I have an invention idea. What do I do first?

The first steps that protect your idea and your money, in the right order.

Most inventors I talk to have the same problem. They've had the idea for months, sometimes years, and they don't know whether to tell someone, look it up, file a patent or start building. So nothing happens, or the wrong thing happens first.

I've been inventing for a long time myself. I've designed a high-efficiency pump, built pool automation systems nobody sold, and made an exoskeleton back in 2011. The ideas that went well all followed roughly the same order. This guide lays it out.

1. Write it down and date it

Before anything else, get the idea out of your head and onto paper. Write down:

• the problem it solves, and who has that problem
• how it works, as far as you know right now
• rough sketches, even bad ones
• what makes it different from anything you've seen

Date it and keep it somewhere safe. A notebook works, and so does a document emailed to yourself. This won't legally protect the idea on its own, but it gives you a clear record, and it forces you to think it through properly. Half the time, writing it down shows you the part you haven't solved yet.

2. Check what already exists

This is the step people skip because they're scared of the answer. Do it anyway. It's free, and finding out now costs a lot less than finding out after you've paid for a prototype.

• Patent searches: search IP Australia's patent database and Google Patents using a few different ways of describing your idea
• Product searches: look through Amazon, eBay, Alibaba, Kickstarter and trade suppliers for anything similar
• Industry knowledge: ask people who work in the field, without giving away the clever part

Finding something similar isn't always the end. Often the existing product is expensive, clunky or solves a slightly different problem. That can be your opening. But you need to know exactly what you're up against.

3. Keep it confidential

Until you've had advice, treat the idea as a secret. Talk about the problem it solves, not how you solve it. When you do need to share details with a designer, engineer, manufacturer or possible partner, ask them to sign a non-disclosure agreement (NDA) first.

A good builder won't be offended. I'm happy to sign an NDA before anyone tells me the details of their idea, and I'd be wary of anyone who refuses.

4. Talk to a patent attorney before going public

This is the step that matters most, and the one people get backwards. In Australia, telling the public about your invention, through a social media post, a crowdfunding page, a market stall or a news story, can affect whether you can patent it later.

Australia does have a 12-month grace period for some disclosures, but relying on it is risky, and many other countries don't have the same rules. If you might ever sell overseas, a public post could close that door. So the safe order is: get advice first, then go public.

Some things a patent attorney can walk you through:

You might come across older articles about the "innovation patent". It has been phased out, and no new applications have been accepted since 25 August 2021.

5. Prove the core idea works

Before spending real money, prove the one thing your invention depends on. If it's a new mechanism, build a rough version from whatever's in the shed. If it's a sensor idea, wire up an off-the-shelf board on the bench and see if the reading is reliable.

This "proof of concept" doesn't need to look good. It just needs to answer the question: does the clever part actually work? When I was designing my pump, the early versions looked nothing like a finished product. Their only job was to prove the flow idea before anything was refined.

6. Scope it properly

Once you know the core idea works, work out what the whole product involves before you pay anyone to build it. That means:

• what parts it needs (mechanical, electronic, app, cloud)
• where and how it will be used (outdoors, wet, battery or mains)
• what it might cost to make, roughly, at different quantities
• what the first prototype needs to prove
• which rules apply if you sell it, such as electrical safety

If your idea includes electronics or an app, my guide on scoping an IoT product goes through this in more detail.

7. Build the prototype in stages

Prototypes work best in steps, with each one earned by the results of the last:

1. Proof of concept: rough, on the bench, proving the core idea
2. Looks-like model: a CAD design and 3D print to check size, shape and feel
3. Working prototype: it looks right and works right, tested where it will really be used
4. Pre-production: designed for manufacturing, with real materials and costs

8. Decide how you'll take it to market

With a working prototype, you can make the big decision: manufacture and sell it yourself, or license it to a company that already makes and sells similar products. Each suits a different kind of inventor. I've written a separate guide on licensing vs manufacturing your invention.

← More guides