An invention disclosure is the internal record made before any patent application exists.
It is not a filing. It establishes no priority, confers no rights, and means nothing at the patent office.
What it does is capture the information nobody can reconstruct later — who conceived what, when, and crucially whether anything has already been disclosed or sold.
Those last two decide deadlines, and they are the items most often discovered too late.
It is not a filing
| Disclosure | Application | |
|---|---|---|
| Establishes priority | No | Yes |
| Confers rights | No | On grant |
| Filed at the USPTO | No | Yes |
| Publishes | No | At 18 months |
What the form must capture
| Section | Why |
|---|---|
| The problem | Frames everything |
| The solution — how it works | The substance |
| Alternatives considered | Cannot be added to the specification later |
| Ranges, materials, parameters | Same — amendment room |
| What was tried and failed | Supports non-obviousness |
| Who contributed what | Inventorship |
| Date of conception | Evidence |
| Any public disclosure | Starts deadlines |
| Any offer for sale | Starts deadlines |
| Known prior art | Duty of candour input |
| Who else knows | Confidentiality assessment |
The disclosure and sale questions are the ones that change outcomes. A trade show six months ago alters the entire foreign filing analysis, and counsel cannot know it unless told.
Common gaps in disclosures
| Missing | Consequence |
|---|---|
| Alternatives | No amendment room later |
| Ranges | Cannot retreat to a sub-range |
| Disclosure dates | Deadlines miscalculated |
| Contribution detail | Inventorship unclear |
| Known prior art | IDS incomplete |
Why alternatives belong in the disclosure
| If prior art appears against | You need already described |
|---|---|
| The specific mechanism | An alternative mechanism |
| A stated value | A range with sub-ranges |
| The material | Alternative materials |
| The arrangement | Alternative geometries |
Nothing can be added to a specification after filing. Every alternative the inventor knows and does not write down is permanently unavailable during prosecution.
Inventors have this information and counsel does not. That is why the technical detail must come from the engineers rather than being extracted in a thirty-minute interview. See how to write a patent application.
Confidential does not mean unusable
| Use | Permitted |
|---|---|
| Internal review and triage | Yes |
| Counsel review | Yes |
| Wide internal circulation | Risky |
| External sharing without NDA | No |
| Investor diligence | Under NDA |
Dates and deadlines
| Event recorded | Consequence |
|---|---|
| Public disclosure | US: 1-year grace period. Elsewhere: rights forfeited |
| Offer for sale | Same, even for one unit |
| Conference paper, trade show | Public disclosure |
| Crowdfunding page | Public disclosure |
| Disclosure under NDA | Generally not public |
| Internal use only | Generally not public |
Most countries have no grace period. The US one-year window preserves US rights and does nothing for anywhere else.
Capture the exact date, not the month. Deadlines run from the day.
Who submits and when
| Trigger | Quality |
|---|---|
| On solving a problem in a non-obvious way | Correct |
| Before any external disclosure | Critical |
| At a project milestone | Catches inventions late |
| At annual review | Far too late |
| When someone remembers | Unreliable |
Milestone-triggered capture catches inventions after trade shows, by which point foreign rights may already be forfeited.
Train engineers on the trigger and on what counts as public disclosure.
Witnesses after first-to-file
| Pre-AIA | Now | |
|---|---|---|
| Establishes priority over another applicant | Yes | No — first to file |
| Evidence of conception | Yes | Yes |
| Evidence of contribution | Yes | Yes |
| Evidence against derivation claims | Yes | Yes |
| Still worth doing | — | Yes |
Witnessed, dated records no longer win priority races, which they did before the America Invents Act.
They remain useful for inventorship disputes, derivation proceedings and internal records, which is why most organisations still date and witness.
Inventorship starts here
| Contribution | Inventorship |
|---|---|
| Conceived a claimed element | Yes |
| Funded the work | No |
| Managed the project | No |
| Built it to instructions | No |
| Ordinary skill applied | No |
Inventorship is decided against the claims, which do not exist yet at disclosure stage. So the form records contributions and the determination happens later.
Record what each person contributed, not just their name. When claims narrow during prosecution, inventorship may need correcting, and the description is what makes that possible. See inventorship.
What counsel needs from the form
| Input | Why counsel cannot supply it |
|---|---|
| Alternatives | Only the inventor knows them |
| Ranges and tolerances | Domain knowledge |
| What failed and why | Not in any database |
| Which competitors do something similar | Field awareness |
| Disclosure and sale dates | Only you know |
Counsel drafts; inventors supply. A thirty-minute interview cannot extract what a properly completed form captures, and the specification cannot add it later.
Storage and retention
| Practice | Why |
|---|---|
| Keep every disclosure, filed or not | Evidence, and prior art context |
| Retain dates | Deadlines and conception |
| Searchable | Finds related work later |
| Access controlled | Confidentiality |
| Linked to any application | Traceability |
The triage process
| Stage | Question | Outcome |
|---|---|---|
| 1 | Is it commercially relevant? | Most stop here |
| 2 | Free prior art search | Many stop here |
| 3 | Is it patentable? | Counsel assessment |
| 4 | Patent, publish, or keep secret? | Strategy |
| 5 | Where to file | Markets and manufacturing |
| 6 | Budget approved | Filing |
A good process says no quickly. Disclosures that sit unreviewed for months cost nothing to store and can cost foreign rights if a launch happens meanwhile.
Step two is free and eliminates more than any other. See patent an idea free.
A minimal form
| Field | Required |
|---|---|
| Title and date | Yes |
| Inventors and contributions | Yes |
| Problem | Yes |
| Solution, in detail | Yes |
| Alternatives | Yes — the critical one |
| Ranges and parameters | Yes |
| Any disclosure or sale, with dates | Yes |
| Known prior art | Yes |
| Who else knows | Yes |
Nine fields. A form longer than this gets completed badly; one shorter than this misses something that matters.
Patent, publish or keep secret
| Route | Fits when |
|---|---|
| Patent | Visible in the product, worth the cost |
| Trade secret | Undetectable from outside |
| Defensive publication | You only need to stop others patenting it |
| Nothing | Not commercially relevant |
Filing forecloses trade secret permanently, because the application publishes at eighteen months whether or not it grants.
Defensive publication is cheap and underused where the concern is being blocked rather than blocking. See patent application publication.
Worked example: two disclosures
| Disclosure A | Disclosure B | |
|---|---|---|
| Technical detail | Two paragraphs | Full description |
| Alternatives listed | None | Five |
| Ranges | "About 15 degrees" | 5–40, preferably 12–18 |
| Prior disclosure recorded | Not asked | Trade show, 4 months ago |
| Contributors described | Names only | Contribution each |
What happened
| Disclosure A | Disclosure B | |
|---|---|---|
| Foreign filing | Proceeded — rights already forfeited, unknown | Assessed correctly |
| Prosecution | No fallback when art appeared | Amended to an alternative |
| Inventorship | Correction needed later | Clean |
| Outcome | Abandoned | Granted |
Disclosure A's failures were all recorded at disclosure stage, or rather were not. None was a drafting problem; all were input problems.
The trade show question is the one that cost most. Nobody asked, so nobody knew foreign rights were already gone.
Invention disclosure software
| Provides | Value |
|---|---|
| Structured intake form | Consistency |
| Routing and review workflow | The main benefit |
| Decision tracking | Audit trail |
| Docketing connection | Deadlines flow through |
| Reminders | Nothing sits unreviewed |
| Search across disclosures | Finds related work |
At low volume a document and a shared folder work. The tooling earns its cost when disclosures arrive faster than one person can triage them.
Evaluate on workflow and the audit trail, not on the form. The form is the easy part. See patent prosecution software.
Reviewing disclosures fairly
| Practice | Why |
|---|---|
| Respond to every submission | Silence discourages future ones |
| Explain a no | Teaches what qualifies |
| Decide within weeks | Deadlines run meanwhile |
| Recognise contributors | Whether or not you file |
| Track outcomes | Improves triage over time |
An unanswered disclosure teaches engineers not to submit. That cost is invisible and compounds.
Confidentiality
| Practice | Why |
|---|---|
| Mark it confidential | It is |
| Restrict circulation | Wide circulation can become disclosure |
| No external email without an NDA | Same |
| Store securely | Trade secret option depends on it |
| Train engineers on what not to post | Social and forum posts count |
Circulating a disclosure widely can itself be a public disclosure, which is exactly what the process exists to prevent.
What most disclosures become
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
Most disclosures should not become applications, and most applications that do grant are eventually released. Triaging honestly at the front is cheaper than discovering it at the third maintenance fee.
Invention disclosure: the checklist
- It is an internal record, not a filing. No rights, no priority.
- Ask about public disclosure and offers for sale. Those set the deadlines.
- Capture exact dates, not months.
- Have the inventors write the technical detail. Counsel cannot invent alternatives.
- List every alternative, range and material. None can be added after filing.
- Record what each contributor did, not just names.
- Date and witness it, even though priority no longer turns on it.
- Triage quickly. Slow review costs foreign rights when launches happen.
- Decide patent, publish or secret before filing forecloses the last two.
- Treat the disclosure as confidential. Wide circulation can itself be disclosure.