I got a request last week for a post about self publishing technical books, so I thought I'd give the disc jockey career a spin. I got my start in publishing by writing technical documentation to help bring money into a start-up I was working for around twenty years ago. The writing wasn't too challenging, manuals for a software product, but the production using Ventura Publisher 1.0 destroyed my self confidence for, well, forever. My first venture in technical self publishing was a short manual I wrote for technicians I was training back in 1991/92, an expanded version of which I published on the web as a free ebook in 1995. Around the same time, I co-authored a technical report on new electronic media which retailed for $125, and a little later, I signed my first contract with McGraw-Hill to write The Hand-Me-Down PC.
That's the quick background, but none of the above really describes the type of technical book that my correspondent was asking about, which would fall more into the category of a literary reference book. If anybody thinks that historical engineering type books or encyclopedic references can't be literary, I'd suggest reading some civil engineering histories for starters. At the purely technical end of the spectrum, we have all the C.R.C. books published by the Chemical Rubber Publishing Company (all the techies think it stands for Cyclic Redundancy Check), while at the purely literary end we have books like The Great Bridge by David McCullough, published by Simon & Schuster, which may never go out of print:-)
The potential market for a technical book pretty much depends on which end of the spectrum it falls on. Books on the purely technical end of the spectrum sell to engineers, university libraries and large corporate libraries (those with labs). The size of the market for those books depends almost entirely the subject. Somewhere I own a copy of a book on the radio frequency properties of certain materials that I think was published by the M.I.T. Press. The author was also the experimenter, so many of the constants measured appeared in other references. The audience for the book was limited to engineers who were interested in what happens to radio waves hitting those particular materials, so I suspect they got by with one print run. Another purely technical reference, The CRC Handbook of Chemistry and Physics is currently in its 87th edition, which may be some sort of record.
In the middle of the spectrum are the cross-over books, and I hope for my correspondent, a cross-over hit. These are books that contain enough detailed data and appendices to be of use as a reference text for an engineer or a manager making decisions about an engineering project, while maintaining an interesting enough storyline to be readable by engineers or technically oriented people who aren't planning on building or fixing one of the things. A sort of The Way Things Work (by David Macaulay and Neil Ardley) for just one thing, or the Soul Of A New Machine (Tracy Kidder, a local whom I've never met), but with a happy ending.
The main things authors or publishers of such technical books have going for them is a built-in audience and a (usually) noncompetitive environment. Judging the size of the built-in audience is tricky, and reaching a good portion of them in the year the book is released, as opposed to sustaining a trickle of sales forever, still requires some marketing. I think the best and most obvious place to do this marketing is on the web, with a website built around the book, but if there are any journals, magazines or IEEE type groupings around the specific subject, it would probably pay to try a little print advertising as well. The remaining decision for a publisher facing a non-competitive environment is whether to price for maximum sales or maximum profits. If the book is likely to be widely ordered by libraries, gouging is probably the order of the day. If the focus is on the retail customer, keeping the price in the literary, rather than technical range, is critical.