23 July 2015

The last 10% of the poster should take more than 10% of your time


Over the last few weeks, I’ve been teaching the #SciFund poster class (compiled material here). It was a learning experience for me as well as the students, because I’ve never used Adobe Illustrator before. I made a poster based on some research I hadn’t presented yet. (The paper is in press; I’ll share the poster here once the paper is out of production and ready to read.)

I had something that I could have hung on a poster board at pretty much any conference in the world around the end of the second week. I felt the poster was maybe 90% of what I wanted it to be. But to get to the point where I thought it was almost 100% of what I wanted it to be, took two to three more weeks.

To be clear, I’m not talking about working continuous eight hours a day on a poster for a couple of weeks, but working on it briefly each day for a couple of weeks. You need to be able to step away from the work and look at it later with fresh eyes.

I spent that time adjusting the leading of the text. I made key numbers bigger. I proofread the text, refined it, and proofed it again (and someone will probably still find errors when I show it). I moving around a logo. I tried a different logo, realized it didn’t work, and switched it back to the first one. I tweaked the colours. I added lines, made them thinner, thicker, adjusted the line colours, then made them thinner again.

Those tiny little adjustments may not be something that an average viewer can easily identify when they read your poster, but the difference in the overall impression it leaves on a viewer is huge.



You have to leave yourself time to make all those tiny little adjustments. When you first start making a poster, improvements come fast, and the zones of “I could never hang that up” and “I could show it, but I wouldn’t be proud of it” are narrow.

The range of what is a passable poster is large. And somewhere in that big gray zone of “I like this and nobody would give me grief about it” is where you hit the point of diminishing returns. Sure, the poster is getting better, but not as much as when you started blocking it out.

To get to something that truly stands out, you have to keep working past that point of diminishing returns. You have to be willing to keep adjusting, coming back the next day, and adjusting again. The final improvements will come in at a crawl, not a sprint.

16 July 2015

Critique and makeover: Fine lace

Melissa WilsonSayres found this poster at this year’s Society for Molecular Biology and Evolution conference:


Yes, that’s a bra in the background. The authors made their poster into a boob joke.

The reaction to this on Twitter was not positive. Comments ranged from “shocked” to “mind boggling” to “poor choices all around” to “what where they thinking” to “speechless” to “great way to get people to miss the point of your poster.”

Okay. There is much to discuss here, and I’m not sure I’ll get to it all. There are a lot of things to talk about, I know, with everyday sexism and gender roles and appropriate behaviour in a professional setting and more.

I’m not one to talk about this. I am not about to cast the first stone on using knickers in design. I did, after all, write an entire post about lessons from lingerie. But this poster... that wasn’t what I meant.

Putting aside all that, this is a poorly designed poster, both conceptually and graphically. Let’s go back to a basic principle:

The design of a conference poster should be in the service of what the audience wants to know. Here, the design is in the service of a joke. And the joke doesn’t make any point about the scientific content.

For instance, Andrea Kirkwood noted:

The background is overemphasized to the detriment of the data.

Bad at Being Human had the most memorable critique:

The combination of the title style and the background comes across as the textual form of motorboarding.

I completely agree with this. It’s not just that there’s a bra as a background image, but the authors squeeze the title down into the cleavage to emphasize the breasts.

It’s not just the bra that’s the problem, either. There is a lot of room to improve in almost every aspect of the poster. Roberto Marquez added:

Plus 3D plots (why does anyone even consider...?)

The title is well below eye level, and will be blocked by anyone standing in front of the poster.

The typefaces seem to be chosen to be “feminine,” but they are hard to read. I cannot make out the text of the introduction in the photo, for example, even at the highest magnification. The swishy text might have made for a good heading, but is a horrible choice for the majority of the text.

Here’s what I would have done.

When I do poster makeover, I try to not to destroy the spirit of what the authors wanted. The title indicates that wanted something a little sexy. I am not opposed to making something sexy. Appealing to our sexual side can be a powerful way to communicate, if you can get past the inherent craziness and irrationality that comes along with sex appeal. See the “four organs of communication” in some of Randy Olson’s writing (summarized here).

In search of sexy, the authors went with the bra image. The problem is that the imagery they used is too literal and in your face. It violate the Sommese rule and treats the audience like morons.

I’ve talked before about the power of pastiche: imitating something that is a proven and recognizable template. You want to evoke lacy bras? Pick a well known and recognizable brand that one associates with lacy underthings.

I would go look at a Victoria’s Secret catalogue. (Purely for research!)

I’d look at the type used in the Victoria’s Secret logo and in the main pages. Bell MT is close to their main logo. I also notice that they use a mix of small caps and italics in their display text. Victoria’s Secret sometimes use a grotesque sans serif that I can’t identify. I tried a Franklin Gothic as a substitute.

I’d take a few representative pages from the catalogue to figure out what the colour palette might be. It would probably be pinks, pastel blue, creamy or pearly off-whites.

My version of this poster might be more like this (click to enlarge):



The key element is the lace of the title. The lace is now a very light, subtle pattern in the background. I took a large image of lace, and used the corrections in Microsoft Publisher to adjust the lightness and recolour it. To reinforce the lace theme, I kept something the authors had in their original poster: a little bow, which I put at the bottom instead of the side.

This is just my first draft, and certainly isn’t the only or best way to do it. The makeover shows that colour, type, and patterns alone can evoke a little bit of the sexiness implied by the title.

09 July 2015

Make this your working title for every poster

When you’re laying out your poster, instead of typing in the title you put in your abstract, put this:


“No one has to read this crap.”

Frame grab from this interview with Ed Yong. Ed has this posted above his desk as advice for freelancers, but the advice is equally appropriate for poster makers.

Nobody owes your their time at a conference. Nobody has to stop at your poster. Nobody has to talk to you.

Let that harsh realization guide your editing and design to make something that another person, who is not you, who is not invested in the project, wants to read.

02 July 2015

Critique: Galaxy quest

This week’s poster comes from Chelsea Sharon, and is shown with her kind permission. Click to enlarge!


Chelsea writes:

I feel like hardly anyone ever has useful design critiques for me, and I’ve certainly settled into a specific template.

I would try to calm down some of the type in the title and headings by removing outlines and underline. The title is particularly busy, with the title and the authors’ names and the authors’ institutions each getting a different combination of colour and outline: red with white outline, white with blue outline, blue with white outline. The outline of the authors’ names is making those two lines come perilously close together, touching in several places:


While the photo is reasonably subdued, it contributes to the overall feel of clutter. 

I’m not a fan of author photos on posters, but this one could be incorporated more smoothly. The picture breaks the symmetry of the title, and again contributes to the crowded feel. If you’re going to have a non-symmetrical layout, own it, by not centering the text. My revision might look something like this:


(You’ll have to image the background image of the radio telescope array.)

One other little point in the title bar is that Fabian
Walter’s name gets broken across two lines (as I have imitated in this paragraph). Where to put line breaks in text can be a tricky business, but it’s probably best not to break up an author’s first and last names if possible.

Travelling down to the main body of the poster, the underline on the headings could be removed.

The bullets are not bringing anything to the table. Bullets are effective for calling out short lists in larger blocks of text. Here, they fall victim to the Syndrome syndrome:



The amount of text is intimidating, and makes me wonder if some cutthroat editing might be in order.

Having most of the the main text all in the center of the poster helps provide a thread for the reader to follow through. Having the figures on the sides of that central column meas the reader has to weave back and forth between the text and the figures, but it isn’t too confusing. Maybe there could be few signals to link the text and the figure: some emphasis with bolding, and perhaps even a hint of colour?


Finally, those big heavy blue lines separating each block? Let’s see what happens if we remove those:

The poster seems to hold up fine without them.

Related posts

Mug shot

External links

Run ragged

25 June 2015

Link roundup for June 2015

The self-declared contender for the best poster. Not just academic poster, no.


Yes, that tap at the bottom? It works.


Coverage of this poster can be found here, here, and here. Probably other places, too. Hat tip to Jeffrey Bemis.

Back to science now, with a blog post about posters for Twitter.

This week on Twitter, I came across an image that was a hybrid between a science poster and an infographic. ... The simplicity of a tweetable poster makes it easy to highlight a project’s impact or identify solutions, and by sharing them on Twitter, the reach of these posters goes far beyond that of the traditional posters you find at conferences.

A reminder from Max Roser about why you should not use pie charts:


You will know the name from your font dropdown menus on your computer, all the way at the end: Zapf. An obituary of type designer Hermann Zapf is unexpectedly rich. Hat tip to Zach Seward and Amanda Krauss.

A poster! Hat tip to figshare.

A great look at science photographer Felise Frankel.

Frankel’s goal is to capture scientifically honest photographs that, in her words, “frankly, makes you want to look at it.” Since her first image ran on the cover of Science in 1992, her images have landed on some 30 journal covers.

While posters are generally static, there’s a lot to think about in this interview about turning nanoscale bioloical processes into movies:

In promoting the biomedical animations I should avoid overstating how accurately I have depicted the reality of the molecular world. It is vastly messier, random and crowded, and it’s physical nature is unimaginably alien to our normal perception of the world around us.

The Sociobiology blog has a post describing how to organize a fab small meeting. This particular meeting emphasized talks over posters, but I include it here anyway, since I am always hopeful conference organizers are lurking here.


18 June 2015

#SciFund poster class links

We’re in the thick of the #SciFund poster class now! One of the fun things for me about being involved is that we’re doing stuff that I haven’t covered in this blog.

In particular, Anthony Salvagno has written a lot about how to use Adobe Illustrator to make a poster. I had not used Illustrator before I started working on this class. It is powerful, but not simple. Anthony’s tips and suggestions are just the thing if you have been curious about using Illustrator for making posters.

You can download Illustrator and use the full version for free for 30 days.

I’m going to collect all the #SciFund poster class links here for archival purposes. As I post this, just two are up, but I will add the next three weeks as the become available.

#SciFund poster class links

Week 1: Focusing on message and getting started with Adobe Illustrator
Week 2: Developing a draft and building your wireframe with Illustrator
Week 3: Creating images and graphs
Week 4: Working with text
Week 5: The home stretch

11 June 2015

Critique: Shape perception

Today’s contribution comes from Arvid Herwig, and is shown with his permission. Click to enlarge!


My first reaction to this poster was incredibly positive. It’s an interesting mix of the bold and the restrained. The dark red bands surrounding each section are very large and visually dominant. Yet there are so few of them, and they are placed so precisely, that they don’t feel overwhelming. The muted background also helps calm the overall design.

To give an idea of how important those colour choices are, here’s a quick and dirty replacement of the brick red of the lines with a straight red, and the light gray grid with straight white:



Suddenly, the warmth is gone and you have a look that has all the appeal of a traffic sign.

The logos are corralled down in the corners, making them unobtrusive.

I’m impressed by how well the images and text fit within the section borders. Circles and triangles are not easy shapes to fit text or graphs into, but there is little wasted space here. The text and images follow the contours of the shapes very nicely.

The one problem I had was when I started to read it. I immediately read it the wrong way. I went from 1 across to 3, instead of from 1 down to 2. It seems that a major challenge for this poster is how to signal that it should be read in columns, not rows.

The first cue,spacing of the sections, tries to guide me. Section 1 and 2 are closer to each other than 1 and 3.

Part of the problem may be the labelling. The callout for “01 Introduction” is about the same distance, or maybe even a little closer, to “03 Methods” as it is to “02 Objective”.

Second, the labels for 01 and 03 point in the same direction. This provides a subtle cue that the two sections may be related to each other. I tried this alternate, making the headings for the first two sections match, attempting to strengthen the link between those two.


While I’m not sure this makeover works yet, I think the theory is sound. If you can get the reader to go down in the first instance, the rest falls into place.

Overall, some quite lovely work here.