Do some PIs who publish in Nature, Cell or Science collaborate with professional designers?

The thing is that I have never seen it mentioned in their work. Is it a service offered by the journals? Or do the PIs pay for these services and outsource them?

  • 18
    Not a duplicate, but the reasons why academics' personal websites tend to look ugly is addressed here: academia.stackexchange.com/questions/69275/…
    – mhwombat
    Jul 10, 2017 at 5:46
  • 3
    related: academia.stackexchange.com/questions/38619/…
    – Bitwise
    Jul 10, 2017 at 5:47
  • 9
    Could you add an example of a figure you find particularly well designed?
    – Cape Code
    Jul 10, 2017 at 10:53
  • 21
    The question in the title (about researchers’ design skills in papers vs on websites) is quite different from the question in the body (about who designs figures in published papers). This is causing major confusion in the answers — some address one, some address the other. Could you edit so that the title and body match, and it’s clear what the actual question is?
    – PLL
    Jul 10, 2017 at 12:53
  • 4
    Good design, like beauty, is in the eye of the beholder - or in the case of web sites, the poor chaps like me out here on the other side of the browser trying to use those at best marginally usable "well designed" web sites. I mean web sites that force you to use particular browsers, or particular versions of that browser, or which present marquee images or sliding popups running by too fast to read, or which insists on using particular colors, or... In my experience, most academics have good, usable web pages, while most web designers would serve society better as ditch diggers.
    – jamesqf
    Jul 10, 2017 at 16:22

11 Answers 11


There are a few reasons not to waste time with web design:

  • academic web pages are read mostly by other academics who are looking for information;
  • the information a professor needs to put on pages is about group members, papers published and talks, courses, books and none of these seems to require sophisticated web design skills;
  • learning proper web design skills is time taken away from research. Most professors don't have that time. But, it can happen that graduate students or postdocs are capable of making nice web pages, so the professor asks them to do that.

The beautiful figures that appear in journals, are most often the work of postdocs and graduate students. Again, most professors won't spend time making them, unless they can't rely on anyone else. It was quite common back in the day to have someone else from within the university or research institute make figures and typeset, but now, there are plenty of great software packages one can use to make figures.

  • 4
    @0x90 Maybe you want to give examples for what you assume is the work of professional designers? There are many programs around that can make really, really nice graphics if you know how...
    – Dirk
    Jul 10, 2017 at 6:36
  • 35
    @0x90 Most of the links you gave are graphs of some functions/sets/... There are tons of functions to plot these, so someone familiar with Maple, Matlab or similar software should have it easy making them (I even know a professor who programs such things purely in Latex...). If being familiar with a software that can be used to make nice graphics automatically turns you into a "professional designer" can of course be argued, but I don't see anything in your links that would require more than "program, plot these points in the following way:".
    – Dirk
    Jul 10, 2017 at 6:49
  • 8
    @0x90 Some of the figures you cite are an afternoon's work with software like MATLAB, ORIGIN, PowerPoint. For some others, I don't know the software, but I would find out if I had to produce something close to that. You should never underestimate a graduate student who wants to publish.
    – user21264
    Jul 10, 2017 at 6:52
  • 5
    @0x90 That might be true and not that surprising, but every time I asked people at conferences how they made figures, they sent me to specific tools. The people I worked with always made their own figures, and yes, some of our papers made it to Nature and Science.
    – user21264
    Jul 10, 2017 at 7:03
  • 13
    @0x90 Reply to "No SW package can turn a natural science student into an artist, unless he's already a professorial designer." Actually it can. But, a researcher while publishing needs innovation and high-level thinking to make the paper look beautiful and quite well received by its readers. SW package actually can aid to that innovative thoughts. I am no good designer of web pages, actually, I hate it. But, still I spend at least 15min to 1 hr for a single figure in my article. Don't always be skeptical about all these stuffs. Paper and web pages are highly uncorrelated with each other.
    – Coder
    Jul 10, 2017 at 7:08

What is good web design depends on the purpose. I am sure Wells Fargo Bank and Google both have access to expert web page designers, but I find the current versions of news.google.com and the Wells Fargo on-line banking web pages extremely frustrating.

The problem is they both use lots of white space and severely limit the information that is available in one view. I am used to reading textbooks and academic papers. I like information-dense web pages, where I can see a lot of words and numbers in one view. That preference may be commoner among academics, who tend to be very good at reading and processing information, than in the general population.

Some, at least, of the academic web pages that don't conform to current web design fashion are well designed for people like me.

  • Aren't you ever intimidated by a wall of text, especially when you need to read it in order to solve a more important problem?
    – Ooker
    Jul 10, 2017 at 17:52
  • 12
    @Ooker, imagine a printed dictionary with one definition per page. In a nice stylish font, well-chosen to integrate with the word being defined (so the words "harsh" and "soothing" and "water" are all in different fonts and colors). In nice large letters, center-aligned. Now imagine trying to actually use it to clear up a term from your textbook.
    – Wildcard
    Jul 10, 2017 at 22:31
  • 12
    ...and yet, actual printed dictionaries solve the "wall of text" problem just fine. Guide words at the top of the page. Entry words in bold, slightly less indented than definition continuation lines. Example sentences in italics. The result is an extremely information-dense page that is very easy to scan through to acquire useful information (assuming you know the alphabet well enough).
    – Wildcard
    Jul 10, 2017 at 22:33
  • 6
    I'm not sure how your preferences about Google News and your online banking system relate to this question about the quality of artwork in journals... Jul 11, 2017 at 12:31

Researchers don't have much interest in and time for design (exception: Edward Tufte), otherwise they would have become designers. Luckily, they can rely on the work of others for "beautiful figures" - division of labour and all.

The beautiful figures are

  • generated by preset LaTeX packages like Tikz/PGF
  • generated by preset R libraries like ggplot2
  • preset designs created by publishers.

By contrast, many universities only provide some webspace or a CMS in corporate design for researcher's personal websites.

  • 5
    Of course LaTeX was designed by a researcher (Knuth).. precisely because tools to typeset equations didn't yet exist, and he needed something. Similarly, I had to design my own graphing software in the 80's to plot the output of my instruments (from a 6502 based "computer" with 32 kB or RAM). But those days are mostly in the past now.
    – Floris
    Jul 10, 2017 at 14:50
  • 2
    @Floris Not really important here, but "La"tex comes from Lamport, not Knuth.
    – deviantfan
    Jul 10, 2017 at 16:11
  • @deviantfan And TikZ/pgf (it's to day most powerful drawing tool) even by someone else, namely Till Tantau :-)
    – yo'
    Jul 10, 2017 at 16:15
  • 3
    To add to your list: Python packages like matplotlib (my preference) and bokeh.
    – Ébe Isaac
    Jul 16, 2017 at 3:49

Successful researchers react to incentives—and there's little incentive for webpages to look beautiful. I doubt a better webpage design even attracts better graduate students, and I'm sure it won't help you get tenure or a grant.

Conversely, there are incentives to communicate clearly to reviewers and readers. To whatever extent authors prepare well-designed figures, it's to write clearly to reviewers. Some figures can make or break a paper.

A few journals also employ professional designers/typesetters (like journals you mention apparently, see https://academia.stackexchange.com/a/92944/8966), but that's the exception, not the rule.

  • 3
    "I doubt a better webpage design even attracts better graduate students," Well a pretty web page probably wont, but one with useful information certainly will. Actually lots of people I know respond well to a web page that also includes some hobbies and interests, if you want to get a position under someone its good to know they are human. But lots of things that many web designers consider 'pretty' are actually just annoying and take too long to load.
    – Clumsy cat
    Jul 11, 2017 at 12:42
  • 2
    "one with useful information certainly will" agreed, but that can still be ugly (or plain). Jul 11, 2017 at 16:36

These are different skills: Web-design vs Figure Design.

  • Web-design is closely related to layout of whole pages. It needs to do things like margin sizes, and text-widths. How much white-space around a figure. These kinds of problems. In many fields laying out papers is a solved problem - use LaTeX.
  • Web-design is actually more difficult than paper layout, since it needs to handle different screen-sizes. And all the other things like menus openning up, and keeping a nav-bar in place (or not) etc. There is a lot of scope to handle and a lot of places to mess up.
  • Web-design has very little shared skill with figure design.
  • Even assuming one has both the aesthetic sense to know what looks good in both, the ability to actually pull it off does not leverage one skill for another.
  • Basic web-design is HTML+CSS, plus maybe some form of templating language. Figure creation normally involves either a plotting tool eg Matlab, ggplot2; or a vector graphics tool, eg Tikz or Inkscape.
  • The time spent also varies. I regularly spend 4-8 hours on a single figure. I might spend 4-8 hours every 3 months on the appearance of my whole website.
  • +1, though laying out some figures is not a problem solved by LaTeX—and it overlaps with web design. Jul 11, 2017 at 16:41
  • 1
    I'm not sure what you mean? Layout specifically is positioning figures and text on a page. That is done up-to occational tweaking needed in latex. And the occational tweaking needed is things like setting various penalities for too many floats on a page (nothing like webdesign). Jul 12, 2017 at 0:42
  • 1
    I meant layout within the figures. Alignment, proximity, consistency and contrast (the basic design concepts I know) all apply in both cases. Yes, LaTeX best practices typically overlap. Jul 12, 2017 at 18:00

Researchers at large experiments or laboratories (e.g., CERN or Fermilab) tend to use the same templates, scripts, etc as those who wrote papers before them. Some experiments even have their own guidelines about how any graphics published or presented must look, on top of any requirements from the journal.

As such, the same set of scripts and templates tend to get passed around, to ensure all of those guidelines are met. This also ensures a consistent style is used for all papers from that lab / experiment / group.

Those scripts and templates are developed / passed around largely by grad students and postdocs.

There may be similar templates available for faculty webpages, but they won't be as pretty looking because the need just isn't there. In a journal or conference presentation, you are limited by number of pages or time; therefore, you need to make sure every word is needed, and every figure is conveying as much information as it can while being as clear as it can. You may pay some amount for each figure you put in the paper, so you make sure the ones you have are stellar. There is no such limitation on faculty websites, so there is no incentive to make them look as nice.


Professors at my University hire undergraduate students (me included) to format their papers as well as other small tasks. Nobody has ever asked me to do the same to their website. My guess is that they really do not care as long as the information on their website is correct and (sometimes) up to date.

I generally use LaTeX because I like the programming feel and you can make the paper and figures look however you want it to.


Figures in a paper communicate information about the results. Graphic design elements don't.


Professors are well versed with using LaTeX, R in their teachings/ lectures hence their publications viz. Papers, presentations, etc. are so beautiful in terms of figures and presentation.

However it is not the case with their websites because it involves a lot of time to implement various web designing techniques which they are constantly running out of. Also usually the different Professors follow a template given by University to keep the whole website uniform.


Papers, and where they get accepted are regarded by many (especially the publishing business, lol) to be the measure of success in academia. Therefore it is not strange if more effort is made to try get your papers into good journals. We humans as well as many animals are easily mesmerized by shiny and pretty things, so the papers have higher chance of being accepted into good journals if they are shiny and pretty.


Without disclosing more information. I can say that I know a research group which published a paper in either science or nature and they used the help of a graphic designer for 2 of their figures.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .