An envelope arrives, eighteen months after filing, and every claim in it has been rejected. Claim 1 rejected. Claims 2 through 14 rejected. The word appears so many times that most first-time applicants read the first page, feel the floor go, and never read the rest. That reaction is understandable and completely wrong. The document in that envelope is not a refusal. It is the examiner's opening position, and it was written with the expectation that somebody would answer it.
Patent examination is adversarial by design. An examiner is paid to find reasons the claims should not be granted, because the alternative — an office that grants what it is handed — produces patents nobody can rely on. The applicant's job is to answer those reasons, in writing, on the record. What follows is what that document actually contains, how to tell its four very different kinds of objection apart, and what separates a reply that moves the case from one that wastes a year.
Base Rates · 01
Rejection Is the Default Setting
Across most major examining offices, the share of applications allowed on first examination sits somewhere between one in ten and one in six. The overwhelming majority draw an initial report rejecting some or all claims. That is not a statement about the quality of those inventions; it is a statement about how the process is structured. A first filing is deliberately drafted broad, because broad claims are cheap to narrow later and impossible to widen once filed. The examiner's first job is to push back against that breadth with whatever the search turned up.
Understanding the economics helps. An examiner typically has somewhere between twelve and thirty hours of credited time for the entire life of an application — searching, reading, writing the report, considering the reply, and writing the next one. That time is spread across a docket of dozens of cases. The first report is therefore written fast, against a search that may have taken three or four hours, and it will sometimes map a reference onto a claim in a way that does not survive careful reading. Roughly a third of first-action rejections are withdrawn or substantially narrowed after the applicant's first substantive reply, without any amendment at all.
Anatomy · 02
What Is Actually Inside the Envelope
The report has a predictable structure, and reading it out of order makes it far less frightening. Skip the cover sheet. Go to the summary page, which lists exactly which claims stand rejected, which stand objected to, and — crucially — whether any claim has been indicated as allowable. A single dependent claim marked "would be allowable if rewritten in independent form" is the most valuable sentence in the entire document, because it tells you precisely where the examiner believes the boundary of the prior art sits.
Next comes the substantive body: each ground of rejection, the claims it applies to, and the references relied on, usually with column-and-line or paragraph citations pointing at the specific passages said to disclose each claim element. Then the search notes — the classification areas searched, the databases queried, the keyword strings used. Those notes are routinely ignored and shouldn't be: they reveal what the examiner was looking for, and occasionally reveal that they were looking in the wrong place entirely.
Two words are also worth separating. A rejection goes to the substance of a claim — it says the claimed invention is not patentable over what came before. An objection goes to form: a typo in the specification, a drawing missing a reference numeral, an inconsistent antecedent. Objections are usually fixed in an afternoon. Confusing the two makes a routine housekeeping list look like a catastrophe. Learning to separate the citations from the arguments built on them is a skill in itself, and one worth studying against a real example, which is why it helps to work through how the cited references are structured before drafting anything.
An examiner's first report is a hypothesis about your claims, written quickly against one search. Hypotheses can be wrong. Saying precisely why is the whole job.
The frame that makes the document readable
Grounds · 03
Four Grounds, Four Different Fights
Nearly every rejection falls into one of four categories, and each demands a structurally different answer. Treating them as one undifferentiated problem is the most common error in self-drafted replies.
Novelty. One single reference is said to disclose every element of the claim. The answer is elemental: identify a limitation in the claim that the reference does not disclose. One missing element defeats the ground entirely. These are the most winnable rejections and the fastest to resolve.
Inventive step. Two or more references are combined, with an asserted reason why a skilled person would have combined them. The fight is rarely about what the references say and almost always about the reason for combining them.
Clarity. A claim term is ambiguous, has no antecedent, or is a relative term with no yardstick in the specification. Usually cured by amendment, and usually cheap.
Sufficiency. The description does not teach a skilled person how to make and use the full breadth of what is claimed. The most dangerous ground, because new matter cannot be added after filing.
The second category is where most of the argument happens, and where the most ground is needlessly conceded. An answer built on the missing rationale for combining references is very often stronger than an amendment, because it preserves claim breadth rather than trading it away. Practical help with patenting invention ideas at this stage usually consists less of legal fireworks than of a disciplined reading: what does the claim actually say, what does the reference actually disclose, and where exactly is the gap between them.
The Clock · 04
Deadlines That Do Not Negotiate
Every report carries a response period, commonly two to four months from the mailing date, extendable in many territories up to six months on payment of fees that escalate sharply month by month. Miss the outer date and the application goes abandoned. Revival is sometimes possible, but it requires a petition, a fee, and in some systems a statement about why the delay was unintentional — a process that costs more than the extension would have and does not always succeed.
The second timing feature that catches people out is finality. In many systems the second report can be issued as final, which does not mean the case is over but does mean the applicant's right to have new amendments entered as of right has closed. From that point the options narrow to a small set: an after-final amendment the examiner may decline to enter, a request for continued examination that reopens prosecution for a fee, an appeal to a review board, or a fresh application claiming the earlier filing date. Board appeals are thorough and slow — commonly twelve to thirty months from filing the brief to a decision — which is why a well-timed refiling often beats an appeal on both cost and calendar. That calculation is the whole reason keeping a family deliberately open is treated as strategy rather than paperwork.
The Reply · 05
What a Strong Response Contains
A reply has two halves: amendments, and remarks. The amendments are the claim text as it will now read. The remarks explain, ground by ground and claim by claim, why the rejections no longer apply. Weak replies assert; strong replies quote. The examiner cited column 4, lines 22 to 31 — so the reply reproduces those lines and shows that what they describe is a different mechanism from the one the claim requires. Argument anchored to text is checkable. Argument anchored to adjectives is not.
The most expensive habit in first-time prosecution is reflexive narrowing. An applicant reads a rejection, panics, and adds three limitations to claim 1 to escape it. The claim is allowed, and the patent that issues covers a version of the invention so specific that any competitor can design around it in a week. Worse, the record now contains a statement about why that narrowing was necessary — a statement that will be read back in any later dispute about scope. Narrow when you must, argue when you can, and never do both to solve the same rejection.
Interviews are underused. Most offices permit a scheduled conversation with the examiner, typically twenty to thirty minutes, before or after a reply is filed. Agreement reached in an interview still has to be put on the record in writing, but the conversation frequently collapses a dispute that two written rounds would not have resolved — often because the examiner and the applicant turn out to have been reading a single claim term differently the whole time. Accounts of long-running inventor advocacy, such as this profile of a career spent working alongside independent inventors, tend to describe the same pattern: the cases that resolve are the ones where somebody kept talking.
None of this makes a rejection pleasant, but it does make it ordinary. The record is full of household products that took three rounds of prosecution to reach a grantable claim — the story of one kitchen invention brought to market by a first-time inventor follows the pattern almost exactly, and so does much of the work now happening in independently developed green technology, where crowded prior art makes an initial rejection close to inevitable. What separates the applications that grant from the ones that lapse is almost never the strength of the idea. It is whether somebody read the whole document, understood which of the four fights they were in, and answered on time.
Never amend to escape a rejection you could have argued away. Every limitation you add narrows the patent for the whole of its twenty-year life.
The rule that survives contact with the office
Read the summary page first. Find the claim the examiner has already conceded. Separate the objections from the rejections, sort the rejections into their four kinds, diary the outer deadline the day the report arrives, and answer the citations with quotations rather than adjectives. That is the entire method, and it is available to anyone willing to read past page one.
End of report