as unsupported. Age now travels inside user.data segment objects from data providers, or inside a platform's logged-in graph, where it is modelled rather than declared.

Modelled age is approximate, and platforms say so. Display & Video 360 documentation on on-target reach names the reasons: users misreport or withhold details, unsigned-in sessions defeat inference, shared devices blur individuals into households. Its worked example puts a $100,000 budget aimed at women aged 25 to 44 in the United States at 5.84 million people reached out of 43.8 million, a 13.3% on-target figure. Planning against a cohort means buying a probability distribution, not a census.

Where the boundaries came from

The alphabetical convention started with Generation X and continued by default. An Advertising Age editorial in August 1993 coined Generation Y, a placeholder later displaced by Millennials, the term Neil Howe and William Strauss introduced in 1989. Generation Z arrived as the next letter. Rival names lost: Homeland Generation, from Strauss and Howe in 2005, and iGen, from the psychologist Jean Twenge in 2017. Zoomer stuck informally.

Pew Research Center set the dominant boundary. On January 17, 2019, Pew president Michael Dimock published the decision to treat 1996 as the final Millennial birth year, making anyone born from 1997 onward part of a new generation. A 16-year Millennial span matched the 1965 to 1980 span Pew uses for Generation X. The substantive argument was that social media, constant connectivity and on-demand media were innovations Millennials adapted to, and that those born later simply assumed.

Other definitions circulate. McCrindle Research, the Australian firm that named Generation Alpha, uses uniform 15-year spans and places Gen Z at 1995 to 2009 and Generation Alpha at 2010 to 2024. McKinsey has published 1996 to 2010. The education network Backpack Media, launched March 4, 2026, built its audience proposition on 1997 to 2012, with Generation Alpha from 2013. The spread is two or three birth years at each edge, and there it decides which cohort a buyer describes.

What the cohort does with media

Distribution behaviour is where the label earns its keep. Research on European broadcast markets found Gen Z viewers giving 16% of their viewing time to linear television against 55% for baby boomers, with YouTube and social video above 40%. Nielsen's 2026 Upfront Planning Guide, published March 12, 2026, put streaming at 66.7% of ad-supported television time among adults 18 to 49.

Audio runs heavier than the stereotype allows. Edison Research's Share of Ear data for the first quarter of 2026 recorded 13 to 34 year olds spending four hours and thirty minutes a day with audio, more than any older group, and 84% of Gen Z monthly podcast listeners engage with video podcasts.

Discovery has partly migrated off search engines. Around 57% of TikTok users search within the app, although complaints about result quality have accumulated. Ashley Liddell of the agency Deviation said "Gen Z is leading that shift." Pinterest reported the cohort passing half its global user base during 2025 and, in its 2026 Men's Trend Report, said Gen Z men account for more than 62% of male monthly searches globally. A January 2025 Zety survey of 896 United States Gen Z employees found 46% had secured a job or internship through TikTok.

Commercial behaviour is less uniform than vendor decks suggest. RTB House research from March 2026 found Gen X and boomers nearly twice as likely as Gen Z to cut spending in 2026, and Adobe reported Gen Z 73% more likely than older groups to buy on impulse for stress relief. NIQ data on European online fashion cuts the other way: shoppers aged 18 to 29 are a growing share of buyers, while consumers aged 40 to 44 post the highest annual spend. Volume and value sit in different cohorts.

The cohort regulation cut in half

Gen Z is the first cohort with an advertising eligibility line running through its middle: the youngest members are minors, the oldest approach thirty.

Google consolidated five child and teen advertising policies into one hub on January 15, 2025, then began machine learning age estimation in the United States on July 30, 2025, disabling personalisation and restricting sensitive creative for users the model flags as under 18. Reddit switched every European teen account to restricted settings and disabled ad personalisation from June 24, 2026.

Platform-level exclusion goes further. Australia's under-16 restrictions took effect on December 10, 2025 across nine named services, with penalties reaching A$49.5 million, after which Meta reported removing 756,000 accounts through June 30, 2026, split 462,000 on Instagram and 294,000 on Facebook. The United Kingdom announced a comparable measure on June 15, 2026 for spring 2027. France went the other way, its Constitutional Council striking down an under-15 ban on August 14, 2026.

Direct revenue at stake is modest: minors were already constrained. The second-order effects are not: removing an age band alters the engagement signals feeding recommender systems, shifts frequency distributions, and leaves lookalike seeds describing a population that no longer exists in that shape.

Limitations and disputes

The sharpest criticism comes from the institution that fixed the boundary. On May 22, 2023 Pew published guidance warning that generational categories "are not precise, definitive or universally agreed on," that labels invite stereotype, and that conventional generational portraits carry an upper-class bias skewed toward college-educated experience. Pew committed to using the labels sparingly and only where data supports a generational lens rather than an age or period effect.

That distinction is the technical objection. Separating cohort effects from life-stage effects requires age-period-cohort analysis that most commercial research does not attempt. When the IAB found 39% of Gen Z consumers negative toward AI advertising against 20% of Millennials in fieldwork running October 2025 to January 2026, the design could not establish whether the gap belonged to the cohort or to being under 30.

A commercial incentive sits alongside it. Much cohort research comes from firms selling access to the cohort, and the figures are often self-reported platform data rather than independent measurement. Definitional drift compounds it: a study on McCrindle's range and one on Pew's are not measuring the same people, and the difference rarely reaches the headline.

Not the same as

Millennials. The preceding cohort, born 1981 to 1996 under Pew's definition and aged 30 to 45 in 2026. The two overlap inside the 25 to 34 bracket, which is why studies pair them so often and why bracket-level results attribute poorly.

Generation Alpha. The succeeding cohort, dated 2013 onward by Pew-aligned sources and 2010 to 2024 by McCrindle. Entirely minors, and governed by children's advertising regimes rather than teen ones.

Zillennial. An informal label for those born roughly 1993 to 1998, on the cusp. No standard boundary, no platform representation.

The 18 to 24 demo. A media buying band, not a cohort. It refreshes annually as birth years roll through, so it will describe Generation Alpha by the 2030s.

Recent developments

Cohort language grows more prominent in platform marketing even as the operational unit stays the age bracket. Amazon Ads staged a Wisconsin campus concert for three brand sponsors in September 2026, built around the college transition as a purchasing moment.

Trust measurement is the recurring theme. Yelp research published July 30, 2026 found 43% of Gen Z respondents trusting AI summaries for restaurant decisions, and Gracenote's 2026 study recorded 78% of Gen Z verifying chatbot results, matching Millennials and above boomers at 58%. Findings of that kind complicate the description of the cohort as credulous digital natives, and they are why the label keeps being tested rather than retired.

Timeline

  • 1989: Neil Howe and William Strauss introduce the term Millennials
  • 1991: Douglas Coupland's novel popularises Generation X, establishing letter-based cohort naming
  • August 1993: An Advertising Age editorial coins Generation Y, later displaced by Millennials
  • 2005: Strauss and Howe propose Homeland Generation for the post-Millennial cohort
  • 2017: Jean Twenge publishes iGen, offering a competing name for the cohort
  • January 17, 2019: Pew Research Center sets 1996 as the final Millennial birth year, making 1997 the start of Gen Z
  • April 2022: OpenRTB 2.6 is released, deprecating the user.yob and user.gender fields
  • May 22, 2023: Pew publishes cautions on generational research and commits to using labels sparingly
  • January 15, 2025: Google consolidates five child and teen advertising policies into one hub
  • July 30, 2025: Google begins machine learning age estimation for United States ad protections
  • December 10, 2025: Australia's under-16 social media restrictions take effect across nine platforms
  • March 12, 2026: Nielsen's Upfront Planning Guide puts streaming at 66.7% of ad-supported viewing among adults 18 to 49
  • June 15, 2026: The United Kingdom announces an under-16 social media ban for spring 2027
  • June 24, 2026: Reddit disables ad personalisation for European teen accounts
  • August 14, 2026: France's Constitutional Council strikes down the country's under-15 ban
  • 2026: The cohort spans ages 14 to 29, leaving four of its sixteen birth years below the age of majority

Summary

Who. Pew Research Center fixed the dominant boundary and publishes the cautions against over-reading it. McCrindle Research, McKinsey and others maintain competing ranges. Media buyers translate the cohort into platform age brackets; Google, Meta, Amazon and Microsoft define those brackets; regulators in Australia, the United Kingdom and the European Union decide which part of the cohort can be addressed at all.

What. A demographic cohort born 1997 to 2012 under the dominant definition, aged 14 to 29 in 2026, expressed in advertising systems as a combination of the 18 to 24 and 25 to 34 age brackets plus an unreachable minor segment reported as unknown.

When. Named as the alphabetical successor to Generation Y during the 2000s, bounded by Pew on January 17, 2019, qualified by Pew's own methodological cautions on May 22, 2023, and progressively split by age assurance and minimum-age rules from January 2025 onward.

Where. Used worldwide, with boundary variants by research house and country, and with addressability varying by jurisdiction as under-16 and under-18 rules diverge between Australia, Britain, France and the United States.

Why. It gives planners a shorthand for a population whose media distribution differs sharply from older groups, concentrated in streaming, audio, short video and in-app search rather than linear television. The shorthand is imprecise, the boundaries are contested, and the platforms that sell against it can only approximate it.