Before anything is filed
Whether a patent is the right protection at all, what the prior art position looks like, and whether you are free to operate in the first place.
Intellectual property
A patent is the only right you have to earn by giving something away. You publish how the invention works, and in return you get the right to stop other people doing it. Almost everything that decides whether that trade was worth making happens before anything is filed.
A patent protects an invention: a product, a process, a device, an apparatus, a formulation, a composition, a material, or a technical method of doing something. If it is a thing that works, or a way of making something work, it is in scope in principle.
What sits outside it matters just as much, because it is where most disappointment happens. A discovery, a scientific theory, a mathematical method, an aesthetic creation, a scheme or method for doing business, a computer program, and the presentation of information are all excluded as such. The words as such are carrying real weight there: software that produces a technical effect beyond the ordinary running of a computer is a different question from software described as a business idea, and the line between the two is where these applications are won and lost.
Methods of treatment and diagnosis practised on the human or animal body are excluded, although a substance or composition for use in such a method is not. Plant and animal varieties have their own separate route rather than this one.
The commonest version is disclosure. The invention was shown at a trade fair, sent to a manufacturer without anything signed, described in a pitch deck that went to twelve funds, or posted about. Novelty is judged against what was already available to the public, and once something is out there it is generally out there for good. No amount of money later reopens that door.
The second is a specification that reads well and claims badly. The description explains the invention beautifully and the claims cover one narrow way of doing it. A competitor reads the document you published, builds the version your claims do not reach, and thanks you for the instructions.
The third arrives years later, usually in a raise or a sale. The invention was made by a contractor, or by a founder before the company existed, or jointly with a university, and nothing was ever assigned. The patent is real and the company does not own it. That is a diligence problem with no quick answer.
The fourth is quiet. Renewals lapse on something still earning, because nobody was watching the dates and the reminder went to an address that stopped existing.
The claims are the right. Everything else in the document is context. When a dispute happens, the argument is about the words in the claims and almost nothing else, which is why drafting is the whole job rather than the paperwork at the end of it.
Disclosure decides whether you have anything at all. That means the order of events matters more than the quality of the invention: file, then talk. Where talking first is unavoidable, it needs to be under terms that keep the disclosure confidential, and the terms have to exist before the conversation rather than after it.
Ownership comes from the contract, not from the inventing. An employee inventing in the course of their duties is one position and a contractor is another, and the second one catches companies constantly because paying an invoice does not transfer anything.
Territory is a series of separate decisions. A patent is national. Protection elsewhere means separate steps in each place you care about, and the honest question is not where you could file but where you would actually be prepared to sue.
Timing cuts the other way too. Filing early fixes your date but freezes the invention at the stage it had reached, and inventions usually improve for months after the first version works. The useful question is when the thing you would want to claim in three years actually exists, and that is a judgement about the product rather than about the law.
Whether a patent is the right protection at all, what the prior art position looks like, and whether you are free to operate in the first place.
Claims written to be enforced rather than merely granted, with technical specialists in the relevant field working alongside us.
Examiner objections, amendments, and the decisions about scope that get made under time pressure and matter for decades.
Getting the invention into the company that is supposed to own it, including the ones made before the company existed.
Infringement of a granted patent, and reaching the operation actually manufacturing rather than the shopfront selling.
What to keep, what to let go, and knowing which of your rights is carrying the value.
Also under intellectual propertyBrands and trade marksContent and copyright
Plenty of inventions should not be patented, and saying so is the most useful thing we can tell you. A patent is a bargain: you publish the method, and you get a right that lasts a fixed period. If the thing you have is genuinely hard to reverse engineer, keeping it confidential may be worth more than any patent, and it does not expire.
The harder test is enforcement. A patent is only as good as your willingness and ability to litigate it. If you would not sue, you have paid to publish your method for the benefit of everyone who would. That is a commercial decision rather than a legal one, and it is better made at the start than discovered at the end.
And there is a version of this that is worse than doing nothing, which is filing narrowly to have something to point at. A granted patent that covers one embodiment of an idea tells a competitor exactly where the edge of your right sits, and reads to an investor as protection you do not have. Either the claims are worth defending or the filing is decoration.
Positions harden the moment the other side takes advice, and the quiet routes stop being available once a demand has gone out. While nothing has been sent, everything is still open.