Single-Source Publishing: The Song Remains the Same

This article was originally published as part of a series of articles on our LinkedIn page:

I fell under the spell of music when I was twelve years old. It was the 1980s, the Age of Synths (and hairspray – so. much. hairspray). But all the songs that are still with me today are essentially melodies; melodies that will work in every conceivable arrangement.

When I was young, there was a Cambrian explosion in studio technology. Synthesizers, samplers, and signal processors allowed musicians and producers to create breathtaking soundscapes and unheard effects. But when the needle is dropped on all those wonderful albums by Eurythmics, Prince, David Bowie, and Kate Bush four decades later, the first thing we hear is melodies. A truly great track works not only as a complex multitrack production, but also when played live by a four-piece band, or as a campfire song, with only a six-string and your friends singing along.

Which brings us to single-source publishing.

The equivalent of melodies in publishing is your story/content. It’s the words – articles, essays, white papers, and yes, manuals – that you typed on whatever device you had with you in that moment, on a plane, train, or pogo stick. Everything else is (post) production.

In the bygone analog age, content and presentation were one: handwritten and typed stories, pages with edges and smudges and scribbles. Then came “word processors” (an odd, steampunk-ish term evoking cogwheels and assembly lines full of lead letters) and layout/design applications. But still, most tools would handle one file format, where styles were mapped to presentation, and documents were printed or “rendered” as PDFs.

So what about the web? Microsoft first introduced an HTML export option in Word 97, and it was … something else. A whole cottage industry was built around cleaning up those HTML rat kings, and even today, exporting anything but PDFs from InDesign can lead to … interesting results. At the core, these are still tools for creating brochures and magazines, with everything else tacked on.

Modern single-source publishing, on the other hand, brings to writing what software developers have enjoyed for many years: separation of concerns. And what a beautiful thing that is. Well, actually, it’s three: content (what you want to say, i.e., the melody), presentation (how it should look or sound), and logic (what you want to do with the content, or allow your readers to do) can now be explored individually. And the more flexible such a system is, the more additional dimensions/vectors you can unlock.

Write your piece. Export it as HTML. Attach a beautiful classic style sheet with custom serif fonts for a specific use or audience. Add night mode for your app. Use JavaScript to generate a table of contents or a pagination feature. Extract important terms and build a linked index. Export EPUB and convert to AZW3 for Kindle. Go wild.

In the age of large language models and text-to-media models, authors and brands have more opportunities than ever to rethink the nature of content and which transformations could add depth and meaning (or sales, because someone’s got to pay the rent). LLMs can annotate content or create digests (akin to remixes in music production). Workflows can be set up to insert relevant illustrations or videos in documents automatically. If you’re too busy to apply structure to your plain-text document, have ChatGPT do it. Markdown (see my earlier posts in this series) is the perfect format for merging content from multiple sources without worrying about formats – they are applied “downstream”. In short: When done right, single-source publishing allows your content to shine, in every context, now and in the future.

To paraphrase the last panel of the last Calvin and Hobbes strip: “Let’s go exploring!”

Next week: Translation Workflows: From Sausage Making to Smart Collaboration

↻ 2025-10-24