Why does my novelty or overall score change between runs?

Every IP assessment runs a fresh search of a large, continuously updated body of patents and published research sourced from all non-patent literature databases. Often several references teach the same number of core concepts concepts **and are about equally relevant to your invention. Each new run may surface a different mix of them, or list them in a different order. The specific references and some concept ratings may shift, but the overall assessment stays consistent.

What to expect:

• Overall and market scores fall within ±15 points between runs and are typically much closer.

• Core concept ratings are stable. In more than 99% of cases, a concept's rating stays the same or moves one step, such as Partially Taught to Fully Taught.

• The patentability label (Low / Moderate / High) comes from the overall novelty score, so it changes only when a score is near the line between two labels.

How prior art references are evaluated

Each report identifies a set of core concepts for the invention — the key technical capabilities that define what it does (shown as C1, C2, C3, and so on in the report). Each prior art reference found is then evaluated against every concept:

• Fully taught — the reference describes this concept completely

• Partially taught — the reference covers part of it

• Not taught — the reference doesn't address it

A reference's similarity rating (High, Moderate, or Low) reflects how well it covers all of the invention's core concepts together — not just one. A reference that fully covers one concept but misses the others entirely will score lower than one that partially covers all of them. Breadth of coverage matters.

Novelty is then calculated based on how well the strongest matching references cover those concepts in combination. Novelty is based on the average strength of the prior-art matches. Stronger weighted average matches mean a lower novelty score; how those matches are distributed across individual core concepts does not drive the score. The score typically falls between 20% and 95% — a score of 95% does not mean no prior art exists; it means the closest references found cover your concepts only weakly.

How to read the novelty score

Our system surfaces the closest known prior art, then identifies the white space — the areas where the patent landscape is sparse or where no reference was found that fully teaches a concept. That white space is where your claims may have the most room. A lower novelty score doesn't mean the invention isn't patentable — it means the system found references that cover the concepts well, and you should be aware of them. A higher score means the specific combination of your concepts is less crowded in the existing landscape.

White space is a qualitative signal, not a calculated score. It does not indicate patentability or freedom to operate.

Start your prior art search in Tradespace

Tradespace’s IP assessment is designed to give you a fast, structured read on the landscape — summarizing relevant prior art, identifying claim angles, and surfacing competitive products. It is not a replacement for a full freedom-to-operate or patentability opinion from a patent attorney. Use these reports as a starting point for understanding the prior-art landscape— not as a substitute for advice from your patent counsel. Discuss the findings with your patent counsel, especially before making filing decisions.