When an inventor pictures the examination of a new machine, the picture is usually a quiet specialist who will sit with the idea until its novelty is settled, as if the office were a laboratory and the file were a research problem without a clock. The person who actually opens that file inside the United States Patent and Trademark Office is working a production craft whose hours were assigned before the claims were read. The look is already budgeted. The work is to finish a reasoned judgment inside that budget, on a file that may be a fishing lure or a satellite, and to treat both as units of the same counted fortnight.
The hours that arrive with the technology
In the USPTO document on examiner performance plan changes for fiscal year 2026, expected time for an application sits at about 19 hours at the low end and about 35 hours at the high end. That range moves with the technology of the filing rather than with the inventor’s sense of how large the idea is. A fishing lure and a satellite can both be inventions, and the 2025 report of the Government Accountability Office numbered 25-107218 puts different clocks on the two of them. An experienced examiner is given 16.6 hours to take a fishing lure application through the look. A satellite is given 27.7 hours. A novice examiner on that same lure is given 30.2 hours, which is seniority written as a longer clock on the identical invention. The average across the files in that 2025 report is 20 hours. That midpoint sits between the lure and the satellite and does not describe either file.
The experienced examiner and the novice are not being asked to have different thoughts about the lure. They are being given different lengths of time in which to finish the same class of judgment. The satellite, which looks to an outsider like a larger idea, is simply a different technology on the same production sheet, because complexity here is a scheduling category rather than a license to keep reading until the idea yields.
The inventor who wrote the lure and the inventor who wrote the satellite may both believe the examination is a scientific reading of their thought. The office has already classified the thought as a quantity of hours. What looks from outside like a difference of importance is, on the sheet, a difference of allotment.
The biweek as the unit of the job
The working unit of the job is not the calendar day and not the life of the invention. It is an 80-hour biweek. In the same FY26 examiner performance plan document, 77 percent of that 80-hour biweek is spoken for as production time, which leaves the remaining slice of the fortnight for the other work an examiner still has to do between the files. An interview with the applicant is credited as an attribute of 1 hour in that same FY26 sheet. That 1 hour is not a conversation that expands until the idea is clear. It is a named increment on the same sheet that already holds the 19 to 35 hours.
The hours are on the sheet before the claims are understood.
Two counts and the intermediate refusal
A December 2025 USPTO video, Understanding a patent examiner’s role, names the scoring that turns those hours into a fortnight’s work. Each application brings two counts toward the production sheet. The biweek is not a stretch of thinking but a run of files that must close with those counts attached. The agency’s own training material puts 62.5 percent of the credit for a file on the first substantive written response, the document that lays earlier filings against the claims and tells the inventor where the idea already exists. That is a share of credit, not a measured share of the hours actually spent. Searching for those earlier filings is the substance of the look. The refusal is where that search has to land on paper. The biweek does not expand because the invention is elegant. The invention is fitted to the biweek.
Fully Successful is the whole quota
On 17 August 2026, Bloomberg Law reported that a Fully Successful rating now requires 100 percent of the production quota, in place of the former 95 percent that once sat between a full sheet and a passing one. The examiner who used to be allowed a thin miss is now measured against the whole quota, which is the current definition of a successful fortnight rather than a verdict on the quality of any one invention. Hitting 100 percent of production is now the floor of a successful rating, so the examiner who spends extra hours on a difficult satellite in order to be more certain is spending hours that the lure files in the same biweek still require. Those hours arrive through a secondary 2025 GAO report, a paywalled 17 August 2026 Bloomberg Law account, and a December 2025 office video that does not show the internal method.
A public standard, not a price
The hour here is not a price the professional sets, which is where it differs from what an hour of expertise costs for people who bill their time. It is an allotment the office publishes and the examiner works inside. That is a description of one agency in one performance year, and it does not settle whether the allotment is generous or thin. It does settle the shape of the work: the time for a file is budgeted in advance, and an examiner who spends longer on one application is spending time the schedule expects back somewhere else.
Sources and statuses
- 1GAO 25-107218Verified