How much genuine information content carries per unit of length: the ratio of substance to words, the axis our Cyberfluff substance work scores.
Information Density is how much genuine information a piece of content carries per unit of length: the ratio of substance to words. A dense page states specific facts, data, and findings with little padding; a sparse one uses many words to say little, however fluent it reads.
The measure has roots in information theory, where the information in a message relates to its entropy and, for a single object, to its Kolmogorov complexity, the length of the shortest description that reproduces it. Content that compresses well is often content that repeats itself; content that resists compression tends to carry more distinct information. This is the same axis semantic compression works along.
We built a model to score this directly. Our Cyberfluff work on content substance classification separates genuine substance from eloquent fluff, flagging text that pads word count without adding information (detailed in this write-up). Density matters for AI visibility because retrieval favours it: a system grounding an answer pulls the passages that carry the most relevant information for its limited word budget, so a dense, specific page earns its place in a grounding chunk where a padded one is passed over. It connects closely to information gain, which measures the new information a page adds on top of what is already known.
