Definition
Liquid content is content that changes shape to fit the reader and the screen they are on. One story becomes a vertical video, a newsletter, a social post, a podcast segment or a chat answer, each built at the moment someone asks for it. Publishers stop making articles and start making knowledge that travels.
Origin
Coca-Cola coined it in 2011. Journalism took it over in 2025. The Reuters Institute made it the word of 2026.
In 2011 Coca-Cola announced a plan it called "liquid and linked": ideas so strong they spill across every channel on their own. Good slogan. Different subject.
The meaning publishers use today arrived in three steps. In February 2025, INMA asked its members why the article should still be the unit of news. The journalist Pete Pachal gave it an edge: liquid content is what happens when AI remixes a story into other formats. Then in January 2026, Nic Newman wrote it into the Reuters Institute's annual predictions report and supplied the definition everyone now quotes. Content that "adapts in real time based on the viewer's context, location, time, or interaction."
The report's larger point is the one that matters. Media companies are moving "away from authoring articles" and toward "structured knowledge that can flow into different formats, surfaces and interfaces." The article is a format. The knowledge inside it is the asset.
Two things this page is not about. DNN Software sells a CMS feature named Liquid Content. Shopify's template language is named Liquid. Both are fine products. Neither is this.
Source: Journalism, media and technology trends and predictions 2026, Nic Newman, Reuters Institute for the Study of Journalism, January 12, 2026.
Comparison
Static content is finished before anyone reads it. Liquid content is finished the moment someone reads it. Everything else follows from that.
| Static content | Liquid content | |
|---|---|---|
| What you publish | A finished article | The facts, quotes and sources inside it |
| When the format is set | Before you press publish | When the reader shows up |
| Adding a channel | Another production job | Another output of the same source |
| Cost of the fifth format | About the same as the first | Close to zero |
| Who it is built for | A segment, at best | The person asking |
| Inside an AI answer | The model has to summarize your prose | The facts are already in shape |
Multimodal means you made three versions. Liquid means the version gets made when it is asked for.
People use the two words as if they mean the same thing. They don't.
A multimodal newsroom publishes the article, cuts the video and records the podcast. Three formats, each fixed on the day it ships. That is where most publishers stand today.
A liquid newsroom holds the story as parts. A reader on a train at 7am gets a forty-second vertical cut. The same story, requested inside a chat assistant, comes back as five sentences with sources. Neither version existed until someone asked.
Multimodal is a distribution decision. Liquid is a data decision. You can have the first without the second, and most publishers do.
How it works
Five steps. Two are hard. Most teams skip the hard ones and blame the software.
Facts, quotes, sources, dates, names, places, pictures, audio. Not paragraphs. A paragraph is a format decision you have already made.
What it is. Who it names. Where and when it happened. Which source it came from. This is the layer that lets a machine rebuild the story instead of guessing at it.
Vertical video, newsletter, social card, audio, search page, chat answer. Each gets a length, a tone, an aspect ratio and a rule about credit.
Make the formats people ask for. Skip the ones they don't. This is the step that pays for the other four.
Nothing ships without a human OK. One wrong fact in a new format costs more than a year of production savings.
Where Nota fits
Studio does steps one through four. Your team does step five. The first clip is ready in under five minutes.
You drop in an article, a broadcast, a recording or a transcript. Studio takes it apart, labels the pieces, and returns every format you have asked for, in your voice, with your style guide applied. Video clips, social posts, newsletters, summaries, SEO copy and audio, from one upload.
Step five stays with you. Every output waits as a draft until someone on your team approves it. Nothing goes out on its own.
The engine underneath is Polaris, a language model we built for media and trained on licensed editorial data. It is the reason the outputs read as if your newsroom wrote them.
Cox Media Group, a top five U.S. broadcast group, used it to lift average clicks per URL from search by 613%. Lee Enterprises grew social interactions by 1000% from the same articles it was already filing. Both numbers, and the rest, are in the case studies.
FAQs
Content that changes shape to fit the person reading it and the screen they are on. The publisher stores the story as structured parts and assembles the format when it is requested, instead of publishing one finished article and hoping it travels.
Nobody owns it. Coca-Cola used "liquid and linked" in 2011 for advertising. INMA brought it to publishing in February 2025. The journalist Pete Pachal tied it to AI remixing stories. Nic Newman put it in the Reuters Institute's January 2026 predictions report, which is the definition most people now cite.
No. Repurposing takes a finished piece and makes another finished piece, one job at a time. Liquid content has no finished piece. The source is structured data, and the article is one rendering of it, no more canonical than the video.
Multimodal means the story exists in several formats, each fixed when it shipped. Liquid means the format is chosen at the moment of consumption. One is a distribution choice. The other is a data choice.
It does when the voice lives in the prose. Put it in the spec instead: a style guide the system reads, sample writing from your own staff, and a tone rule per surface. Then the voice survives every format change. Nota Studio does this with a tool called Tone Builder.
Three layers. Something that pulls structure out of source media, a model tuned to your voice, and a spec for each surface you serve. Nota Studio does all three in one product. The alternative is a headless CMS, a clipping tool, a scheduler and a newsletter platform, which is four contracts and four invoices.
Per story, not per post. Count formats produced per story, minutes from source to first output, and reach and revenue per story across every surface. Add one new number: how often AI assistants cite you. Per-post metrics hide the gain, because the gain is one story doing ten jobs.
Yes. Any organization that makes content on a schedule has the same problem: one asset, many surfaces, rising cost per surface. Sports properties, broadcasters, brand studios, creators and B2B marketing teams run the same model. The output specs are the one thing that changes.
Book a demo. Bring one article or one broadcast. We run it through Studio while you watch and hand you every format it produces.