Back to Truss Tales

Was Warren Buffet Wrong?

Published September 26, 2025 by
Joe Kannapell, PE
Joe Kannapell, PE

Can truss software work well without being bundled with the purchase of truss plates? That is a 46-year-old question that many have asked over the years. While I haven’t evaluated the newest software-only offering, I did personally evaluate two prior offerings while I was working in a truss plant and two while working for MiTek. To that experience, I can also add that I have been immersed in software development since studying it in school, applying it as a programmer myself, managing its development, and writing about it for nearly a decade.

Some of the hard lessons I’ve learned about software development are:

  • To make good software, you must have had experience making good software.
  • If you have no experience in crafting a similar task, you can’t predict development time or cost.     
  • You don’t know if your software is good until users test it.
  • The biggest and most diverse group of users will yield the best software.
  • 100% consistency and reliability, which are both essential, is impossible to achieve.

All of these favor the largest and most experienced incumbent, and that’s why there are only a few software companies that dominate the specialized areas of computer technology – Microsoft in operating systems, Oracle in database management, Salesforce in sales management, etc.

Now let’s review the history of software-only offerings in the truss industry:

1979: Forest Products Inc. (FPI), introduced by Lenny Sylk, succeeded because it was based on software developed by Sanford, a leading plate manufacturer, and due to its early adoption of the first desktop engineering-capable computer, the HP 9845.

Mid-1980s: Merlin program succeeded because it was developed by two PhDs, Frank and Paul Little, who had extensive Alpine experience, and due to its early adoption of generic IBM-compatible PCs.

Late 1990s: A.C.E.S. layout program succeeded because of a brilliant programmer who had experience at Gang-Nail, the largest plate and software company, and because he applied the best mathematical model to solve an essential task largely overlooked by the incumbents, 3D truss layouts.

The result: None of these programs operated independently for more than a decade. That’s right – even though these software-only offerings were successful, they did not remain independent companies over the long term.

And then there’s Warren Buffett’s judgement. He bought MiTek primarily because of the product, truss plates, and only secondarily because of the software. We all know Buffett’s decision has paid off over the past 25 years, as I can attest as a former shareholder from my time at MiTek. Based on his actions, I would argue that Warren Buffet would answer the question I posed at the start with “no.” Whether or not he would be wrong in the latest case remains to be seen.

Photos from BCMC Louisville circa 1972 — left photo: Chris Dudek, PE is on the computer and Lenny Sylk is in the background; right photo: Keith Myers (current GM of Woodhaven Lumber Truss & Panel Operations) is checking out the computer screen in the FPI booth.