Trade secrets have been a lively topic for IPKat readers recently. Neil's posting on "Term of Trade Secret Protection: the Dirty Little Secret" (here) has generated some good comments, while Christopher F. Marki's "The Case Against Patents" (here, flagged on this weblog here) extols trade secrets in his own industrial sector as a far superior tool to patents has also sparked off some thoughts. The IPKat received the link to Marki's piece from Ron Camp (Kilburn & Strode), who now writes:
"I thought [Marki's post] might be of interest as one of the few examples I am aware of that sets out the usefulness of know-how, but I agree that the author is working in a specialist field where protection by know-how happens to be appropriate. The extent to which companies rely on trade secrets and know-how is normally something of an unknown quantity to outsiders, as otherwise it wouldn't be secret!
I guess it comes down to "horses for courses": the microwave field has long relied heavily on know-how. In the 1970s, the Plessey company managed to tweak what was a fairly standardised process for making ECL logic circuits to allow frequency dividers to operate at the then-astronomical rate of 1.2GHz, about twice as fast as the best the competition could manage. They had the field to themselves during the many years it took for the competition to catch up.
When I was with the patent department of what used to be a large UK-based multinational electrical company, we relied on both types of protection as appropriate for the technology and the needs of the business. A pile of patents (the more the better) was certainly absolutely essential for most businesses when it came to periodic cross-licence negotiations with competitors: individual patents were rarely significant per se (for the representative situation in that particular field of technology, see GEC Avionics' patent  RPC 107). However, where the invention related to manufacture of an improved version of a known type of hi-tech component, and the means by which the component had been processed during manufacture to obtain the improvement was not apparent from inspection of the device, the usual practice was not to patent, because how would you police such a patent?
On the other hand, relying solely on know-how can be dangerous: I am aware of one business which -- due to bad advice from their own legal advisers who were not well versed in patents -- had traditionally relied solely on know-how, nearly came to grief when their new and highly innovative product fell foul of a competitor's patent: only the ability to file a US patent application 11 1/2 months after the first public disclosure of their new product gave them sufficient bargaining power to negotiate a cross-licence and stay in business.
A UK Intellectual Property Office web page here gives the following advice on the situations where trade secret protection might be appropriate:
If it would be difficult to copy the process, construction or formulation from your product itself, a trade secret may give you the protection you need.
My own experience is that a trade secret can also be appropriate where a manufacturing process involves a processing step that is very easy to perform but which no-one would think of performing. A process that requires specialist equipment to carry it out, is arguably more readily detectable than a process which involves the use of something that would be found in any manufacturing establishment and which would therefore not involve the "infringer" in the purchase and installation of equipment that he had not hitherto needed".