Provisional Form Worksheet
Three plain questions that place your material in one of four forms observed in 32 US software provisional applications, and what the sample shows about the later applications that followed each form. Read it below, copy it, or download it.
Observed in 32 US software provisional applications. The four forms come from all 32 documents; what followed each form comes from the 24 families whose later application was compared with the provisional element by element. Method and limits: how we read these filings.
What this worksheet does
It takes something you have written (a design note, a paper, a slide deck, a draft application, or several documents) and asks three questions about its form. The first yes places it in one of the four forms the reading found. The worksheet then reports, for that form only, what the sample shows about the non-provisional application that came later: how many families there are, the median and range of the share of the later application classified as additional, and the kinds of content with the most elements in that share.
The questions are about the material's form, not about the idea. They ask what is on the page: separate documents, a claim set, numbered paragraphs. They need no patent vocabulary beyond that.
What it does not do
It describes. It does not say which form to file, it does not rank the forms, and it does not answer whether any document can be filed as a provisional or what a filing would preserve. Those turn on the specific material and the claims that come later, which is work for a registered patent practitioner.
How the questions were tested
The questions come from the corpus's own selection guide. On 8 held-out families, the form the questions gave matched the reading's form in 6 of 8, against 3 of 8 for always giving the most common form. The answers in that test were taken from the provisional itself, the same document being classified, so the result shows that the questions separate the four forms. It does not show that they predict anything about a later application. The questions were revised once before that test, after a first test on 8 other families (6 of 8 there, against 4 of 8).
The Worksheet
Your Material
What the Material Is
Name each document you are answering from and say in a line what it is: a paper, a design note, a slide deck, a free description, a draft specification, or something else. Note its date or version.
How Many Separate Documents
List the documents one per line. For each, note whether it has its own title, its own order of sections, or its own paragraph numbering, and whether another document refers to it and explains it.
Three Questions About Its Form
Answer in order, from the material itself rather than from memory of it. The first yes places the material, and you stop there. Each answer describes the form the material is in. It is not a recommendation.
Is it more than one separately written document? Question 1
Is the material more than one document, each written separately with its own structure (its own title, section order or paragraph numbering), such as two specifications? An appendix of code or data tables that the main text refers to and explains is part of one document. So is one document assembled from parts written at different times, however its sections restart: the question counts the documents as they stand, not how they were written. If yes, the material is in form P4, a compiled multi-document filing. Write yes or no, and the evidence.
Does it contain a claim set? Question 2
Does the material contain numbered statements that each name the elements of the subject matter, with at least one statement that refers back to another and narrows it ("the method of statement 1, wherein...")? A page headed "claims" that holds one sentence or a list of features does not count: 4 of the 8 P2 documents in the sample carried a page like that. If yes, the material is in form P1, an application with a claim set. Write yes or no, and the evidence.
Are its paragraphs numbered? Question 3
Is the description written with numbered paragraphs ([0001], [0002] and onward)? Standard headings and numbered figures do not decide it, because the reading found them in P1, P2 and P3 alike. Numbered sections in a paper are not numbered paragraphs. In the sample, all 21 P1 and P2 documents had numbered paragraphs and none of the 9 P3 documents did. If yes, the material is in form P2, a formatted specification without a claim set. If no, it is in form P3, an informal technical disclosure.
The Form Your Material Is In (P1 / P2 / P3 / P4)
Write the form the first yes gave. If an answer was close, note why: a claims page with only a few statements that refer back, or a draft with numbered paragraphs in some sections and not others. The reading treats P1 against P2 as a matter of degree, separated by how fully the claims are drafted.
What the Sample Shows After Each Form
For 24 families, each element of the later non-provisional application was compared with the provisional and recorded as present, partly present, or additional. Read only the entry for your form. Every figure is of the families in that form, not of all 24. "Additional" means a reading did not find that element in the provisional. It does not mean anything was missing or had to be added. Counts of kinds are pooled across a form's families, so a family with a long later application and many additional elements weighs most.
If P1: Full Application Draft With a Claim Set
13 of the 32 provisionals in the sample; the later application was compared for 8 families. Share classified as additional: median 1.5% (8 families), range 0% to 22%. The three kinds with the most elements classified as additional, by count: descriptions of what a figure shows (51), background on existing technology (48), citations of other documents (39). Two of the 8 families hold 568 of the 659 additional elements, so this ranking mostly describes those two.
If P2: Formatted Specification Without a Claim Set
8 of the 32 provisionals in the sample; the later application was compared for 7 families. Share classified as additional: median 5.2% (7 families), range 1.0% to 47%. The three kinds with the most elements classified as additional, by count: background on existing technology (96), steps or elements recited in a claim (54), examples of what a general part can be (49). Two of the 7 families hold 576 of the 852 additional elements.
If P3: Informal Technical Disclosure
9 of the 32 provisionals in the sample; the later application was compared for 7 families. Share classified as additional: median 20.7% (7 families), range 5.7% to 56%. The three kinds with the most elements classified as additional, by count: definitions of terms (196, in all 7 families), examples of what a general part can be (101, in all 7), descriptions of what a figure shows (80, in 5). Two of the 7 families hold 982 of the 1,705 additional elements.
If P4: Compiled Multi-Document Filing
2 of the 32 provisionals in the sample; both families were compared: 0.3% and 2.2% of the later application's elements were classified as additional. Two families are too few for a median or a ranking of kinds, and the reading marks this form as tentative.
Your Notes
Kinds From Your Form's Entry, Checked Against Your Material
For each kind named under your form (the P4 entry names none), note whether your material states it: yes, in part, or not. This records what the material contains. It does not say whether anything needs to be added.
Questions for a Patent Practitioner
Anything the questions left uncertain, or that you want to ask about before deciding what to do with the material.
The four forms in the sample
Every one of the 32 provisionals carried a specification. The forms differ in what came with it.
| Form | Documents | Numbered paragraphs ([0001] markers) | Claims document | Abstract |
|---|---|---|---|---|
| P1 Full application draft with a claim set | 13 of 32 | 13 of 13 | 13 of 13 | 10 of 13 |
| P2 Formatted specification without a claim set | 8 of 32 | 8 of 8 | 4 of 8, nominal (one sentence or a list of features) | 5 of 8 |
| P3 Informal technical disclosure | 9 of 32 | 0 of 9 | 3 of 9 | 2 of 9 |
| P4 Compiled multi-document filing | 2 of 32 | 1 of 2 (read from scanned text) | 2 of 2 | 2 of 2 |
These are observed patterns, not categories anyone has to choose from. The names are the reading's own. P3 covers a paper, a design note, a slide deck or a free description filed as it stands. P4 rests on two documents and is the least settled of the four.
What followed each form
For the 24 families with alignment records, the share of the later application's elements classified as additional:
| Earlier form | Families | Median | Range | Each family |
|---|---|---|---|---|
| P1 | 8 | 1.5% (8 families) | 0% to 22% | 0%, 0.2%, 0.3%, 0.8%, 2.3%, 12%, 21%, 22% |
| P2 | 7 | 5.2% (7 families) | 1.0% to 47% | 1.0%, 1.6%, 2.6%, 5.2%, 29%, 47%, 47% |
| P3 | 7 | 20.7% (7 families) | 5.7% to 56% | 5.7%, 7.9%, 12%, 21%, 27%, 43%, 56% |
| P4 | 2 | not reported (2 families) | 0.3% and 2.2% | 0.3%, 2.2% |
Read the ranges, not the medians. They overlap. Two P2 families were followed by applications 47% additional, higher than six of the seven P3 families. Groups of 7 and 8 are small enough that one family moves a median. The pattern is a tendency in this sample. It does not predict any particular filing.
Which kinds of content made up the additional share
Each element of a later application is coded with one of 84 content kinds. The tables below rank, for each form, the ten kinds with the most elements classified as additional, by count. "Families with any" is how many of the form's families had at least one element of that kind classified as additional. The last column gives the same kind after the other three forms pooled, for comparison. Kinds marked drafting material are about the document or its claims rather than the system; the reading's own taxonomy classes them that way. P4 is not ranked: its two families hold 17 additional elements between them.
| Rank | Kind of content | Additional, after P1 | Families with any (of 8) | After the other three forms (16 families) |
|---|---|---|---|---|
| 1 | Descriptions of what a figure shows (drafting material) | 51 of 338 (15%) | 3 | 111 of 577 (19%) |
| 2 | Background on existing technology | 48 of 196 (24%) | 4 | 115 of 664 (17%) |
| 3 | Citations of other documents (drafting material) | 39 of 46 (85%) | 7 | 46 of 141 (33%) |
| 4 | Steps or elements recited in a claim (drafting material) | 28 of 453 (6%) | 3 | 102 of 1,200 (8%) |
| 5 | What a part or step is for | 24 of 207 (12%) | 3 | 48 of 299 (16%) |
| 6 (tied) | The shortcoming of existing practice | 23 of 136 (17%) | 4 | 45 of 280 (16%) |
| 6 (tied) | Rules for calculating or mapping a value | 23 of 104 (22%) | 2 | 73 of 263 (28%) |
| 6 (tied) | How an operation is carried out | 23 of 208 (11%) | 2 | 73 of 289 (25%) |
| 9 (tied) | Illustrative examples | 22 of 156 (14%) | 3 | 52 of 292 (18%) |
| 9 (tied) | Reported implementation or testing | 22 of 48 (46%) | 1 | 10 of 36 (28%) |
Two of the 8 P1 families hold 568 of the form's 659 additional elements, so this ranking mostly describes those two. Five of the 8 had 2.3% or less of the later application classified as additional.
| Rank | Kind of content | Additional, after P2 | Families with any (of 7) | After the other three forms (17 families) |
|---|---|---|---|---|
| 1 | Background on existing technology | 96 of 419 (23%) | 2 | 67 of 441 (15%) |
| 2 | Steps or elements recited in a claim (drafting material) | 54 of 523 (10%) | 4 | 76 of 1,130 (7%) |
| 3 | Examples of what a general part can be | 49 of 278 (18%) | 4 | 104 of 824 (13%) |
| 4 | Other ways or forms of doing it | 39 of 146 (27%) | 5 | 88 of 433 (20%) |
| 5 | What each part does | 33 of 238 (14%) | 5 | 44 of 604 (7%) |
| 6 | The user interface and what it presents | 32 of 170 (19%) | 2 | 29 of 312 (9%) |
| 7 | Descriptions of what a figure shows (drafting material) | 31 of 243 (13%) | 3 | 131 of 672 (19%) |
| 8 (tied) | Statements that the description does not limit the subject matter (drafting material) | 30 of 127 (24%) | 3 | 67 of 380 (18%) |
| 8 (tied) | The shortcoming of existing practice | 30 of 138 (22%) | 2 | 38 of 278 (14%) |
| 10 | How parts connect and communicate | 28 of 232 (12%) | 4 | 33 of 412 (8%) |
Two of the 7 P2 families hold 576 of the form's 852 additional elements.
| Rank | Kind of content | Additional, after P3 | Families with any (of 7) | After the other three forms (17 families) |
|---|---|---|---|---|
| 1 | Definitions of terms | 196 of 446 (44%) | 7 | 45 of 478 (9%) |
| 2 | Examples of what a general part can be | 101 of 425 (24%) | 7 | 52 of 677 (8%) |
| 3 | Descriptions of what a figure shows (drafting material) | 80 of 259 (31%) | 5 | 82 of 656 (12%) |
| 4 | Other ways or forms of doing it | 75 of 224 (33%) | 6 | 52 of 355 (15%) |
| 5 | Rules for calculating or mapping a value | 68 of 160 (42%) | 3 | 28 of 207 (14%) |
| 6 | Statements that the description does not limit the subject matter (drafting material) | 58 of 183 (32%) | 7 | 39 of 324 (12%) |
| 7 | How an operation is carried out | 52 of 152 (34%) | 4 | 44 of 345 (13%) |
| 8 | Illustrative examples | 46 of 231 (20%) | 5 | 28 of 217 (13%) |
| 9 (tied) | What a determination is based on | 45 of 195 (23%) | 5 | 26 of 254 (10%) |
| 9 (tied) | Steps or elements recited in a claim (drafting material) | 45 of 573 (8%) | 7 | 85 of 1,080 (8%) |
Two of the 7 P3 families hold 982 of the form's 1,705 additional elements.
A rank and a contrast are different claims. After P3, definitions of terms rank first by count. Rules for calculating or mapping a value rank fifth, and how an operation is carried out ranks seventh. What sets those three apart is the contrast with the other forms: 44% of the definition elements in the later applications were classified as additional after P3 (196 of 446), against 9% after the other three forms (45 of 478). For calculating or mapping rules the shares are 42% (68 of 160) against 14% (28 of 207), and for how an operation is carried out 34% (52 of 152) against 13% (44 of 345). Those are three of the contrasts in the last column, not the only large ones, and none of them is a share of filings: they count elements of one kind inside documents, not documents lacking a feature.
What "additional" means here
- It means a reading compared two documents and recorded an element of the later one as not stated in the earlier one at any level of detail. An element the earlier document stated only in part is recorded as partly present, not as additional.
- It does not mean the provisional was deficient or that the content had to be added. Nobody measured necessity, and applications change between filings for many reasons.
- It does not measure quality. A form followed by a smaller additional share is not a better form, and the ranges above overlap.
Frequently Asked Questions
Does this worksheet tell me which form to file?
No. It describes the form your material is already in and reports what a sample of 32 US software filings shows for that form. Which form to file, if any, is a decision for you and a registered patent practitioner, made on your specific material.
Does a smaller share classified as additional mean a better provisional?
No. The share counts elements of the later application that a reading did not find in the provisional. It does not measure quality. The ranges for the forms overlap, and groups of 7 and 8 families are small.
Can I file my design document as a provisional?
The worksheet does not answer that. It can tell you whether your document is in the form the reading calls an informal technical disclosure, which 9 of the 32 provisionals in the sample were. Whether to file it is a question for a registered patent practitioner.
Where do the figures come from?
The four forms come from a reading of 32 US software provisional applications. What followed each form comes from the 24 families whose later non-provisional application was compared with the provisional element by element. The method and its limits are on the page about how we read these filings.
Related
- What real provisional applications contain, the findings this worksheet is built on
- How we read these filings, the sampling, the definitions and the limits
- What a provisional application actually is, from the statute
- What the twelve months buy
- Idea Disclosure Template, for writing the material down in the first place
- Disclosure Quality Check
Scope. Completing this worksheet does not file anything, establish a filing date, or say what any document can be used for. It describes the form of your material and what a sample of 32 US software filings shows. Rules can differ by country.
Disclaimer. This worksheet is used solely to describe the form of a technical document and to report observations from a sample of published US patent filings. It and any output are for informational and technical-documentation purposes only and are not a substitute for advice from qualified intellectual property counsel. Use of this worksheet does not constitute legal services or legal, patent, patentability, inventorship, ownership, claim-drafting, filing-strategy, deadline, or freedom-to-operate advice, and does not create an attorney-client relationship with Obviously Not. Intellectual-property protection can be time-sensitive: public disclosure, product release, open-source publication, sale, offer for sale, demo, or other external sharing may affect available rights. You are solely responsible for promptly consulting qualified intellectual property counsel about any legal or IP issue.