Generational differences in online search are differences in the starting point: which screen lights up first when someone has a question. In the UK that starting point is still a search engine for most age groups. What changes is second place: for a 20-year-old it is social media or a chatbot, for a 75-year-old it is the television news.
Almost everything written on this subject comes from US surveys, often three or four years old. Here I try to put UK numbers where they exist, and to state the market measured where they do not.
The 2022 quote that still props up the first page
On 12 July 2022, on stage at Fortune's Brainstorm Tech, Prabhakar Raghavan, then senior vice president of Google's Knowledge & Information division, said: "In our studies, something like almost 40% of young people, when they're looking for a place for lunch, they don't go to Google Maps or Search. They go to TikTok or Instagram". TechCrunch reported it the same day.
Four years on, that sentence still holds up part of what Google UK shows for how different generations search online. On 26 September 2026 one of the first-page results summarised it in its snippet as "nearly 40% of Gen Z prefer to search on TikTok and Instagram over Google Search and Maps", and another stated that Gen Z searches TikTok and Reddit before Google (DataForSEO, Google UK).
It is worth reading for what it is. It was not a published study, it was an executive quoting from memory an internal piece of research that was never made public. It covered one category only, restaurants. The sample was American, aged 18-24. And it is four years older now.
In 2026 Gen Z's preference for TikTok over Google halved
Adobe Express repeated its survey in January 2026 with 1,007 people in the United States, 807 consumers and 200 small business owners, collected through SurveyMonkey. The results point in two opposite directions.
Use is growing: 49% of US consumers have used TikTok to search for something, against 41% in 2024.
Preference is falling: the share of Gen Z who prefer TikTok to Google as a search engine went from 8% in 2024 to 4% in 2026. Only 25% of Gen Z consider TikTok effective for finding information, and that falls to 17% among millennials, 12% in Generation X and 5% among baby boomers.
In the same survey 14% of consumers say they would rather turn to ChatGPT than Google. That is three and a half times the TikTok figure, and it is almost identical across age groups: 12% in Gen Z, 15% among millennials and Generation X, 14% among baby boomers.
These are US figures, collected by a vendor on a sample that was 53% millennial and 15% Gen Z, with no stated margin of error. The limits matter. But the direction is readable, and the first page has not caught up with it yet.
In the UK the real generational gap runs through artificial intelligence
On using social media as a search tool, there is no UK figure showing a dramatic generational gap. On AI there is, and it is wide. These numbers come from Ofcom's Adults' Media Use and Attitudes report, published in May 2025 with 2024 data, and refer to having used AI tools such as ChatGPT or Gemini.
| Have used AI tools, 2024 | United Kingdom | Change |
|---|---|---|
| All adults | 31% | from 23% in 2023 |
| Aged 16-24 | 53% | not reported |
| Aged 25-34 | 48% | not reported |
For context, Eurostat's 2025 ICT survey puts generative AI use at 32.7% of 16-74-year-olds across the EU and 63.8% of 16-24-year-olds. The questions and years differ, so the figures are not directly comparable, but the UK sits in the same range as the European average rather than far ahead of it.
At the other end of the age scale the gap shows up in attitudes. In a Tony Blair Institute and Ipsos survey of 3,727 UK adults (fieldwork May-June 2025), nearly half of respondents said they never use generative AI, at home or at work. Only 16% of 16-24-year-olds see AI mainly as a risk, against 41% of 65-74-year-olds.
The age groups that use AI most are the ones now entering the workforce, and they bring their habits with them. That is what makes the figure relevant for anyone producing content today, not in 2030. If you want to understand how a language model behaves when it meets your content, the starting point is generative engine optimisation.
How people in the UK get their information, age group by age group
Ofcom's News consumption in the UK 2025 report, published on 21 July 2025, shows where people start when they look for news.
- 70% of adults get news online, 68% from TV including on-demand: online is ahead.
- Around eight in ten 16-24-year-olds use online platforms for news.
- Among those aged 75 and over, nine in ten use TV for news and only four in ten use online sources.
- 51% of adults use social media for news, with Facebook alone at 30%.
- Six in ten adults use an online intermediary (social media, a search engine or a news aggregator), and Google search reaches 34%.
This picture contains something the American pieces leave out: television. In the UK reading of generational behaviour, the search engine's competitor for the over-75s is not another website, it is the BBC evening news.
How much the shape of the query really changes
US studies say younger people write longer queries that look more like questions. Fractl, with 956 US consumers recruited through Mechanical Turk, measured about five words per query for Gen Z against about four for the other generations, with more frequent use of "best" and "how to". WebFX, analysing over 3,000 queries, found an average of 5.83 words for Gen Z and 43.57% of searches phrased as questions.
On real Google data the curve is much flatter. I measured the Search Console queries of two sites we manage over the same period, 1 May to 20 August 2026. Both are Italian and the figures cover traffic from Italy only, so read them as an example of the method rather than a UK benchmark.
| Metric (Italy, 1 May - 20 August 2026) | Outdoor equipment e-commerce | B2B search marketing blog |
|---|---|---|
| Unique queries | 8,300 | 986 |
| Impressions | 107,554 | 15,642 |
| Words per query (impression-weighted average) | 2.6 | 3.15 |
| Impressions from queries starting with a question word | 0.99% | 8.6% |
| Impressions from queries of two words or fewer | 55.4% | 26.7% |
| Share of impressions from mobile | 65.3% | 25.8% |
Two sites are not a national sample. But they say something worth keeping in mind: the gap between 2.6 and 3.15 words is not generational, it is about intent. The e-commerce site catches people who already know what to buy, the blog catches people with an open question. The audience of an outdoor equipment shop is not older than the audience of an SEO blog, it is simply closer to buying.
In practice: before segmenting content by generation, segment it by search intent, because that is where the variance can actually be measured. It is also why keyword research based on the data you already have in Search Console almost always beats a demographic segmentation drawn up at a desk.
The four generations, one by one
The figures below come from a SEO.com survey of over 700 people in the United States. They show a direction, they do not describe the UK market. Where there is a UK figure on the same behaviour, I have put it alongside.
Baby boomers, born 1946 to 1964
92% start from a search engine, 78% prefer text results, 72% say they avoid AI and chatbots. The UK equivalent points the same way: among the over-75s nine in ten get news from TV and only four in ten from online sources, and 65-74-year-olds are the group most likely to see AI mainly as a risk.
A useful correction, though: according to Ofcom, 21% of over-65s have no internet access at home, which means roughly four in five do. The idea that this age group cannot be reached online is ten years out of date. The group that is still harder to reach is the over-75s, 32% of whom had no internet access in 2024.
Generation X, born 1965 to 1980
87% start from a search engine, 71% put reliability first among their criteria, 50% avoid AI and chatbots. It is the generation that compares most before deciding, and in B2B it is often the person comparing suppliers: they read the article, then search for the company name, then ask for a quote. I wrote about how to build that path in the guide to B2B SEO strategy.
Millennials, born 1981 to 1996
84% start from a search engine, 79% search on mobile, 41% check several sources before trusting one. In the UK they are also the group where AI use is already close to half: 48% of 25-34-year-olds have used AI tools. They are the group most likely to come across a generative answer and then go and check the source anyway, which makes appearing in Google's AI Overviews a different problem from simple traffic.
Generation Z, born 1997 to 2012
79% search on mobile, 34% often use social media as a starting point, 24% use forums such as Reddit, 23% AI and chatbots. The figure usually left out when this generation is described: 96% say they want text and video together.
The Reuters Institute's Digital News Report 2025, across 48 markets, finds that 15% of under-25s use an AI chatbot for news at least once a week, against 7% of the whole population. In the same report, 54% of 18-24-year-olds name social media and video as their main source of news, thirteen points more than the previous year.
On the scale of AI use there is one number that helps more than any other. In the study How People Use ChatGPT, written by OpenAI researchers and published by the NBER on about 1.1 million conversations sampled between May 2024 and June 2025, almost half the messages sent by adult users came from people under 26. Over the same period the "seeking information" category grew from 14% to 24% of all use of the platform.
What changes for people who produce content in the UK
Google is still the floor. Even in the youngest UK age group, only just over half have ever used an AI tool, and nearly half of all adults never use generative AI. Moving budget from organic search to visibility on LLMs means getting ahead of the market. Our SEO for AI service starts from that assumption: you add a channel to the work that already performs.
Pages that win on Google also win in generative answers. Models cite what they find indexed and well structured, so the groundwork does not change. What changes is the answer formats, which is why we have collected the specific tactics for ranking in answer engines and for appearing in ChatGPT answers.
Video goes alongside text. 96% of Gen Z ask for both. An article with an embedded video covers two behaviours with a single URL, and video SEO is less crowded than text.
Social media is for being discovered, not for ranking. The link between social signals and rankings is weaker than people claim, as we found when looking at the data on social media and SEO. If your audience is young it is still worth covering the platforms' internal search, starting with TikTok, and remembering that there are alternative search engines with their own index.
An example from our own work, already public. Macropix, a Milan-based LED wall manufacturer, went from position 88 to position 2 on Google Italy for a keyword with 4,400 monthly searches, and in July 2026 held 25.55% share of voice on the "ledwall" cluster. None of that result came from a strategy built around generations: it came from separate pages for separate queries. The method is described in full in the case study on SEO for manufacturing, and the same logic underpins the SEO consultancy projects we run.
Segmenting by age is the most convenient option and the least useful
The industry rarely says it, but segmenting an audience by generation is nearly always the worst shortcut available. It ends up in the slides because it is the only variable the client already knows, not because it explains anything. In search data, the variables that really separate behaviour are two others: the device people search from and where they are in their journey. A 50-year-old searching on a phone inside a shop behaves like a 20-year-old searching on a phone inside a shop.
When the generational difference really does matter, as with artificial intelligence in the UK, the gap is so wide that it does not need segmenting. Over half of 16-24-year-olds have used AI tools, and a 65-74-year-old is two and a half times as likely as a 16-24-year-old to see AI mainly as a risk. You can see it without a spreadsheet.
Frequently asked questions
No. In the Adobe Express survey of January 2026 with 1,007 people in the United States, the share of Gen Z who prefer TikTok to Google as a search engine fell to 4%, half the 8% of 2024. Use of TikTok for searching is growing, but it sits alongside Google, and only 25% of Gen Z consider it effective for finding information.
According to Ofcom, 31% of UK adults had used AI tools such as ChatGPT or Gemini in 2024, up from 23% in 2023. The figure rises to 53% among 16-24-year-olds and 48% among 25-34-year-olds. In a Tony Blair Institute and Ipsos survey from mid-2025, nearly half of UK adults said they never use generative AI.
In US studies yes, but only slightly: Fractl measures about five words per query for Gen Z against about four for the other generations. On the Search Console data of two sites we manage, between May and August 2026, the impression-weighted average length is 2.6 words for an e-commerce site and 3.15 for a B2B blog. The difference is explained by the type of search, not the age of the searcher.
Rarely. In search data the variables that separate behaviour are the device used and the stage of the buying journey, not the year of birth. Segmenting by search intent produces measurable results faster. The exception is artificial intelligence, where in the UK over half of 16-24-year-olds have used AI tools and older adults are far more likely to see AI mainly as a risk.
Among those aged 75 and over, nine in ten get news from TV and only four in ten from online sources, according to Ofcom's News consumption in the UK 2025 report. Most over-65s can still be reached online: Ofcom reports that 21% of them have no internet access at home, so roughly four in five do.
