1. A correction, before anything else
An earlier version of this article said we had analysed 10,000 highly shared infographics and reported what we found: that 78% of viral graphics led with a metric, that stat-first headlines lifted click-through by 42%, that whitespace improved recall by 65%. We ran no such study. Those numbers were invented and they have been removed. We are leaving the article at the same address rather than quietly deleting it, because a page that made things up should say so where the making-up happened.
2. Why every statistic in this genre should be treated as marketing
Search for evidence that infographics spread and you meet the same handful of figures on every page: infographics increase web traffic by 12%, they generate 3x more social engagement, they are 30 times more likely to be read than an article, they account for 41.5% of all engagement with brand content. Follow any of them back and the trail ends at a vendor survey, a decade-old blog post, or nothing at all. They are quoted by the companies selling the format — we sell it too, which is exactly why we are not going to add another number to the pile. The useful position is that nobody has a reliable measurement of this, and anyone offering you one is selling something.
3. Distribution decides it, and almost nobody writes about that part
The infographics that travel are the ones that were pitched. They were sent to editors, submitted to sites that publish them, posted by someone with an audience, or attached to a piece of original data that a journalist wanted. That is the variable, and it is the one the design listicles skip because it is unglamorous and cannot be fixed with a colour palette. If you are producing one and there is no plan for who will see it beyond your own followers, the design decisions below will not rescue it.
4. One idea, and one number worth repeating
The graphics that get shared are almost always reducible to a single sentence somebody can repeat without looking at them again. That is a content decision, not a design one, and it is made before anything is laid out. A sheet carrying nine findings is a sheet nobody can quote. Pick the one that would make somebody stop and say the number out loud, put it at the top at a size that is almost uncomfortable, and let the rest support it.
5. Reading order: the one layout claim with research behind it
Eye-tracking work on web reading has consistently found that people scan in rough F and Z shapes rather than reading top to bottom, which means the top-left of a graphic is the most valuable space on it and the bottom-right is where attention has already gone. Put the finding top-left, anchor the middle with the chart that supports it, and leave the source line and any call to action at the bottom. This is genuinely well-established, unlike most of what gets asserted about infographic design.
6. Build it for the feed it is going into
A graphic composed at the wrong proportion arrives cropped, and a cropped headline is a wasted one. Portrait 4:5 suits Instagram and LinkedIn, square 1:1 travels across several platforms at once, 16:9 suits X and slides, 9:16 suits stories, and 3:4 suits Pinterest and long scrolling pages. Choose before you compose rather than resizing afterwards. Our social media infographic guide goes platform by platform.
7. Are infographics still worth making?
For the right material, yes — and the honest qualifier matters. They work when the point is comparative, numerical or sequential, when the reader has not yet committed to reading anything, and when the underlying finding is genuinely interesting. They do nothing for a point nobody wanted, and no amount of design turns a thin idea into a shared one. That is a less exciting answer than twenty-one secrets, and it is the one that will save you the most time.