A method of systematically and effectively identifying new markets and product lines.
You start by identifying experts in the field using a literature search. Then you use the experts to help you find the lead users in your industry or, more usually, some related industry. These are the people whose needs are unmet and so have had to devise their own solutions. These are not early adopters. If someone has bought your product or service and is at all satisfied with it then they're disqualified.
Once you've got the lead users, you sit them all down and combine their knowledge, insights and prototypes into an entirely new product line. It's unusual that one lead user will have a complete marketable product as a prototype or design, but it happens.
This method has been successfully applied to products, services and processes. In fact, the lead user method was formally discovered by an application of itself. When 3M found EricVonHippel, they discovered that they were the lead user of the method and had to devise a prototype in collaboration with him.
The lead user method is straightforwardly applied to software. A lot of exceedingly crappy software wouldn't exist if the method were applied. For example, had a software developer taken the prototypical world wide web and actually designed a product to meet the need it represents, we wouldn't be stuck 20 years later with those broken protocols.
The only way that non-application of the lead-user method doesn't result in horrible software (barely working prototypes) becoming widespread and locked in is if the lead user is also a competent designer by some freak accident. We're talking about a convergence of three radically different skill and knowledge sets here, at least two of them mutually exclusive given ProgrammingVsDesigning.
This explains the general miserable state of software. And when I reflect on that miserable state of software, it makes me (a designer and lead user multiple dozens of times over) consider letting programmers suffer out of spite. After all, it's your own damned fault for not being professionals.
Note that non-application of the lead user method also explains the miserable state of CPU designs, which are built with early adopters or even the mass market (ie, mainstream software and languages) in mind and not lead users.
-- RK