CPC Based Deal Search (i.e., not a COVID post)

I lied. We’ll talk about COVID-19 a little. The lockdowns are hitting the economy and some of my clients quite hard. This is the first in a series of tools I’m providing to help out.

Consider the following:

“I want a pet. Pets are animals. Animals live in the forest. Therefore, I shall go to the forest to find a pet.”

My logic isn’t wrong – but it is stupid. The minor term of the syllogism is incomplete: “Pets are domestic animals.” Domestic animals are NOT found in the forest (alternatively, for the logic pundits, the major proposition was of form I and not A).

Equivalent incomplete reasoning often occurs when technology companies prosecute patent portfolios, are looking for joint research partners, are looking for buyers for their patents, etc.

“I want to sell my patents to a large company. My patents are useful to large companies in technical field X. Therefore, I shall look for large companies operating in technical field X.”

Based on that reasoning a Google search should suffice, which, unsurprisingly tends not to work. You find folks, but again and again, they just aren’t interested. What should the major premise have been instead?

“My patents are useful to large companies in technical field X that are motivated to do a deal.” Timing is everything and not everyone has reason to talk to you.

Ok, so how do we find companies motivated to do a deal? Lacking a crystal ball to peer into their boardrooms, we should instead look for companies that have done deals in our technology area in the recent past. Whatever motivated them to do that deal may still be relevant now.

But how do we find those deals? Most deals, even for public companies, don’t announce their terms publicly, let alone mention the exact technologies involved.

Well, under 35 USC 261, my ownership of an asset is only good against a subsequent bona fide purchaser if I record it at the patent office. So there’s a strong incentive for folks to record their asset transfer after the deal.

The good news? That data is public and accessible through a REST API.

The bad news? The REST API only permits searching by patent number, not by patent category (say by the CPC code).

Manually finding a patent relevant to our deal portfolio, then searching the assignment database to see if it was involved in a deal similar / relevant to the one we’re planning, really isn’t feasible.

There is however, tucked away in the APIs, another database relating CPC Codes to issued patents . . . so if some kind person were to patch up a CPC input frontend, they, in theory, (and in a hideous mockery of a SQL table JOIN) could patch up the backend to the assignment REST API so that we could effectively search for deals by their CPC codes (i.e., deals, where at least one of the patents involved in the deal falls within our CPC search query).

I have done such a thing.

Follow the link and you’ll be taken to a minimalist page like the above.

The USPTO databases are extremely fragile and slow. Searching by CPC group will take forever and returns too much anyway. So in the top left, you can only input a subclass, e.g., “A01B1/00.” Not sure which subclass your technology falls under (it’s on your patent/publication incidentally)? Click on “CPC Code Search” and it will take you to the CPC dictionary.

After you hit submit you’ll see the system working:

It may take a while (sometimes several minutes) so be patient. Only give up after, say 5 minutes without a response.

When it finishes, you’ll see this:

This is a list in descending chronological order of all recordations involving at least one asset (on the right) falling in your CPC code. By reviewing the reel/frame you can review the document and start putting together a pitch for your own assets. You can export the results to an Excel file by hitting the top right button (Excel will probably give you a security warning once it opens).

This isn’t perfect – but it’s probably a better start for considering deal targets than a Google search or blasting your LinkedIn network.

Certainly don’t rely on these results dispositively. As mentioned above, only recorded deals will appear here. Folks do do deals without recordation for various reasons. Also, the USPTO assignment database *IS* faulty. It’s pretty good, but there are omissions, broken fields, misplaced fields (a lot of entries confuse patent and application numbers for some reason). I have separate tools for cross-referencing data to mitigate some of the errors, but even those can’t correct all the errors (some are hard-coded in the database). Finally, this is very much a beta and I haven’t (and won’t for the near future) have time to address any bugs / optimizations of which I’m sure there are some of the former. YOU HAVE BEEN WARNED.

So if the economy’s giving you a hard time and you need / want to do a deal to make some cash, you might hunt around here first to get a feel for who the viable players may be.

Next time I’ll share the results from a tool that does a deeper dive on some of these databases to provide lower-level data on a target.

Also, I will get off technical posts and get back to case / legal posts soon. Just need to work through the queue to get there . . .

Stay safe.