Re: @Krzysztof's answer, the range of types of products and platforms that are possible is extraordinary - ranging from smartphones and embedded software to enterprise applications. You say "I'd like to have a software product for my consulting company" - it sounds like you want this as "social proof" for your company in order to promote your consulting services. Unless the product is extremely well aligned with the area of your consulting, I really don't see how this could support your consulting business. In fact it sounds like a huge distraction from everything else that needs to be done.
However, I get tired of answers that say "you're asking the wrong question" so assuming that you have your reasons for doing this - a few random thoughts:
Developers, like other independent business people, tend to attribute too high a salvage value to their abandoned products and efforts. I would not expect to find many bargains. I'd expect to find bargaining dynamics somewhat like trying to buy an obscure domain name from a squatter.
So much software dies because the platform shifted away from it and the developer had no intention of porting it. So the DOS or Windows 3.1 or Windows 95+Win32s API code may require so much reengineering (as already pointed out in the thread) that it is not worthwhile to re-use it directly.
I will surmise that software that was not abandoned only because of OS platform shifts suffered from at least one of the following problems:
- Too late or otherwise uncompetitive - it missed market windows and no longer made sense to sell (this happened to a couple of products that I developed for past clients - they kept adding features and the market window whizzed right past.)
- Was not the right product - it did not address a problem that anyone really wanted to pay to solve. (Again, I have experienced this as a software contractor.)
- Was marketed and/or positioned very poorly - it was being marketed to the wrong groups, or was not differentiated properly. (Ditto - this is fairly common too.)
I will also guess that the most viable acquisition target would be a business related (B2B) product that is written in a relatively modern language like .Net or using a neutral platform like C++ with Qt. It would fit the last description above the best - it could have been sold successfully as a product but the owner did not market it competently.
I would surmise that this combination of characteristics would also make for the most expensive acquisition. People usually know what they have in hand already.
Lastly, to find such products, you will probably have to market your need to groups of ISVs, and make yourself known as someone who has the budget and willingness to buy such a product's assets.