Three Separate Timing Questions

Inventors ask when to file as if it were one decision. It is three. There is a legal deadline you cannot miss without losing rights permanently. There is a technical readiness point where the invention is defined well enough to claim broadly. And there is a financial point where the spend is justified by what the patent will actually do for the business. Those three dates rarely coincide, and filing well means managing the gap between them.

The outcomes are asymmetric, which should drive the decision. File too late and nothing can be fixed: the right is gone in the markets that matter. File too early and you have spent money on a narrow claim set aimed at a design you abandon two revisions later. Expensive, recoverable.

The Deadline You Cannot Move

US law gives an inventor a one-year grace period to file after their own public disclosure. Almost nowhere else does. Europe, China, Japan, and most other major markets apply absolute novelty: any public disclosure before your filing date destroys patentability there permanently, with no cure and no appeal. If the product has any export ambition, treat the US grace period as an emergency rescue, not a plan.

What counts as disclosure is broader than most inventors expect. A trade show booth, a crowdfunding page, a published spec sheet, a conference poster, a demo video, a quote sent to a distributor without an NDA, an offer for sale even if nobody buys. Any of these can start the clock or wipe out foreign rights outright, and the mechanics are worth reading in detail on public disclosure before filing a patent before you show anything to anyone.

The practical rule: your filing receipt should predate the first uncontrolled conversation about the invention. Everything before that date runs under a signed agreement, and the limits of that protection are covered in NDAs for inventors. A contract manufacturer quoting your part, a machine shop cutting a prototype, and a potential licensee reviewing a deck are all disclosures unless paper says otherwise.

When the Invention Is Ready to Claim

A patent protects what the claims describe, and claims can only be supported by what the specification actually teaches. You cannot add new matter later. So the question is not whether you have a finished product, but whether you understand the invention well enough to describe every variant you might want to cover.

You are ready when you can answer three things: what the core mechanism is, independent of the specific implementation; what the plausible alternatives to each element are, including materials, geometries, and sequences you have not built; and how someone skilled in the field would build it from your description without further invention. If you can only describe the one version sitting on your bench, filing now buys you a claim scope that a competitor walks around with a different spring, a different sensor, or a different order of operations.

You do not need a working prototype. You do need enablement. A CAD model, test data, and a clear mechanism narrative will usually carry a filing further than a rough physical build with no theory behind why it works.

The Provisional Bridge

A provisional application exists precisely for the gap between the legal deadline and technical readiness. It secures a filing date for whatever it discloses, costs a fraction of a full utility application, and gives you twelve months to develop, test, and talk to the market before committing to the expensive non-provisional filing. The filing mechanics are in how to file a provisional patent application.

The trap is treating it as a placeholder. A provisional only protects what it discloses. A three-page description with two hand sketches gives you a priority date for three pages of content, and anything you add in the non-provisional gets the later date instead. Since the priority date is what the examiner and any future litigant will measure prior art against, write the provisional as though it were the real thing. The consequences of the date itself are laid out in patent priority date.

A second useful pattern: file a provisional, then file an improved provisional four months later when testing has revealed the real variables, and claim both when the non-provisional goes in. Two provisionals still cost less than one wasted utility filing.

When Filing Is the Wrong Move

Some inventions should never be patented. A patent is a bargain: you publish a complete recipe, and in exchange you get roughly twenty years of exclusivity you must enforce yourself. If the invention is a process nobody can see from the outside of the product, publishing that recipe hands it to every competitor the day the application publishes, and the tradeoff is analyzed in trade secret vs patent.

Skip or defer filing when the product lifecycle is shorter than the examination timeline, when the market is too small to fund enforcement, when the advantage is brand or distribution rather than technology, or when a prior art search shows the concept is already published. Run that search first: a few thousand dollars of prior art searching regularly saves a filing that would have died at first office action. Budget matters here too, and the real numbers are in how much a patent costs in the US.

Filing Triggers Worth Watching

  • A date is scheduled for any public showing. Trade show, launch page, press briefing. File before, not after.
  • A partner, investor, or factory needs the full design. Paper first, then the filing, then the files.
  • Testing confirms the mechanism. The moment you know why it works, you can claim the family of ways it could work.
  • A competitor files in your class. Under first-to-file, the second application loses regardless of who invented first.
  • Tooling is about to be cut. Once steel is committed, the design is frozen anyway, so claim it.

Get the Timing Reviewed Before You Commit

Projects House works with inventors on the engineering side of this decision: what the invention actually is, which variants belong in the specification, and how the filing calendar fits the development and manufacturing schedule. Send a description of where your project stands through our contact form and we will tell you what still needs defining before a filing is worth paying for.