You are absolutely correct. This is indeed a first step in a large process (resources permitting). It is important to remember, however, that a trivial innovation in any product will prevent outright copying. By that same token, a trivial innovation that is difficult to design-around can protect more than just its increment.
I agree with your comments. It is unfortunately very difficult yo predict the current (or future) value of a "before its time" patent. Blocking patents are a little more identifiable by their relation to a base technology. Still not a simple analysis. You bring up a good point, however, in the link between valuation and inventor recognition. How do you recognize inventors who develop IP which is difficult to value, but nonetheless is providing (sometimes great) value to the company. I would like to find a good solution for that issue.
Many times, the "heavier stack" analysis does play a part in licensing negotiations, unfortunately. The long and short of it is that detailed patent valuation is an arduous (and imprecise) task.
By valuing your patents by their relation to competitor technology and revenue streams, you can bring an extra data point to the table. This data point can help counterweight an opposing numerical advantage.
Good question. Valuing or at least setting up a high level priority structure is useful to assist in making a number of IP related decisions. Patents cost money to maintain. Those with high value should be maintained, and those with low value should be further analyzed for potential abandonment and $ savings. High value patents related to my products may be reviewed for to determine if additional patents may be filed (continuations) based on the original disclosure, or brainstorming sessions may be setup with engineers to further develop and protect potential design-around ideas.
High value patents related to competitor products will be moved to the top of the list for potential infringement analysis.
These are just s few examples of uses for this simple value assessment.
As far as this analysis not being the only indicator of value, I completely agree. Think of this analysis as one (easy) indicator of positive value. We can agree that if a patent protects a large revenue stream, it likely has high value. I do not mean to suggest, however, that those patents that are not tied to a large revenue stream are of low value. Further analysis on other characteristics is necessary to make the high/low valuation on those patents.
Consider this an easy, step 1, analysis. We will probably agree that it is easier to calculate current revenue streams than future ones.
Assigning a patent to its most relevant revenue stream is a fine start, but you must also look at the patents themselves. For example, a company could have patents directed to an important product, from a revenue perspective, but the actual innovations covered may be trivial.
Important patents often reflect concepts developed "before their time". A patent may have no revenue flow in the early years until the world catches up and appreciates its importance. Also, blocking patents may prevent competitors from entering the space and yet not generate any "revenue" - even though they protect critical revenue streams. These issues can be important to inventors if their contributions to the success of the company are not fully appreciated until after the fact.
This is an interesting concept -- assigning value to your own company's patents based on tying them to your own company's revenues. That doesn't necessarily relate to any sort of intrinsic value, or what a licensee would pay -- but is perhaps a good way to determine how much you stand to lose if you get sued and some of the key claims in any of those patents get invalidated.
When it comes to cross licensing, is it not standard procedure for each company to just weigh their stack of patents, and the company that has the heavier stack receives money from the one with the lighter stack? :)
I am curious as to why the patent portfolio needs to be valued? If the company owns the IP already does it matter that some are more valuable than others. I guess I am missing the reasoning behind assigning value to particular patents. Anyway, I would also suggest that just because you are not earning a revenue stream from a patent does not detract from its "value". On the contrary it could be the most valued if it is in a future product. Just a few thoughts and questions. Thanks!
I don't disagree with your comments. By no means would I suggest that the techniques described in this article represent a complete patent portfolio valuation.
Really what I describe in this (brief) article is an exercise in finding value as much as it is an exercise in efficiency.
I suggest focusing on patents relating to established products with easily predicted revenue streams because the analysis is simplest, and if the products are creating $$s, it is hard to question the value of the corresponding patents. If nothing else, these patents need to be identified first.
Once that analysis is completed, and time and resources permit, the rest of the portfolio can be attacked from any number of angles (including the factors that you describe).
p.s. I have not seen a House-of-Quality approach applied to a patent analysis before, but it makes sense - I like the idea, and may have to give it a run on a future project.
This approach would seem to work for established products with easily predicted revenue streams, but how about new products where you don't know the revenue stream. Do you just go with the marketing estimate? This can be dangerous. On the other hand, a patent on a product outside the core may open up a large market.
I would suggest a broader muti-dimensional measure be developed to identify the potential (perhaps using a House-of-Quality like approach) for a patent over a wider range of criteria. It might be interesting to identify a first cut criteria list. Anyone have a suggested starting point? Would you include things like sunk cost? Numbers of recent patents in this or similar fields? Market size (from outside sources)? Rankings of importance from key customers?
Blog Doing Math in FPGAs Tom Burke 18 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...