By 2016, Don Norman had publicly articulated how misunderstood the term "UX Design" was. Looking at the degree of evolution of UX design in the modern sense, I don't think there is a better description than the originator of UX design can't define the word. To understand this, we need to start with the history of design and development since the 1990s, which are intrinsically linked.
Traditional User Experience Design
The purest form of UX design is based on the waterfall model:
(Image courtesy of Swipecubes)
A product team working through the waterfall model needs to learn everything that goes into the process before it starts building the simplest prototype. This preliminary preparation work takes months or even years to complete, and the results of the preliminary research determine the practical implementation of the design team. Product requirements need to be locked before design, and design prototypes need to be locked before development is executed. There is no turning back until version 2.0 arrives. This is how the so-called waterfall model works.
The traditional UX design process is usually taught to college students in this way:
Do more research to find the problem;
Categorize the problems you find;
Create user profiles and behavioral paths;
Run ideation exercises to stimulate new ideas;
Create and test prototypes;
Take the final prototype for development;
launch your product;
Return to step 1 based on user feedback.
This is a basic waterfall model. The classic elements of user experience design also follow the waterfall model, building and executing from the bottom according to clear needs.
But it can also be seen that the classic user experience design is fundamentally incompatible with agile and fast.
Towards a lightweight direction
Innovation in Silicon Valley has long been driven by Moore's Law, which states that the number of transistors in a dense integrated circuit doubles roughly every two years. Superimpose the evolution of a waterfall model, and you'll find it's perfect for a two-year, 24-month cadence. The business, design, and development cycles work as punctually as you can with Swiss clocks, perfectly fitting the release cycle of new Intel chipsets.
But suddenly one day, Sony, Toshiba and IBM (STI Alliance) felt that Moore's Law was too slow.
The STI consortium created the first single chip, which stacked 8 microprocessor cores on a single wafer, and this shit actually worked and got to b2b data work (it powered the Playstation 2), and the multicore architecture has changed since then everything.
Moore's Law is not broken, but its life cycle is threatened, and it is the waterfall model that is actually broken.
Almost overnight, speed and flexibility replaced the precision and predictability that have long dominated the competition. Companies are turning to another mature but underutilized development method across the board - lightweight development.
Lightweight products focus on rapid iteration. It releases a new set of incremental feature releases every 2-4 weeks, revising the product over time, rather than rolling it out all at once. It is based on hypothesis, experimentation, rapid release and real-time measurement. There is no editing phase in the process of lightweight development, there is no perfection. As Bre Pettis said in 2009 "the completion of one iteration is the impetus for more refinement".