Most advice for new inventors fails in the same way: it is a list of everything you will eventually need, handed to somebody who needs to know what to do on Saturday morning. This page is the short version. Three steps, in order, inside a fortnight, using nothing you do not already have.
Two things are deliberately not on the list. Building a prototype is not step one, and neither is finding a manufacturer. Both are reasonable things to do later. Done first, they are the two commonest ways an idea loses its protection or its owner's savings, usually both.
Step One · Paper
Write Down Exactly How It Works, and Date It
An idea in your head is not property. Nobody owns "a better bicycle lock". What can be owned is one specific technical solution, described precisely enough that somebody in the trade could build it from your words without inventing anything to fill the gaps. So the first step is writing, and it should be finished before anything is spent.
Sit down and answer four questions on paper. What are the parts? How are they arranged? What does each part do? By what mechanism does the arrangement produce the result? If you can answer all four, you have something a professional drafter can work from. If you cannot, you have an intention rather than an invention, and no amount of money converts one into the other.
- Date every page. A file created on a known date, kept unaltered, is worth far more later than a beautifully written undated document.
- Sketch it with rough dimensions. Artistic quality is irrelevant; the relationship between the parts is everything.
- List the variations you considered and rejected, with the reason for each. These frequently become the narrower fallback claims that save an application when the broad claim is refused.
- Photograph any mock-up, however crude, and save the photographs as separate dated files instead of overwriting them.
- Start a list of everyone who has already heard about the idea, with the date they heard it. You will need this and you will not remember it.
Save versions rather than replacing them. It takes a fortnight at most, and it is the only stage of the whole process where perfect protection costs nothing at all. That is why every sensible list of first steps for inventors begins here rather than with a workshop.
Step Two · Silence
Stop Telling People Until You Have Filed
This step consists of not doing something, which makes it the hardest of the three. A new idea produces a strong urge to tell people, and the response is usually encouraging, which makes the urge stronger. Resist it anyway, because this is the step where rights are lost permanently.
Describing your invention publicly before filing can end your ability to patent it. Some jurisdictions allow a grace period of twelve months, during which your own disclosure is not held against you. Many allow nothing at all: the rule there is absolute novelty, meaning the first public description of the invention counts against you from the day it appears. A market stall, a funding page, a detailed post explaining the mechanism, a short video of it working — any one of those can be enough to close the window for good.
"Public" is broader than it sounds. It does not need an audience or a sale. Making the description available to somebody who was under no obligation to keep it secret is generally sufficient. A conversation with a patent attorney is confidential by default. A conversation with a manufacturer, a colleague or a helpful relative is not, unless a written confidentiality agreement was signed before the conversation rather than afterwards.
Two weeks of silence costs you nothing. Two minutes of enthusiasm in the wrong room can cost you the invention outright.
Step two, in one sentenceThere is one exception worth naming, because people worry about it: telling a professional adviser is not disclosure. Neither is telling somebody who has signed an agreement first. Those two channels are open to you throughout, and using them early is not the risk. The risk is the friendly, unrecorded, unprotected conversation that felt too small to matter.
Step Three · The Record
Find Out Whether Somebody Already Did It
The third step is a search of the public record, and it is largely free. You are looking for prior art — everything already published anywhere, in any language: granted patents, published applications that never granted, expired patents, product catalogues, trade manuals, academic papers. Expired documents count fully. They defeat a claim just as effectively as a live competitor's patent, because the test is whether the idea was already described, not whether anybody still owns it.
Search in plain descriptive terms first — what the thing does, what problem it solves, what it is made of — then in the technical words you find in the results. Read the claims of anything close, not the abstract. Then write down what you found, what it covers, and why your invention is or is not distinguishable from it. That written summary is worth as much as the search.
Searches almost never come back empty, and that is not the bad news it feels like. Almost nothing is entirely new. The useful question is whether this specific combination — this mechanism, applied this way, to this problem — has already been described. Very often the search works by redirecting you: the headline concept turns out to be anticipated, while a detail you thought was incidental turns out to be the part nobody has claimed. Redirection now costs a few hundred at most. Redirection after a refusal costs a year and a drafting bill.
After the Three · What Changes
What the Fortnight Buys You, and What Comes Next
At the end of two weeks you should hold three things: a dated, versioned description complete enough for a drafter to work from; a list of everyone who knows and whether each conversation was covered; and a written search result with your own honest assessment of it. Total spend, in most cases, nothing.
From there the decisions get real. Whether to file a short placeholder application, which typically costs a few hundred in official fees and buys twelve months, or go straight to a full one. Whether a patent is even the right instrument, given that a registered design is cheaper and faster and a trademark lasts indefinitely. Whether to work with an organisation that does this professionally; groups that have spent four decades supporting independent inventors through this exact sequence tend to say the same thing about it — the expensive mistakes are cheap to make and slow to surface.
It is also worth reading how other people's fortnight turned into a product, because the accounts are less mysterious than they look. The story of how a household mixing device reached the market is mostly about the unglamorous early stages. So are accounts from other fields: work on inventions aimed at environmental problems and the case for letting inventors lead design in the built environment both describe the same pattern of long, cheap preparation followed by short, expensive commitment.
None of these three steps requires any special aptitude, and the sequence is the whole of the skill. The full route from here — what filing involves, what the granting office does with your paperwork and how long each part takes — is set out in our plain guide to getting an invention idea protected, which picks up exactly where this fortnight ends.
If you take only one thing from this page, make it the order. Written record, then silence, then search. The most common reason first steps for inventors go wrong is not skipping one of the three; it is doing them in the wrong sequence, and finding out too late that the third one would have changed the first.
Two weeks, three steps, no money. Everything after this is a decision — but only if you still have something left to decide.