What is a graphical abstract?

What journals ask for

Requirements differ by publisher, so the author guide of the target journal is the first stop. Elsevier's guidance is typical in spirit. It asks for an image of at least 1328 by 531 pixels, width by height, at a minimum of 300 dpi, keeping that proportion if the file is larger, in TIFF, EPS, PDF or Office formats. It recommends Times, Arial, Courier or Symbol at a size that survives reduction, a clear start and end that reads top to bottom or left to right, and no additional text, outline or synopsis beyond the image itself. It also asks authors not to label the image as a graphical abstract.

How a good one is built

Most successful graphical abstracts follow the logic of the paper in miniature: what was studied, what was done, what was found. That usually means three panels or zones joined by arrows or simple ordering, with the finding given the most space. Icons stand in for participants, samples or settings; a small chart or a single number carries the result. Colour is used to link the exposure and its effect across panels rather than to decorate. The test is whether a colleague outside the field, shown the image for five seconds, can say what the paper found. If they describe the method instead, the result panel is too small.

When it matters, and when to skip it

It matters most where the journal displays it prominently and where the audience is broad, such as multidisciplinary journals or fields with heavy use of social media. It also helps papers whose finding is spatial or comparative and so is naturally visual. It matters less for a methods paper whose contribution is a technique with no single result, or for a qualitative study whose findings are themes that resist a picture. Where a journal lists it as optional, a weak image is worse than none, because it becomes the paper's face in every listing and signals confusion rather than clarity.

Common mistakes

Reusing a figure from the paper unchanged is the most common. Figures are built to be read with axes, legends and a caption; shrunk to a listing thumbnail they become noise. Too much text is next, often a paragraph pasted from the abstract, which Elsevier's guidance rules out. Small fonts that vanish when the image is reduced are third. Missing accessibility is fourth: the W3C guidance on complex images says images such as charts and diagrams need a text alternative, a short description plus a longer one that conveys the information, so supply alt text when the journal allows it. Last, check general figure rules too: PLOS ONE, for example, gives figure widths of 789 to 2250 pixels at 300 dpi.

Where it shows up in the poster builder

The poster builder does not make graphical abstracts. Its two poster canvases are A0 portrait and A0 landscape, neither of which matches a journal's graphical abstract proportions, and a PNG export captures the whole active page rather than a single figure. What it does offer is the same design idea at board scale: the hero figure layouts give one chart roughly a third of the canvas, 36 percent on the portrait board and 28 percent on the landscape one, and the billboard layouts lead with the finding in very large type. A figure that works there is a good starting sketch for a separate graphical abstract drawn at the journal's size.

Graphical abstract or visual abstract

The terms are often used interchangeably, but some publishers mean different things. A graphical abstract in the Elsevier sense is a designed image with minimal text. Some publishers use visual abstract for a template style card that lists the design, population and main result in short text blocks with icons. Both serve the same goal. The journal's own term tells you which it wants, and its examples page is the quickest way to see the expected level of detail before any drawing begins.

Questions people ask