Summary: UX is evolving rapidly, with AI integration and a shift towards distributed expertise. UX will be like a pancake: spread thin but still delicious. As the field expands, career paths will flatten, but opportunities for expertise and impact will grow. Embrace these changes to stay ahead in the field.
The UX profession is melting down, but in a good way. We’re abandoning the ivory tower to live among the peasants, and some old-time UX elitists don’t like the smell of the countryside.
I am an elitist myself. Two of the best jobs in my 41-year career were in places that deliberately limited recruitment to the top 0.1% of IQ in the world: The Bell Labs successor of Bell Communications Research and the team of Sun Microsystems Distinguished Engineers. In such scenarios, a limited super-elite group can leverage a bigger, less-elite staff to monetize their insights (a monetization approach that also works for UX consulting). Sun Microsystems made much more money from the Java programming language (invented by one of my colleagues) than the cost of running the entire Distinguished Engineers program for 20 years. And the cost of running Bell Labs for 80 years was more than recovered by just one of its inventions: mobile telephony.
However, scenarios where you need super-elite work are rare. In many more cases, the difference between zero UX and good-enough UX is all that matters. This difference is immense, so we need every company to embrace some amount of UX, which requires UX to be cheap enough for everybody to afford it.
Only a few companies can afford elite-level UX, which is expensive. But luckily, few companies produce so groundbreaking products that they need expensive UX. Most projects can ship great design at much lower cost. UX costs are dropping rapidly, making it feasible for the number of companies doing UX to explode by several orders of magnitude over the next 30 years.
The benefits of using AI for UX are threefold:
Current AI increases knowledge-worker productivity by around 40%. I expect the productivity of UX staff to double over the next 10 years, meaning that the cost of any given UX deliverable will be cut in half. When something is cheaper, more companies will buy it.
AI vastly increases the creativity of people who use it in a symbiosis-like manner, combining human judgment and direction with raw AI wild creations. With AI, ideation is free, which is a boon for UX.
AI lowers skill gaps and serves as a seniority boost, giving everybody a much wider repertoire of good-enough skills than we had in the past, where the ideal was “T-shaped” UX staff who could only dapple when extending beyond their one core skill. The new ideal is “UX Unicorns,” who are people with a wide range of skills that cover the full UX process. When one person does everything, communication overhead disappears, as does the need for UX personnel to waste half their time on persuasiveness tricks to make other team members pay attention to UX findings.
Commodifying UX
When I started in UX, there were about 1,000 UX people in the world. We were a rarified priesthood of the few people who knew how to make computers easy.
Later, when I was a UX consultant during the dot-com bubble, I was one of the only people in the world who knew how to make a website easy, and I charged super-premium fees as a result. One day’s consulting and I could double a company’s online business. That’s how bad websites were back then.
When the Web took off, there were many UX people in the world, but they looked down their collective noses at something that primitive. (History is repeating itself with AI, which is also ignored by most UX people and being designed by engineers.) The early definition of web design was left to advertising agencies, who knew how to make pretty brochures but knew zero about interaction design. I was one of the few who was willing to work on web usability, and I cashed in accordingly.
Soon enough, Internet executives started reading my books and articles and gave more priority to web usability. Thousands of UX jobs were created. But that doesn’t mean that things were happy: users were still poorly served (if not as poorly as before) because UX was designated as a separate, rare thing closely guarded by a few experts.
The decreasing cost of UX is transforming it from a luxury service to a widely accessible commodity. Some people dislike the word “commodity” because it implies cheapening. For the world, this is good. UX must be cheap to be widely deployed. For people like me who used to charge a fortune for UX consulting, it may be bad that anybody can do UX these days, but it’s great for users.
Commodities for sale in a spice souk. Having commodity spices means that more people get to enjoy tasty meals, and it’s the same with UX: now that UX is no longer exceptional, more users will enjoy better products. (Midjourney)
My allegiance is to humanity: to make computers and technology so easy to use that humans are the masters, rather than being oppressed by overly complicated user interfaces. Meeting this goal requires UX to be a commodity. Simply the way things are done as a matter of course. Something millions of people can do.
For usability and design quality to permeate technology products, UX can’t sit in a corner by itself and virtuously complain that it has no influence. It has to be integrated with development and seen as the best way to make money with technology. In other words, we’re easy to work with, we are cheap, we deliver profits, and we’re the way things are done. A commodity.
Pancaking UX
My even more radical prediction is that the UX field is pancaking. By this, I mean that it’s being spread out across the organization in a flatter way than the hierarchical centralized UX teams that used to dominate.
Pancaking UX means that our skill and efforts are being poured out to form wider, but thinner coverage. Because of improved efficiency, this thinner coverage will suffice for quality results, just as it's yummy to eat even the thinnest crepe. (Midjourney)
Traditionally, we had a pyramidical, hierarchical structure of UX work, with the UX staff put aside in a separate centralized team. With growing UX maturities in many companies, this centralized UX can become quite large and number hundreds or thousands of UXers, all under the management of a big-cheese UX VP.
The career goal for an ambitious UX’er in this traditional structure is to climb the management ladder, from manager to senior manager to director, and hopefully end up as that one VP on the top of the pyramid. Of course, only one person out of 500 can actually be the boss of that 500-person organization, so it’s not really a realistic goal to make it to the top.
Left: Climbing the traditional tall management ladder in a hierarchical organization. Right: Pursuing expertise ladders and design ownership. (Ideogram)
Conversely, with pancaking UX, we distribute UX expertise throughout the company, rather than hoarding it within the purview of that one Grand Poobah of all UX. We do away with all those expensive titles of UX Director and UX VP, though we may still have some UX managers in places where a specific product needs a large UX team.
Mostly, though, pancaked UX consists of UX experts applied exactly where they do the most good: as tightly integrated members of each product team.
What’s the career progression for the top talent now that they can’t aspire to become Grand Poobah (AKA, VP of UX)?
As shown in my illustration, the one management ladder will be replaced by multiple expertise ladders. Instead of managing ever-more underlings, the progression for the best talent will be to own larger scopes of design.
A very junior designer might not own anything, but be solely assigned to assist a senior UX’er. Later, after acquiring a bit of expertise, a junior designer might own an icon or a few buttons. Gradually, a good UX’er may grow to own a dialog box, a full feature, and an entire application. The most senior staff will get titles like UX Principal or UX Architect and own a company-wide design system or UI architecture or be in charge of the company’s UX vision.
With my predicted UX pancaking, are we under-resourcing UX? Not at all. The UX staff will be spread thin around the company instead of being lumped together. But there will probably be more of them altogether since they will have more impact on the company’s profit as they are more tightly interwoven with the products and rewarded more for talent, skill, and expertise, as opposed to their ability to navigate a corporate hierarchy and grow their budget.
With real pancakes, your food intake does not lack calories, even if each pancake is thin. Similarly, with UX pancaking, usability is fattened up.
Even if each pancake is thin, a stack of them sums to a delicious meal with more calories than you really need. (Midjourney)
Case Study: Amateurizing ResearchOps
Fidelity Investments is a huge brokerage in the United States, with about 61,000 employees and about $5 Trillion under management. Jen Cardello is their Senior VP of UX Research (and a former colleague of mine — in fact, one of the 5 most talented UX people I worked with in my 41-year career). A recent case study documents how she led an effort to broaden user research across this immense organization.
Cardello’s ResearchOps team trained over 300 non-research staff in unmoderated user testing, meaning that their teams didn’t have to wait for the official user research teams to get around to running a usability study. As a result of this training, non-researchers conducted over 600 usability studies.
Even people who agree with the famous doll proclaiming that “math is hard” will quickly realize that this means that the average non-researcher taking the course only conducted two usability tests. Clearly not enough to become an expert. But that’s the entire point! User testing is easy enough that any good professional who works on a development team can learn the basics of running simple studies to get quick feedback on his or her design. (More complicated problems will benefit from deeper usability expertise, and Fidelity still has such people on staff.)
Efforts to empower staff who are not full-time UX specialists to conduct their own UX research or design are sometimes referred to as “democratizing” UX, but I dislike this term. It’s not a matter of deferring to majority rule, so that if, say, 3 people vote for design A, but only 2 people vote for design B, we go with A, no matter the expertise of the 5 people voting. In fact, if we had no data and did need to decide based on a vote, I would give more weight to the input from people with more UX expertise.
If we need a word, I prefer “amateurizing” to describe such efforts to support non-UX staff in performing simple UX work on their own projects. “Amateur” is not a dirty word; it simply refers to a person doing something for the love (amour) or appreciation of it rather than as a full-time profession. As UX methods become more systematized and AI-supported, they also become easier to perform, meaning that more and more amateurs can carry out the less complicated parts of the UX process.
User testing is the most basic of the user research methodologies and the one most amenable to amateurization. As long as you shut up and don’t lead the user, you can watch and see what happens. You can’t help but learn something. Unmoderated user testing (i.e., facilitated by a computer instead of a human) is even easier than the setup depicted here, but you tend to learn less from each session. (Midjourney)
New UX Career Paths
My advice is to recognize the current commodification of UX and the coming pancaking of the field. Stop dreaming of Grand Poobah status at the top of a pyramid of centralized UX. Start planning for UX Architect status and cultivate your ability to work in a distributed manner.
The following infographic is an example of how I hope these trends will work out for the young people in our field, even as some oldtimers may lament the decline of the Old Ways.
I made this comic strip using Umesh’s Story Illustrator GPT, setting the style to “3D-rendering animation style.” [Feel free to copy or reuse this infographic, provided you give this URL as the source.]