The web has no primitive for content that updates its own substance.
We're building it. LiquiChart is living content infrastructure: prose that updates its own substance, claims that know when they were last true, and your own polls and charts feeding them, and the lift measured in your own analytics.
A chart gets embedded in a blog post. The post says one thing.
The data, slowly, says another.
The chart updates. The prose around it doesn't. The “73%” in paragraph two becomes 68% on the chart underneath, and nobody notices for eighteen months. The article keeps ranking. The number keeps drifting.
Multiply this by every post in your archive, every benchmark you've cited, every statistic you've quoted from a source that has since revised itself. That's the gap.
LiquiChart exists to close it.
Why now.
Three forces converged. The web's most valuable content is now evergreen and data-dense: benchmark reports, industry surveys, head-to-head comparisons, the long-tail of explainers that earn traffic for years.
LLMs ingest that content and treat its numbers as facts. A stale claim in your 2022 post becomes a stale claim cited by an AI answering a question in 2026.
And the cost of writing has gone to zero. The cost of being right has gone up: citable, current, trustworthy.
The publishers who win the next decade won't be the ones who publish more. They'll be the ones who publish something original, and keep it true.
Six things we'd build the product around even if no one paid us for it.
Numbers are content.
Every statistic, benchmark, percentage, and dollar figure in your writing is a claim, an assertion you can be wrong about. Treat them like content, not decoration.
A claim has a lifecycle.
Born when you publish. Fresh while the source agrees. Stale when the world has moved on. Fixed when you correct it. Most software treats numbers as static strings. They aren't.
Sources should be addressable.
If your post cites a number, that number should know its source, and the source should know it's being cited. When the source changes, every dependent claim should hear about it.
Prose should be reactive.
When the underlying number flips, the sentence around it should flip too. "leads by 3 points" becomes "trails by 1." Not just the digit: the qualifier, the verb, the framing.
Authors stay in the loop.
Automation that publishes without you is a liability. Every rewrite is proposed, diffed, and approved, or set to run autonomously on a per-block basis, your call.
Time is a moat.
Trend data that no one else can backdate, accumulating across periods, becomes the citation. A six-month-old benchmark is worth more than a brand-new one. Living content compounds.
Four pieces of missing infrastructure.
A claim graph
Every number in your content becomes an addressable node connected to its source. When the source changes, the graph propagates. It is the missing data structure of publishing.
A first-party data layer
Polls, sheets, monitored pages: the inputs to the graph. Your own audience and your own collection cadence, not someone else's licensed dataset.
A rewrite engine
When a claim shifts, the prose responds. Living Content blocks re-render with current numbers and language calibrated to current reality. Approval-gated, never silent.
A proof loop
Search Console and Analytics wired into every page you monitor, and experiments that tie a content change to a search outcome. The lift, measured in your own data, not our case study.
Help us build the next primitive
of publishing.
Try it on a post you already published. Tell us where it falls short. We'll listen.