Applicants tend to imagine a single queue: a stack of applications on a desk somewhere, working downward, with theirs near the bottom. The reality is less romantic and much more useful to know, because a stack cannot be influenced and a docketing system can.
What determines when an application is read is a chain of administrative decisions made in the first weeks after filing, most of them invisible to the applicant and one or two of them shaped by choices the applicant already made in the claims. Understanding the chain turns an unbearable silence into something closer to a schedule.
Mechanics · Inside the Docket
It Is Not One Line but Several Hundred Short Ones
Every application is classified by subject matter shortly after filing, using an international classification scheme that runs to tens of thousands of subdivisions. That classification routes the file to a specific examining group — commonly a team of eight to twenty examiners who handle one narrow technical area and nothing else. From that point the only examination queue that matters is the one inside that group, not the one running across the office as a whole.
The consequence is blunt: the wait has almost nothing to do with the quality of the invention and a great deal to do with what technical neighbourhood it lands in. A mechanical closure mechanism and a machine-learning signal processor filed on the same afternoon can receive first letters eighteen months apart, because one group has a two-month intake of new files and the other has a two-year one. Claims are what drive classification, so the drafting choice between framing an invention as a physical assembly or as a data process quietly sets the delay before anything substantive happens.
Examiners typically work two separate piles. New files are worked broadly in order of effective filing date, with internal rules obliging an examiner to act on the oldest unexamined case in the pile within each production period. Files that already have a response on record sit in a second pile with its own clock, often around two months. That structure explains a pattern applicants find maddening: a reply filed in March is not read in April, because the March reply has to wait for its turn in the amended pile behind everything filed in January and February. Knowing which of the several application types and office proceedings a filing belongs to is the first step in predicting which pile it will sit in and for how long.
Numbers · Reading Pendency
Two Figures Describe the Wait, and the Famous One Misleads
Offices publish two distinct measures. First-action pendency is the interval from filing to the first substantive letter, and it commonly sits between fourteen and twenty-six months. Total pendency runs from filing to grant or abandonment, and twenty-four to thirty-six months is normal, with congested fields routinely passing forty.
The trap is the headline average. Office-wide means are dragged around by whichever technical areas happen to be busy, and they conceal a spread that can be a factor of three between the fastest and slowest groups within the same institution. Where per-group statistics are published — and in many systems they are, along with allowance rates per group — the correct number to plan around is the one for the specific examining group, not the institutional average. An applicant told to expect a first letter in eighteen months on the strength of an office-wide figure, whose case sits in a group averaging thirty-one, has been given a schedule that will fail.
Scale explains most of it. Several large granting offices carry standing backlogs in the high hundreds of thousands of unexamined files, and each addition of a hundred thousand filings a year against flat examiner headcount extends every position in the examination queue behind it. The backlog is not a scandal; it is arithmetic, and it responds only to hiring, fee-based triage and the applicant's own choices.
The average pendency figure describes an office. It does not describe your application, and planning around it is how deadlines get missed.
On reading published statisticsAcceleration · Paid Routes
The Places Where Money Genuinely Buys Time
Most systems sell priority in some form, and the offers are more standardised than the terminology suggests.
Prioritised examination is the cleanest of them. A surcharge — often several times the base examination fee, and in large offices among the more expensive single line items in prosecution — buys a target of final disposition within about twelve months, usually with conditions attached: a claim ceiling in the region of four independent and thirty total claims, no extensions of time, and forfeiture of the special status if the claim set is later expanded. Many offices also cap the number of requests they accept in a year, so the route can close in the autumn and reopen in the spring.
A second route trades money for risk. Some offices grant accelerated treatment cheaply, or free, if the applicant supplies their own search and a written statement explaining why the claims are patentable over what was found. The fee saving is real and the exposure is real too: that statement becomes a permanent part of the file and can be read back against the patent for the rest of its life. It suits an applicant with a genuinely clean field and suits nobody who is unsure what the prior art contains.
The third route is work-sharing. Where a first office has already issued a favourable opinion on corresponding claims, a second office will frequently advance the case out of turn on request, generally without a fee, provided the claims in the second filing correspond in scope to those already found allowable. For an applicant pursuing several territories, sequencing matters enormously: obtaining one positive opinion early can pull every later filing forward by a year or more.
Acceleration · Free Levers
The Levers That Cost Nothing but Attention
Before paying anything, there are five moves that shorten a position in the examination queue at the cost of diligence alone, and together they routinely recover a year.
- Pay the examination fee at filing. Several systems allow the request for examination to be deferred by months or years, and treat the case as dormant until it arrives. Filing the request on day one rather than at the deadline is the single largest free saving available.
- Request early publication where examination depends on it. In systems that will not begin substantive work until the application has published, asking for publication ahead of the standard eighteen-month point advances everything behind it — at the price of disclosing early.
- File something formally clean. Drawings at the wrong line weight, an overlong abstract, a missing declaration or improperly dependent claims each generate a formalities letter, and each letter adds two to four months before the file reaches a technical examiner at all.
- Respond early, not on the deadline. A reply filed in week two rather than in month three does not jump the amended-file pile, but it enters that pile ten weeks sooner. Repeated across three rounds, that habit alone is worth most of a year.
- Ask for an interview. A thirty-minute call in which amendment language is agreed before it is filed frequently collapses two written rounds into one, and each round eliminated removes three to six months of calendar time.
Two more sit outside the applicant's control but are worth asking about. Many offices advance a file free of charge where an inventor is above a stated age or in poor health, on nothing more than a declaration. Others run standing pilot programmes for specified technologies — energy efficiency, emissions reduction, medical need — that grant special status on request. Both are underused, largely because neither is advertised.
Acceleration shortens the pending period. It does not lengthen the term, and it can quietly cost you term you were owed.
The arithmetic behind deliberate delayStrategy · Deliberate Delay
When Staying Where You Are Is the Correct Move
Speed is not free of consequence, and a surprising number of experienced applicants pay to slow down instead. The term runs from the filing date rather than the grant date, so a patent obtained in twelve months and one obtained in forty expire on the same day; acceleration buys enforceability sooner, not protection for longer. Worse, systems that add term back to compensate for their own administrative delay generally stop crediting it once the applicant has asked for special handling, so accelerating can forfeit months of term that were about to be granted for nothing.
A pending application also has a tactical value a granted one loses. Competitors reading a published application cannot know what the claims will finally say, and that uncertainty deters design decisions in a way a fixed granted claim does not. Some jurisdictions permit examination to be deferred for as long as three to seven years, and applicants whose market is still forming use exactly that window: keep the disclosure filed and the claim scope fluid until it is clear which embodiment is worth protecting, then narrow toward the thing that actually sells. This is where clarity about what a patent grants and what it does not pays for itself, because the decision is commercial rather than legal.
Cash flow argues the same way. Deferral spreads official fees, translation costs and attorney time across several budget years instead of compressing them into one, and it postpones the point at which annual renewal fees begin. Against that, an accelerated grant is close to mandatory when a licensing conversation, an investment round or an infringing competitor is already on the table — a right you can enforce next quarter is worth more than a broader one arriving in year four. Long-horizon inventing of the kind described in this account of how speculative fiction ends up seeding real inventions often has no counterparty waiting, and those cases are precisely the ones that gain nothing from a surcharge.
Two other objections change the calculus once they land. An office letter arguing that the application actually contains more than one distinct invention resets much of the timing analysis, because each carved-out filing enters the queue on its own terms. And where prosecution stalls entirely, the choice between arguing to a tribunal above the examiner and starting a fresh round of examination is as much a scheduling decision as a legal one. Both sit downstream of the sequence described in the wider account of what happens after filing.
The queue is administrative, not moral. Administrative things have levers.