Jan 17, 2011

The subtle art of articulating novelty

Invention is different to innovation. While I find that many are capable of great innovations, few are able to articulate the invention within the innovation.

I'm not hear to argue for or against the patent system (which varies from country to country) but rather I'd like to explain how it is that one identifies and articulates an invention. Articulating novelty is about explicitly describing your idea and its embodiment in a manner that clearly differentiates it from existing solutions (aka prior art).  This is what conclusively defines it as an invention.

My personal experience is that you get better at this with experience. I've distilled the approach down to a five step method that you can use to tease out the novelty within something you've imagined or built. Keep in mind, its good to take a couple of iterations through the method, as it will only help you improve your precision.

The method

1) Identify the innovation

Presumably, your starting point is that you think you're on to something. You have a solution that you believe has not been done before and you have an idea of what areas of the solution you think are novel.

2) Identify the problem domain 

For the areas you think are novel, what problem did you solve? Now, break the problem down to its most simple form.  Then describe it as succinctly as you can. For instance, say you developed a Web Service Integration Broker. The component you think is novel analyzes incoming payloads and identifies who they came from so the appropriate actions can be taken based on the supplier of the payload. The actual problem here has little to do with web services per se. Your  invention actually solves the problem of "How to identify the originator of an incoming XML document".

3) Narrow the problem domain

Consider the breadth of the problem. For the example above, does the solution identify the originator of any type of document or just XML Documents? Does the transport have to be HTTP or any form of  protocol? Now refine your problem statement to describe the scope of the problem as succinctly as you can.

4) Search for prior art

Using the new problem statement you've just defined, identify the prior art. This means search around and find all the other known solutions to the problem. Usually doing a "Google Patents" and Google/Yahoo Web searches are sufficient. This bit can take some time, but its critical that you're thorough. You will probably be a better candidate for this type of searching than paying a legal team to do it for you, as its unlikely they will understand the technologies in this space better than you.

5) Differentiate your invention

Using the prior art you've discovered from searching around your problem statement, you now set to work on differentiating your idea from the existing ones. This sometimes results in you realizing that your idea was not actually novel and exists already. However, assuming that is not the case, once you've differentiated your idea from the existing prior art you now know conclusively that your idea is novel !

Now what?

Well, there are several paths you can now take. Firstly, you should have no trouble concisely articulating why your innovation is different and possibly better than a competitors. Sales and Marketing will love this. Secondly, you've now done all the work required to file a patent, should you choose to do this. If you go this approach, I recommend you hire a patent lawyer if you can afford it.


Interviews with IBM inventors


No comments: