What Real Provisional Applications Contain

Advice about provisionals is usually normative. File early, describe fully, do not skimp. What is harder to find is a description of what people actually file. This page is that description, from a reading of 32 provisionals and their paired non-provisionals.

Two findings follow, and they rest on different populations. The forms come from all 32 documents. The comparison with the later application comes from the 24 families that have alignment records. Keeping those apart matters, because a figure from 7 families says nothing about 9 documents.

Four Forms, Across 32 Documents

Every one of the 32 carried a specification. They differ in how much apparatus came with it: claims, numbered paragraphs, drawings, an abstract.

These are observed patterns, not four types anyone is required to pick from. The first two are a matter of degree, separated by how fully the claims were drafted, and the fourth rests on two documents and is the least settled of the four.

The headline number is the third row. Nine of 32 provisionals in this sample were not in application form at all. That is a real and common practice, not an aberration, and it is worth knowing before anyone tells you it never happens.

What the Later Application Made Explicit

For 24 of the families, the provisional and the non-provisional were compared element by element, and each element of the later document was recorded as present in the provisional, partially present, or additional.

Read the ranges, not the medians. They overlap heavily. One formatted specification was followed by an application that was 47% additional, which is higher than most of the informal ones. The medians describe groups of 7 and 8 families, which is small enough that a single unusual family moves them. The pattern is real but it is a tendency, not a rule, and it does not predict any particular filing.

Within the informal group, the kinds of content that appeared as additional most often were definitions of terms, calculation and mapping rules, and descriptions of how an operation is carried out. Stated precisely: 44% of the definition elements in those later applications were classified as additional, against 9% after the other forms. For calculation or mapping rules the figures are 42% and 14%, and for manner of performing an operation 34% and 13%.

What This Looks Like in One Family

Aggregates do not show what "made explicit" means. Two families from the sample, picked as the extremes of the range above, do.

A family at the top of the range. Its later application recorded 497 of 945 coded elements as additional. One of them is a definition. The provisional used a term in passing:

"many real estate platforms offer virtual tours, enabling a user to navigate through property in a virtual environment"

The later application defined it:

"may refer to a piece of land or real estate, which may include one or more buildings or other improvements or enhancements"

The reader's note records what changed between them: the provisional used the term in its ordinary sense without defining it or saying it could include buildings. Nothing was wrong with the earlier sentence. The later one does different work.

A family at the bottom. In another family the later application's 786 coded elements included none classified as additional. The provisional had already said it. Whatever the aggregate tendency, this is also what the sample contains, which is why the range belongs beside the median.

Both passages are from published US applications, quoted as filed. Families are identified by the sample's own reference numbers rather than by applicant or firm, because nothing here is an assessment of anyone's drafting.

What "Additional" Means Here, and What It Does Not

The word is doing narrow work and it is worth being exact about it.

• It means: a reading compared two documents and recorded this element of the later one as not present in the earlier one.
• It does not mean the provisional was deficient, incomplete, badly drafted or insufficient. Those are judgments the comparison never made.
• It does not mean the content "had to" be added. Nobody measured necessity. Applications change between filings for many reasons, including ones that have nothing to do with the earlier document.
• It does not mean the later claims lose the earlier date. Whether a provisional supports later claimed subject matter is a legal question about written description and enablement, decided on the specific claims and the specific disclosure.

What This Does Not Decide

• It does not say what any provisional is sufficient for.
• It does not say what you should file, or when.
• It says nothing about the statutory requirements themselves. Those are set out in what a provisional application actually is, and what the year buys is in what the twelve months buy.
• It is 32 documents and 24 comparisons. It is not a survey of US practice.

Educational, not legal advice. Filing dates and benefit claims carry deadlines that do not reopen, which is the practical reason to take them to a registered patent practitioner early rather than late.