Pop science doesn't reach a "general audience"

I've spent the last several weeks traveling to research conferences and talking with scientists. And when I mention that I'm a science journalist, there are generally either two reactions: enthusiasm ("I became an astrophysicist because of reading Scientific American!") or suspicion ("Aren't you worried over-sensationalized headlines are eroding trust in science?")

I'm finding that the suspicious crowd can immediately be brought around with a single sentence: I write for Quanta.

Scientists love Quanta. I get the sense that it's the favorite pop sci publication of the physics-y corners of science — their idea of what pop sci should look like, if science really does have to be "dumbed down" for the public at all. (Many scientists, I get the sense, are caught in a paradoxical catch-22 of believing that their research is so esoteric that no non-scientist could ever understand it and also not understanding that, actually, many non-scientists can't do math, many can barely read, and most couldn't understand their research unless it was greatly simplified.)

And I get it. I, too, love Quanta. It was one of my favorite science magazines before I started writing for them, and it's one of my favorite science magazines to write for now as a science journalist. They are truly unique in their willingness to tackle complex fundamental science and math, they give journalists the resources and time to really go deep on stories, they're not over-sensationalist, and they take real care to ensure factual accuracy — including hiring dedicated fact-checkers, a rarity these days.

But the thing is, Quanta — or Sean Carroll's Mindscape podcast, or New Scientist, or pick your favorite erudite pop sci medium — is not doing what many scientists seem to think it is doing. Researchers often seem to have this idea that if only all pop science were like the science media they like best, the ignorant masses would be illuminated by the light of reason. Everyone would trust scientists, lose their misconceptions about science and the scientific process, and trust experts while simultaneously embracing scientific uncertainty, supporting a carbon tax, buying an electric car, and getting their kids vaccinated.

I'm being a bit facetious here, of course. But I do think it is fair to say that many (not all) scientists who haven't thought much about scicomm do believe that if only pop sci generally looked more like Quanta, many of the problems with "trust in science" would be solved. There are so many problems with that idea that I feel I'll have to come back to it in future posts. But for now, let's stick to one issue: Even if it were perfect, pop science couldn't do any of the magical things we might want it to do for trust in science because it is not actually "pop" at all.

Popular science has an audience that is overwhelmingly university educated, politically left-leaning, and wealthy. I think most scientists — and science communicators, frankly — have truly no clue who the "general audience" actually is and what communicating with the wide public would look like.

I want to do more than write science entertainment

Let’s talk about science and scicomm in a crazy world

Reviewer, too | Elise CuttsElise Cutts

The "general audience" of popular science is educated, wealthy, and left-leaning

Before moving on to discuss what "general audience" science communication would actually entail, I want to make my point about pop sci not really being "pop" at all. The glossy magazines and trade books that scientists and science journalists consider to be of high quality are written for a small, educated elite.

This is not necessarily a criticism. Not everything has to be for everyone — media outlets and creators are absolutely allowed to decide that they want to focus on a specific audience. (In fact, knowing your audience is very important.) My magazine writing, to call myself out, is very much not for everyone. And in this post, I'm only going to "call out" magazines that I write for. If I'm criticizing anything here, it's not that some pop sci is targeted at an educated audience but rather the tendency of scientists and science writers to imagine that they're somehow reaching some enormous, broad "public" with work in magazines like this. Because we just aren't.

Magazines know the truth, though. They have to — and the ones that sell space to advertisers, who absolutely need to know their target demographic, make at least some information about their readership public. New Scientist and Scientific American work with advertisers, so I'm going to use them as examples.

New Scientist's readership of several million across multiple platforms is highly educated and wealthy. 75% have a degree, 88% fall in the "ABC1" sociodemographic category, roughly meaning professional class. They describe themselves as "globally aware" and "tech savvy" and have an average household income of £75,000.^[1] On its site for advertisers, New Scientist itself describes its readers as an "affluent, engaged audience of thought leaders and early adopters."

Affluent thought leaders and early adopters are not exactly a "general audience."

Across the pond in the U.S., Scientific American is far less coy about targeting an elite. Here's how the magazine describes its audience in its press kit for advertisers:

SCIENTIFIC AMERICAN has a worldwide audience of over 9 million comprised of forward thinking, solution-seeking readers who cement trends and set agendas that others follow -  they are the Minds That Matter. (emphasis mine)

There's more:

The SciAm audience that doesn't just read— they cite our work in research papers, reference us in boardroom decisions, and share our content with their networks. When you partner with Sci Am, you're part of the conversation that matters.

Enough said.

EDIT: I realize in this section I forgot to address political leaning for magazines — I do that for YouTube below using Hank Green's audience. MediaBias/Fact Check rates Scientific American as Pro-Science and Left-Center, New Scientist was actually rated as just "Pro-Science," not left/right so the political leaning might not be as obvious there. Quanta is rated as "Pro-Science" and low-bias. That's all encouraging. But just because coverage is mostly unbiased, that does not mean that the audience will be. Educated groups in the US and UK tend to skew left in both political stances and actual voting, especially in the UK, according to this National Centre for Social Research report. (There's a lot to dig into in that report, but observation that stood out to me: " A person with no educational qualifications had around 2 times the odds of voting for either the Conservatives or Reform UK than someone with a university degree or higher.") The type of science content right-leaning and left-leaning groups consume might also differ: One research study found that right-leaning pop sci readers tend to prefer content on "commercial science" like criminology or medicine to content on basic science like physics and biology, which tends to dominate the "best" pop sci magazines. And then there's just the general trend (US data from Pew) of right-wingers tending to distrusting and disliking both science and scientists, compared to lefties, which I think is safe to assume influences the readership of science magazines.

But that's magazines. As we'll get to below, "readerships" are probably never going to be a "general audience." So what about YouTube?

Though the written word still enjoys a privileged place in pop sci and absolutely dominates science journalism,^[2] there are certainly scicomm channels that have earned respect from scientists (science journalism, as opposed to scicomm, doesn't really seem to have gotten the pivot-to-video memo, Cleo Abrams being one notable exception). These scicomm channels can be much bigger than magazines and their audiences are presumably different. 3blue1brown, SciShow, and Veritasium come to mind. I had trouble finding halfway-trustworthy audience demographics for science channels. But one YouTuber in this space does publish audience data that offers some insights:

Hank Green is a science communicator and internet personality who has started a truly unbelievable number of educational content efforts, including SciShow and Crash Course in addition to the more casual science communication he does on his many other channels. He speaks in very down-to-earth language, he's funny, and he's not afraid to be vulnerable — definitely not a huffy, holier-than-thou academic. Helpfully, Hank Green and his brother John Green do a census of their "Nerdfighter" fan community (caveat: not just the science-interested folks) every year, and in the 2024 census something like 38% of their audience said they also watch the popular science channel Veritasium.

According to the census, the Green brothers' audience is mostly Gen Z and millennials, and only 10% identified themselves as extroverts. Less than half identified as heterosexual (about 90% of the general U.S. and U.K. adult population). 36% said they'd never been diagnosed with a mental illness (77% in the general U.S. population). More than half trust the news media at least a moderate amount (~28% in the general U.S. population). Nedfighterians appear a bit underrepresented at the high end of the income distribution — but the audience skews young, so this probably isn't surprising. 20% make no money at all. A whopping 64% have a bachelor's degree or higher educational qualification — double the general U.S. population. And more than 80% have no children (69% of US adults had children in 2023 according to Gallup). I had trouble finding the political data in the census breakdown video, but in a recent video in which John Green laments not being able to reach people outside his bubble, he mentions that more of his viewers voted for Green Party candidate Jill Stein than Donald Trump in 2024 (she got 1.1% of the general vote).

A very important note on this sample being biased: 14% of respondents identified as non-binary, and more than half as women. Hank said in the video breaking down the census that this was not what his YouTube demographic data suggests, implying a mostly male viewership on YouTube (Edit: this video recently confirmed that the YouTube audience is 73% male). This implies that the survey was subject to selection bias (i.e. women in the Nerdfighter community seem to have been more likely to fill out the survey), and that's something to keep in mind in general when considering this kind of survey data. The yearly census is not scientific data collected following rigorous rules to prevent bias. It's an online survey.

With the caveat of this bias problem, this does not sound like anything close a general audience to me — and that's despite the Green brothers making content on a lot more than just science.

What truly "general audience" scicomm might look like

I hope I've been able to convince you that neither Scientific American, Quanta, New Scientist, and other magazines nor science content creators like Hank Green, 3blue1brown, and Veritasium do not reach a truly general audience. So, what is the general audience like? And what would science communication look like — if it really were trying to reach the most general audience possible?

First, let's be a bit more specific about what I mean by general audience: general audience content should be understandable to any adult without an intellectual disability. That means it should assume no experiences beyond those which can reasonably be assumed to be nearly universal among members of the given society (i.e., a general American audience will have specific shared cultural context that a general global audience will not).

But I'm going to go one step further than comprehensibility. If nobody ever hears what you have to say, it doesn't matter how clearly you say it. So "general audience" communication should also be capable of both reaching and grabbing the attention of any adult without an intellectual disability in a given society.

Let's see where that definition of "general audience" takes us:

1. General audience science communication cannot assume any college level education.

Let's focus on the U.S., because that's the country I know best: most Americans do not have a college degree. General audience communication can therefore not assume any college-level education.

According to 2022 US Census data, a bit over one third (37%) of Americans 25 and older had at least a bachelor's degree. The same fraction (37%) had only a high school diploma or less — and seeing as an additional 15% had completed some college but didn't finish a degree, there are actually more American adults with no degree than bachelor's degree holders. (The remaining ~10% are people with an associate's degree as their highest qualification)

Lest you think this is a uniquely American phenomenon, things don't look that different in Austria, where I live now. Here, universities are free and the government offers generous financial support to students that would make Americans drool. You'd think that would mean more grads. But actually, fewer adults in Austria have university degrees than their American counterparts: Just 20.4 % of Austrian 25- to 64-year-olds have a university degree or equivalent (that's up from only 4.5% forty years ago). 17% completed only an equivalent of middle school and high school coursework that would not qualify you for college.

2. General audience communication is not a written article or book.

Most Americans are not good readers. Something like 1 in 5 adult Americans are what you might call functionally illiterate.^[3] Most of these people could write their name or short, simple sentences. But at this level, reading is so hard that text is a barrier to understanding, not a help: The U.S. National Center for Education Statistics describes people in this range as lacking "literacy skills sufficient to complete tasks that require comparing and contrasting information, paraphrasing, or making low-level inferences." The National Literacy Institute calls this level functional illiteracy, the NCES calls it low English literacy, but whatever word you want to use about 20% of Americans really struggle to read and write.

And even among people who do not struggle this much with reading, the written word is still a serious challenge: more than half of Americans read at a 6th grade (12 year-old) level or below.

For context, I ran several of my articles for various magazines through a grade-level estimator and they come out college-level. Oof.

3. General audience communication uses language a 12 year-old could understand.

If most American adults are not comfortable with written language more complex than a 12 year-old can handle, then general audience communication for Americans should stick to that level.

This is especially true for titles and headlines: a study of more than 30,000 A/B tests using Washington Post and Upworthy content found that readers consistently clicked more on headlines written in simple language. Professional writers did not have the same simplicity preference when reading news and were much better at remembering things about content they'd read before than general readers. This suggests that word nerds experience writing differently than their readers. Perhaps most damning, journalists were unable to reliably predict which headline their readers would prefer.

4. General audience communication is video, probably short-ish form (up to 10min).

Most American adults struggle to read complex texts. So it should not be surprising that many now prefer to watch their news rather than read or listen to it: Data from Pew shows that 44% of US Adults prefer video news, while 37% prefer reading and 19% prefer listening. 1 in 5 Americans now regularly get their news on TikTok. Of course, TikTok's demographics skew young: 63% of American adults under 30 year old use TikTok and 43% say they regularly get their news there. Those numbers drop to 44% using and 25% getting news on TikTok among the 30-49 crowd regularly get news on TikTok, and fall off from there in older age bands.

But TikTok is not the only video shop in town. YouTube is actually quite popular among all age groups: 95% of under-30s, 92% of 30-49 year-olds, 85% of 50-64 year-olds, and even 64% of the 65+ crowd uses YouTube. YouTube shorts have 4 billion unique monthly users and receive 70 billion views per day.

Then there's Instagram and Facebook, which are linked together: short-form Reel videos posted to Instagram can also show up on Facebook. They're more popular than TikTok, especially among older demographics (57% of 65+ American adults use Facebook and 19% use Instagram). My mom is in the 65+ crowd and clearly loves short-form video on Instagram.

I'd need to do more research to get a sense of whether very short-form (less than 1 minute) video gets more traction than short-ish form (a few minutes) or long-form video. But my suspicion is that the sweet spot will be something that's on the longer side of what's allowed for a TikTok or Reel — 10 minutes for TikTok, 3 minutes for Reels (Instagram also allows longer videos, but it's kind of a work-around thing). Being short enough to stay on those platforms seems important, though, if you truly care about reaching a "general audience."

5. General audience science communication uses pretty visuals

People like pictures. Pretty visuals get attention. General audience science communication needs to use pretty pictures.

I feel like this is so obvious that it almost feels weird to try and prove it. But if you want numbers, the email marketing platform Vero reported that emails with images prompted almost twice as many clicks as those without (this was based on an analysis of 5000 email marketing campaigns). And if you don't want to trust Vero, trust the UK's National Literacy Trust: over the past 13 years, the proportion of kids who read comics at least once a month has stayed relatively constant around 40%, while the proportion who say they enjoy reading in their spare time has steadily declined from 51.4% in 2005 to 32.7% today. And kids that read comics were more likely to "read something" (comics, one assumes) than kids who didn't. Anecdotally, my mom was an elementary school teacher before retiring and she often had several kids who she could literally only motivate to read with comics — and most kids preferred comics if they were allowed to read them instead of novels.

I've done a little bit of testing with my posts on Bluesky: sometimes I'll post exactly the same post with and without a picture. The picture is never anything special — just a free image from Wikimedia Commons or something. But it almost always improves engagement.

6. General audience science communication piggybacks on popular or trending topics outside of science

General audience communication has to reach people. Which means that it can't just exist and hope people will show up — it needs to go where the people already are, both in terms of platform (as discussed above) and mindset. That means piggybacking on other things people are talking about and interested in already: finding a science angle on a hot political debate, or finding ways to "sneak" science into topics like sports, gaming, pets, beauty, self-improvement, music, activism, health, celebrity gossip, and more. Almost anything can be a science story.

There's some evidence suggesting that connecting science to other areas of life might be especially important for reaching groups historically kept at the margins of science. A survey by the Kavli foundation found that connection — defined as a common thread between one's personal life and scientific interests — was especially important in determining whether Black and Hispanic adults are willing to engage in science activities like museum events. That's not the same thing as science content, but I think it's worth keeping in mind.

7. General audience science communication would, somehow, reach the political right.

Look, I'm not going to claim to know how to do this. But if you're only talking to left-leaning folks, you're not reaching a truly general audience. Roughly half of Americans voted for Donald Trump. Something like 60% of Americans disapproved of RFK Jr.'s job as HHS secretary in October 2025... which means 40% either approved or didn't care. Most of the disapproval came from, surprise surprise, Democrats and independents. I suspect that consciously avoiding anything obviously left-coded, and trying to appear apolitical, mostly wrapping science up in other topics (sports, pets, etc.) that draw politically diverse audiences, and engaging on a more emotional than cerebral level might help with this. But as I said. No idea. People genuinely devote entire research careers to trying to understand polarization...

General audience scicomm looks like... Kurzgesagt?

I'm guess that if you, like me, are a college-educated left-leaning person who likes reading pop science, that last section was maybe a bit of a shock. People tend to assume that others think and feel very much like themselves. So if you're a college-educated science nerd and live and work in a bubble of other college-educated science nerds, it might be a bit jarring to realize that your experience is actually quite far from the norm. You are not the general audience. You like Quanta because you are its target demographic.

After I typed out that section above, I was honestly a bit demoralized. I'm a science writer. I care about telling stories about science — in part because I just genuinely love learning things for myself, but also because I want to reach people. And the reaching people part is actually my job, the bit I get paid for. So if I'm not reaching people, what the hell am I doing?

I've been aware of this for a while, actually. In the past, when I was scrambling to just get into science writing at all, I basically just laughed off the problem. I'd say stuff like: "Oh, it's okay. I'm just writing entertainment for nerds. Nothing wrong with that." But I'm craving more impact these days. The questions scientists ask me at conferences about trust in science, misinformation, and reaching general audiences are good ones. Reaching more people does matter. And maybe it's lame to admit at this point in history, but I am in fact a bleeding heart liberal who believes in democracy and the value of every human being, and I think that scientific truth gives people the best foundation to make decisions and advocate for policies that will improve their lives.

So I'm starting to think more about audience — who I want to reach with what content and why. And that includes thinking about starting to do video. Admittedly, reading through my list of criteria for "general audience" media at first made me think I'd rather chew off my arm than actually produce anything like that. But then I realized that I already know of at least one group that's already kind of done it. And I love their work.

The YouTube channel Kurzgesagt satisfies basically every criterion above except maybe the political one. Their team packages knowledge into visually stunning videos — and now, increasingly, shorts — delivered in plain language. I ran a few Kurzgesagt transcripts through a grade-level estimator and they do indeed come out around a middle-school level (8th or 9th grade, close enough). They aggressively avoid jargon in a way that's only really clear if you're watching a video that you actually have some subject matter expertise in. And they make videos on a wide range of topics, many of which aren't "scientific" at all on the surface, like friendship or exercise. The science sneaks in.

Without a way to pull up Kurzgesagt's audience demographics, I can't say how general their audience truly is. Their main channel has 25 million followers — but Veritasium has 20 million, and as I detailed above I am deeply skeptical that Veritasium's audience is general audience. So it's hard to say if my suspicion that Kurzgesagt truly reaches widely is right. I do have a hunch that they're not reaching older folks, and their "sciencey" branding might scare off people who don't think of themselves as "sciencey" even if the content would connect in a vacuum. I personally suspect identity and identification with science matters more than one might think when it comes to who watches a science video.

But still, they show that it's possible to do deeply-researched, curiosity-sparking science communication that's far more accessible than magazine writing is without it being anything like the clickbaity, sensationalist slop that the magazine-reading might fear from general audience scicomm.

In fact, pretty every college-educated, professional class, left-leaning, Quanta-reading geeky nerd I know loves Kurzgesagt. And they're a tough crowd.

I want to do more than write science entertainment

Let’s talk about science and scicomm in a crazy world

Reviewer, too | Elise CuttsElise Cutts

A final note: I realized 3/4 of the way through writing this that I'd forgotten to include links to my sources. Dang. Magazine writing habits die hard. If people actually read this thing I'll go over it again to add links.

EDIT: People did in fact read this, so I have gone back and added links and some extra context in a few places incl. footnotes.

Footnotes

[1]: A note on the data source. These numbers come from Hurst Media Company, an advertiser, not directly from New Scientists' sites. Hurst cites New Scientist as the souce and claims the data is from 2021.

[2]: I make this claim based on personal experience; science journalism is still very text-oriented and text media dominates both training programs and the field's awards (though there is some movement there). If you want examples, I suggest checking out the former winners of and work honored by/produced through some of the field's top fellowships and honors: the AAAS Mass Media Fellowship, the Kavli Awards, and the National Academies awards are some examples.

[3]: The data referenced by the NCES comes from the Program for the International Assessment of Adult Competencies (PIAAC), a large survey of adults' abilities around the world. The US cohort in the data was surveyed in 2011–12 and 2013–14. About 20% of people in that data set had "low English literacy" meaning they'd struggle to understand a tex. The National Literacy Institute claims that basically the same number of Americans are "functionally illterate," but this wording is controversial and really depends on your definition. Personally, I am okay with it because if someone can't understand a text that requires "comparing and contrasting information, paraphrasing, or making low-level inferences," that is not a very "functional" level of literacy. But the NLI has a financial incentive to make literacy sound as bad as possible, because they are a professional development organization for literacy educators + consultants. They're not a neutral observer on this subject. So make of that what you will.

Thanks for reading

• Subscribe, if you haven't already!
• Share this post on Bluesky, Twitter/X, LinkedIn, Facebook, or wherever else you hang out online.
• Become a patron for the price of 1 cappuccino per month
• Drop a few bucks in my tip jar
• Send recommendations for research to feature in my monthly paper roundups to elise@reviewertoo.com with the subject line "Paper Roundup Recommendation"
• Tell me about your research for a Q&A post (email enquiries to elise@reviewertoo.com)
• Follow me on Bluesky
• Spread the word!