Excel across generations

Will Gen Z be the generation that finally leaves Excel? Anyone who hires analysts or chooses tools for an FP&A team ends up asking some version of the question. If Millennials and Gen Z grew up as “digital natives” who think differently about technology, the case for the grid might only be a case for their managers.

This page is part of the research behind In defence of Excel. Much of the evidence in that essay predates most of today’s analysts, so here we check whether it still holds for the people doing the work now.

Do young finance professionals use Excel less?

No evidence says they do, though surprisingly little data splits spreadsheet use by age at all. The strongest source is independent. In autumn 2024 the Association for Financial Professionals surveyed 362 FP&A and finance practitioners: 96% used spreadsheets for planning and 93% for reporting on a daily or weekly basis. AFP added: “These numbers hold true when considering company size, geography, type of ownership and level of seniority.” Seniority is job level, not age, but it shows that Excel isn’t a junior skill people grow out of.

The only survey that splits finance staff by age is a small poll by a finance-software company: 212 finance professionals in the US and UK, in October 2025. As ITPro reported it, “More than half (54%) of 22-32-year-olds in the CFO’s office said they love Excel, yet that fell to just 39% for both 33-50-year-olds and those older than 51.” The youngest also spent the most time in it: 83% more than five hours a day, against around 70% of older colleagues. Treat this as colour, not proof. The sample is small and not random, the company has an interest in the result, no method is published, and CFO.com gives the youngest band as 22 to 35. One poll also can’t separate age from role: juniors build the models and seniors review them. But it can’t be read as a generation walking away.

To place ages, we use Pew Research Center’s birth years: Baby Boomers 1946–1964, Generation X 1965–1980, Millennials 1981–1996 and Generation Z from 1997 onward. Only the Boomers, Pew notes, are “officially designated by the U.S. Census Bureau”. The rest are conventions.

Do generations really differ at work?

Much less than the popular story suggests. In 2020 a consensus report from the US National Academies of Sciences, Engineering, and Medicine concluded that “a focus on generational characteristics is not supported by science and is not useful for informing workforce management decisions.” Much of the research on generations, it explained, “uses methods that cannot separate generation effects from other changes over time and the life course”.

The studies underneath agree. A 2012 meta-analysis by Costanza and colleagues (a study pooling the results of 20 earlier studies), in the words of a later review, “found no appreciable evidence for generational differences in job satisfaction, organizational commitment, and turnover intentions”. In two large studies, one comparing parents with their own children, Cucina and colleagues found that “over 98 percent of the variance in workplace attitudes lies within groups, as opposed to between groups”. Two analysts born in the same year differ far more from each other than one generation does from the next.

The strongest evidence on the other side compared US high-school seniors in 1976, 1991 and 2006, so everyone was asked at the same age. Twenge and colleagues found leisure values rising steadily from Boomers to Millennials (d = .57, a standard measure of how far apart two averages are). Even that group later found “no clear cut-offs between generations”. And all of this concerns values and attitudes, not how anyone reasons with numbers. A modest difference in how much people value time off may matter for month-end workload. It says nothing about how someone reads a model.

Pew Research Center, which has long published research by generation, changed its policy in May 2023: “We’ll only do generational analysis when we have historical data that allows us to compare generations at similar stages of life.” Its reasoning is a good test for any claim about young analysts:

“When doing this kind of research, the question isn’t whether young adults today are different from middle-aged or older adults today. The question is whether young adults today are different from young adults at some specific point in the past.”

Age, period and cohort, in plain words

Any difference between Millennials and Boomers measured in one survey mixes three things. Age is being 25 rather than 55, with less experience and a different stage of life. Period is what everyone lives through at the same time: the same tools, economy or pandemic. Cohort is the group born in a given span and shaped by the same events growing up. A generation is a cohort with a name.

The three are locked together by arithmetic. In a study run in 2020, everyone born in 1980 is 40. As Rudolph and colleagues put it, “Period 2020 = Age 40 + Cohort 1980”, so “unique effects of age cannot be separated from whatever cohort effect might exist and vice versa.” Where studies have been designed to pull the three apart, the National Academies found that age and period mattered more:

“Many of the research findings that have been attributed to generational differences may actually reflect shifting characteristics of work more generally or variations among people as they age and gain experiences. Indeed, in the few studies designed to separate age, period, and cohort effects, the results point to greater significance of age or period effects than of cohort or generation effects.”

So ‘young analysts struggle with X’ is at least as likely to mean ‘new analysts struggle with X’, or ‘everyone’s tools changed’, as anything about Gen Z. Plan for the experience level, not the birth year.

Is the digital native a myth?

The idea that young people arrive fluent in technology comes from a 2001 opinion essay by Prensky, with no data behind it, which claimed that “today’s students think and process information fundamentally differently from their predecessors”. It hasn’t survived testing. Bennett, Maton and Kervin likened the debate to “an academic form of a ‘moral panic’”. Helsper and Eynon found that “generation is only one of the predictors of advanced interaction with the Internet”, with experience, breadth of use and education sometimes mattering more. And Kirschner and De Bruyckere found “no such thing as a digital native who is information-skilled simply because (s)he has never known a world that was not digital.”

The large assessments agree. ICILS 2023 tested the computer and information literacy of 132,998 14-year-olds in 34 countries. Nearly half were below Level 2, able to work “under instruction but not independently”, and scores were “typically lower in 2023 than in 2018 and 2013” even as students used more technology. The authors were blunt:

“These consistent data across three cycles of ICILS, present confronting evidence of grade 8 students who, despite coming from a generation with high levels of use of digital technologies, have not developed the necessary capabilities to use digital information effectively, both as consumers and producers of digital information.”

Young adults look the same. An Educational Testing Service analysis of the OECD’s 2012 adult-skills survey found 56% of US Millennials aged 16 to 34 below the minimum standard for problem-solving in technology-rich environments. Yet some of our own profession’s research still assumes the opposite: ACCA and IFAC’s 2021 study of about 9,000 people aged 18 to 25 included the advice “Tap into Gen Z’s digital mastery” without measuring spreadsheet skill at all. The risk for a finance team is as much overconfidence as lack of skill.

What students bring to the job

The most direct test of spreadsheet skill is old but telling. In 2007–08 Grant, Malloy and Murphy compared first-year business students’ ratings of their own computer skills with their actual performance. Most of the 173 were 17 to 19, early Millennials. The gap was significant for spreadsheets, and most students “could only perform two of the five basic spreadsheet tasks”. The authors warned of a potential “‘perfect storm’ manifesting between students’ perceived proficiency of computer application skills and the actual assessment of those skills.”

Accounting educators report the same gap. In a 2016 survey of 245 accounting faculty, Rackliffe and Ragland found that “students are not fully proficient in Excel based on faculty’s perceptions.” Employers take the skill for granted: in a 2017 study, 43% to 51% of business-graduate job postings in six US states asked for Excel. The US CPA exam now “includes access to a spreadsheet that can be used to calculate responses and organize data provided”, which is a floor, not proof anyone can build a model. And since Grant’s students were Millennials, the confidence gap is not new with Gen Z.

Where Excel is actually learned

Mostly at work, from other people. Sarkar and Gordon’s 2018 interviews found that “feature adoption in spreadsheets is informal, opportunistic, and social”. One participant, an accountant at a small company, described a path many of us will recognise:

“I’m sure people within the accountancy company taught me. They probably showed me. … But I never did a course in Excel. I’ve never taken a formal learning in Excel.”

Everyone in that study was 35 to 65, so this is how spreadsheets are learned, not how one generation learns them. In a survey of 1,597 business spreadsheet users, 52% had learned from colleagues’ demonstrations and only 38% had ever had classroom instruction. A 2024 survey of 100 users in finance and administration found people less confident in their “general spreadsheet proficiency” than in the spreadsheets of their own job. Your senior analysts and your team’s conventions are the real training programme, so give them time and credit for teaching.

Are Excel’s strengths a generational taste?

The research on why spreadsheets work names a handful of properties: structure for free, instant recalculation, values you can see, direct manipulation (changing the thing itself and watching the effect) and learning a step at a time. We cover them on why the grid fits how we think. Much of that evidence came from users who were then mostly Boomers and Gen X. No study has given Gen Z and Boomers the same spreadsheet tasks, but three kinds of evidence say the properties are general, not generational.

The recent studies were run with young users. Of the 21 people Chalhoub and Sarkar interviewed in the UK in 2021, 18 were aged 18 to 34, born roughly between 1986 and 2003: Millennials and Gen Z. Sixteen said free-form sheets gave them “a sense of flexibility, control, autonomy and ownership”, and the researchers concluded:

“Despite efforts to augment or replace the 2-dimensional spreadsheet grid with formal data structures such as arrays and tables to ease formula authoring and reduce errors, the flexible grid remains overwhelmingly successful.”

In Srinivasa Ragavan, Sarkar and Gordon’s 2021 study, 12 of 15 participants were aged 18 to 40, and the interviewees in a 2019 preprint by Borghouts and colleagues had a mean age of 38. The samples are small, but mostly Millennials and Gen Z, and the classic findings turned up again.

Instant feedback is what we build for children. Scratch, a programming environment for “users (primarily ages 8 to 16)”, rests on the same principle as recalculation: “Just click on a stack of blocks and it starts to execute its code immediately.” Seeing the result of a change at once was chosen for the youngest learners. It isn’t a habit of older ones.

Feedback helps learners at every stage. A 2020 meta-analysis of 435 studies, from kindergarten to university, found a medium effect of feedback on learning (d = 0.48), though it didn’t test age formally. Instant recalculation works on the same principle.

None of this makes the grid a guarantee. A 1997 experiment found that continuous visual feedback “did not significantly help with debugging in general”, and in the 2021 study 12 of the 15 mostly young participants went back to the spreadsheet’s author for clarifications. The next post in this series looks at where Excel breaks. But as far as the evidence goes, the mind a model has to fit is the same at 25 as at 55. What differs is experience, and experience can be taught.

Is there a Gen Z Excel culture?

There is, and it’s competitive. In 2023 more than 3,500 students from nearly 500 universities entered the qualifying rounds of the Microsoft Excel Collegiate Challenge, and its site now counts 18,153 participants. The 2025 winner, Pieter Pienaar, was a 23-year-old accounting science master’s student at the University of Pretoria. He was clear that competition Excel isn’t FP&A: “The challenges are nothing like normal Excel work, so you shouldn’t expect a cash budget. It’s closer to mini-games built inside spreadsheets”. A good ranking is a real signal on a CV. Model structure still has to be taught.

The Microsoft Excel World Championship was streamed on ESPN3 in December 2021 and has held live finals in Las Vegas since 2023. Its champions learned the craft at work: Andrew Ngai, a senior actuary in Sydney, won his third straight title in 2023; Michael Jarman, a Toronto-based head of model development, won in 2024; and Diarmuid Early, who started using Excel professionally at Boston Consulting Group in 2008, won in 2025. Below them the pipeline starts young. The Financial Modeling World Cup runs an under-25 category, and in July 2026 the Microsoft Office Specialist World Championship brought 135 finalists aged 13 to 22 from 36 countries to Anaheim.

There are creators too. Kat Norton, known as Miss Excel, started posting on TikTok in June 2020, and by November 2022 had over 870,000 followers there and over 650,000 on Instagram. Her university called her “the go-to instructor for the ubiquitous software among the under-30 crowd”, a description rather than audience data. Short videos teach tricks, not model design.

What about Google Sheets?

Younger companies and students may well start in Google’s browser suite, though the evidence comes mostly from Google. Its startup page claims that 96% of the startups on Forbes’ Next Billion-Dollar Startups list use Google Workspace, with no year or method given. Its education page says 75% of students choose Google Workspace, a figure attributed to SADA Systems, a Google Cloud partner. Both measure the whole suite, not spreadsheets, and neither looks at finance. Even at face value, they describe a move from one spreadsheet product to another, not away from the grid. A hire who learned on Google Sheets already thinks in rows, columns and formulas.

What the evidence doesn’t show

No one has tested the grid against a generation directly, by giving Gen Z and Boomers the same spreadsheet tasks with and without live recalculation. The case that its strengths are general rather than generational rests on several kinds of evidence pointing the same way, and the recent studies behind it are small and qualitative.

Nor is there an independent survey of Excel use by age in finance. We found none from ICAEW, AICPA & CIMA, IMA, CPA Canada or CPA Australia, and AFP reports seniority, not age. The one age-split poll is small, vendor-sponsored and without a published method. The most direct test of students’ spreadsheet skill is a single-university study from the late 2000s, and nobody has compared Excel and Google Sheets use by age in finance.

What the evidence does show is enough for the argument in In defence of Excel. The labels are weak, the digital native was a myth, and the people who re-confirmed the grid’s strengths in recent studies were mostly under 40. Each new cohort learns Excel the way the last one did, at work and from colleagues. The cognitive research is on why the grid fits how we think, and the surveys and job postings on Excel in FP&A, by the numbers.