The UX Research Reckoning is Here

When I did the human interface course in my comp.sci degree, there was a lot of attention on standardization, making things people could effectively interact with, consistent feature widgets that are clearly those widgets. Remember "HIG"s?

Nowadays, instead of trying to find a way to make an interface pleasant and consistent, it has to be a unique butterfly of some glorious design that just sucks.

Next time you're designing something with a button that doesn't have a border and isn't clearly a button... Step back and have a rethink. Radio buttons indistinguishable from checkboxes? Your design is bad.

Now get off my lawn

What, why are you hiring this many people..

>Incentivized to empire-build, some thought UX leaders (especially at a few larger companies) were disproportionately responsible for over-hiring and creating the market we are in.

…ok, the jokes write themselves.

I just couldn’t get over how full of shit the UX industry seemed to be. It seemed in thrall to itself, and vastly over-rating its own depth and complexity. Sorry for the negativity. I have huge undying respect for dedicated HCI research experts and massive admiration for talented visual designers. I just couldnt shake the impression that a lot of UX had become a blag, a methodology to sell to business like agile etc, and wasn’t on solid ground.

I’m not saying there’s any real insight in that but it was the opinion I formed and this article seems to give some degree of validation to it.

The idea of HCI research and usable product design still strike me as noble & important fields of endeavour. I’d hope if you’re a sceptical UX researcher then we’d be in some kind of understanding & agreement. I admire what it can be.

In the 90s most people had a very hard time using a computer, software came with thick manuals and often they found themselves in situation where needed help.

Nowadays people from all ages use software for many aspects of their life, they interact with their banks, shop online, do their taxes, communicate, interact with their governments... through software. People ranging from kids to elders. Expecting to open an app/web and being able to use it directly, without reading a manual or a help center. This is a great amount of progress.

Over this, you've to take into account that often there's a big gap between the design and the implementation due to many causes (business requirements, deadlines, committees, hurry to release, technical limitations...)

Of course, the field it's not perfect and there's many designers that are not great, or are entitled pricks. But I'd say that design (call it UX/UI, product, HCI, IxD, whatever flavour of the day) has added a lot of value to the table and helped make technology as ubiquitous as it's today.

One knowledgeable individual combining the fields programming, design and usability can sometimes outperform big teams, because there is something to the way big teams decide that dillutes clear ideas and leads to inconsistencies throughout the resulting piece of software. These inconsistencies can be an problem for the user.

UX people generally may claim otherwise, but it is easy to make software without them and often the user experience is better when they are not involved. Imagine two situations. We are going to improve the user experience of the famous text editor vim. Who will make a better job:

1. Random dude who has been teaching vim to people, with some experience in creating user interfaces and a lot of experience with terminal based software

2. UX researcher armed with all the newest knowledge in usability

My money would be on the former, because the latter has a high chance of misunderstanding why people use vim in the first place. I might be mistaken, but my feeling is that in the past 15 years user interfaces have broadly been "dumbed down" or driven towards the smallest common denominator.

That is okay for software that you use once a year and where things should be self explainatory. But there is software where it is okay to have a little hurdle in the start if it means you are going to be so much more productive later. Ideally you will have both a low entrance barrier and a good usability when you use it day in, day out. Newer UX-based software is good at the former, but sucks at the latter.

The way I was finally able to get through to the design team was by showing them an image of a Tesla interior and saying "this is what you guys have built", and then I showed them an image of a well-organized woodshop with hundreds of tools on the walls and saying "this is what the people using our software to build custom apps actually need".

Maybe non-technical end users really do like having all features hidden beneath unlabelled flat buttons, I don't know. But it doesn't work for people using software GUIs to do technical work. I'd go back to 90s-00s style clunky menus in an instant if I could. Just the ability to see what's clickable and what isn't is worth the downgrade in appearance.

Ironically, the author is quick to provide answers that please executives, and slow to offer dangerous questions that would displease them. Questions like how the company's short-term optimization was causing a long-term loss on user retention. Or how certain executive-sponsored beliefs were hurting user trust. Researchers who ask those questions tend not to stick around too long, even before 2021. Those who remain self-select into safer "middle" research; they and their department have been pummeled into submission by the organization.

The author turns a blind eye to this corruption; the author points the finger at researchers, not executives. The author also forgets the history that users made the business successful [1]. Today, however, the internal hierarchies and forward momentum of many tech companies depends on a selective understanding of users [2]. Good research slows down that momentum, it aims for good-enough understanding, it asks important questions. And when important questions collide with the status quo, they are not tolerated.

Or to put in simple terms, from a chief of staff to a middle manager: "Nevermind you 'why', why ain't in your repertoire no more".

[1] http://www.paulgraham.com/before.html "Mark Zuckerberg didn't succeed because he was an expert on startups. He succeeded despite being a complete noob at startups, because he understood his users really well."

[2] https://news.ycombinator.com/item?id=35597279 "The logic that the CPO applied at the time was some pretty sporty mental gymnastics: "we optimize for conversion (looker to booker) because that's a proxy to actually delivering value to the customer."

Academically, 98% of UX people don't have the requisite background to be doing what they are doing.

What the? I had no idea it was so colossal. I didn't even know UX "research" was a thing, as in, someone made their career out of it.

No wonder it takes 100 engineers to write a website, if it takes 100 researchers to dream up some coconuts for people lacking thumbs. Last I checked, humans haven't changed in centuries. The gestalt is all you need to "research" and understand the human perception, imho. Ok maybe in the field of VR, some research need to be made, but this seems like crazy amounts of people.

As a big believer in design, I think the field of UX has collectively allowed it's value to be tarnished by not guarding who is skilled at UX and who is not.

We developers got to be gate-keepy for a long time with the ubiquitous "CS Degree" requirement. Obviously with the ultrademand for programmers, that requirement is vestigial at best. But we have a very established bar for quality of engineering.

The solutions (on here and in the subsequent reflections article) are hilarious though for how general they are and how little they solve the observed problem (UXR dying/being replaced/not justifying its role/etc.). "You'll be fine: just generate value for the business and learn to say no!" The sort of fluff you tell a naive new graduate about to start an internship in a sinking ship of a company to avoid facing reality. You have to read between the lines to realize how devastating the advice actually is.

- Not engaging enough with the customer, i.e., finding out what they actually and truly need as opposed to pandering to their egos because they’re SMEs or whatnot - Engaging too much with the customer by allowing them to have too much say in the design. I’ve seen it go so far as to allow the customer to actually make their own designs (wireframes or mockups, etc.) and then handing that off to the devs to implement, no questions asked, which ended up in a ton of rework for obvious reasons - Relying too heavily on frameworks by allowing yourself to believe that you MUST use one of their built-in widgets or patterns and not create a custom solution that makes sense for the use case - Somewhat contrary to the above, not relying enough on a given framework’s widgets or patterns which can lead to inconsistent design and UX and unnecessary reinventing the wheel - The use of so-called no-code or low-code solutions: Not that these are bad per se, but it leads to the fallacy that a non-engineer (usually mid-level IT manager or “system analyst”) can use these tools to create a complete or even somewhat viable solution (I speak from experience) - Sacrificing a good design for any reason, e.g., ad space requirements, various types of compliance (including even the virtuous ones such as Section 508) - Relying on a framework to do the UX for you, leading to another fallacy that you don't need a UX specialist

UX research still very much needs to be a thing, and I don't say that because it benefits me (it does), but because I wholeheartedly believe in the need for it. I wouldn't design a house or even a pool without the expertise of an architect, so why would an entire dev team build a multi-million dollar UI without a UI/UX engineer?

It can't decide whether it's science or fashion and kind of fails at both. Flat vs. poofy buttons is a fundamentally unserious debate, especially when there's no real barrier to letting the user choose (and not infantilizing them).

The "solution" is understanding that what we're talking about here is fundamentally "communication" and most of these things can be solved through semantic ways of defining these things, e.g. what HTML tries to do.

A lot of UX design has been formalized to the point where the UI frameworks really do provide just enough. Even better, the UI available in native options (browser or native apps) has gotten Really Good.

Really Good is Good Enough for most consumers, and perfect for businesses/enterprise.

I agree too that UXR is disproportionately impacted, and that UXRs focus way too much on user interviews. The challenge for me is that “macro research” is really market research, and UXRs are not well-suited to conduct this type of work, nor do the teams need to be so large to support that kind of work. I feel for anyone losing their job, but it’s tough to justify.

Consider Amazon. Not once in their history did they have a "proper" UX. It's always been ugly, unpleasant, inconsistent, messy. And yet they rule. Craig's list is similar.

These serve as examples that bare bones UX combined with solid business essentials work. It doesn't mean that UX is a luxury in every context, it's a wake-up call to link the work to real business outcomes.

Not to mention, to stand out from the bland average. In this so-called "golden age of UX", I honestly cannot remember a single case where I thought "wow, this is top notch UX, clearly well researched".

First of all, I don't fully recognize TFA's description of what a UX researcher primarily does. It almost sounds like user testing, which shouldn't require a different hat. That's just something you do to evaluate the product. Designers and developers do that anyway (hopefully).

What I did when I put on the "researcher" hat was (I guess, after reading this article) more akin to market research, but within the existing market.

For example, while working with a big holiday/vacation company (operating flights + resorts), we uncovered that a significant portion of the customers staying at their "family friendly" resorts didn't want to be there. They were typically couples who had done fairly adventurous trips when younger, then they'd married, and had kids. They hated the enclosed, catered, walled feeling of these resorts, but thought it'd be the best they could do with a newborn on their hands.

Awareness of this group of people might feel obvious today, but at the time it was a genuine discovery that lead to development of both services and marketing. While you might argue that the company gets their custom anyway, so what does it matter, the fact was that these people left the resort with low ratings, and a low opinion of the company, which they would announce to their peers in $important-demographic on social media and verbally. I.e., long-term damage to brand and customer retention. They freely chose to stay at the resort, and then felt resentment towards the company for it not being a mountain climbing holiday or whatever. It's illogical, but that's what they did.

Secondly, these types of discoveries can't be produced predictably. Either there is something you don't know about your customers or competitors, or there isn't. Initial efforts might yield a lot, but perhaps there is a saturation point along the way, where more research of the same nature at that particular point in time will yield diminishing results? Well, that doesn't matter, because these people have been hired to do research, so they will keep doing it, whether it's useful or not.

I don't understand the separation of design and research into different roles. To me, a designer should be able to go off and do research if it's warranted. Maybe I'm old school, but separating it feels like separating a developer into two people: one whose only job is to find new libraries and read documentation, and one whose only job it is to type into an editor.

Thirdly, to those griping about "UX ruined UI": it did not. UX is what got it to the point you're imagining as "good old days". What damages UX in the long run is shifting incentives (resulting in dark patterns) and a variation of the above, but on the design side: you needed a UX team of a certain size to get to the design you have, but you don't need an engine of that size to maintain it. If your UX department is oversized, I suspect that they will start inventing tasks for themselves and do things that aren't strictly benificial.

This goes for all roles, arguably. More people doesn't equal more productivity, and at some ratio of number of people vs. real problems for them to tackle, the productivity becomes negative.

Also, a UX department as some kind of "other entity" is not a good setup either, in my opinion. In my experience, UX has worked best when it's a cross-disciplinary focus rather than an isolated group of individuals imposing their will and "owning" the UI.

As a UX designer, your most important job is to get people who don't regularly talk to each other into the same room. You don't have to be the cleverest person in the room wrt UI, but you do have to be the person who can bring the user's point of view into a room where other, equally important points of view, are represented. UX can't be effectively done in isolation.

When your profession views math on user & business data as the enemy, and it's for a business...

The problem is the rise of Career Designers™ as opposed student of the arts. The current pipeline is dominated by candidates who realize there are few careers for liberal arts students paying six-figures and also self-aware enough to know they don't enjoy math/programming/excel (although many cashcow masters provide OPT when they don't even any math or even the rigor of a respectable MBA). Google has a propensity to hire people who hold master degrees specifically in UI/UX, but I've found only a small fraction (maybe 1/10, and definitely much lower than 25%) who has any aesthetic intuition. I've seen their non-portfolio work, such as their photography and clothing choices. The average teenager on social media probably have more artistic sensibilities than people who are supposed to be the top-tier designers. Obviously this is subjective and also a proxy metric, but I'd like to believe the ability to compose whether it's photos, interior design, or clothing is strongly correlated to the ability to design a website. Or even just having an appreciation of cinema that isn't Marvel. We end up with people who follow systematic procedure taught in a masters in UI/UX that leads to lots of and lots of overly complicated design patterns that make sense intellectually and in isolation, but will never arrive at something that is hollistically cohesively and usable like Craigslists.

Google has grown into a corporation and select for career oriented students who are studious high achievers who focuses on practical matters rather than frivolous things. Maybe that algorithm works well for programmers, where in math/programmer there is a yes/no, but even then school can't indicate who has good intuition about design. I mean the author of Vue is a design major, not computer science.

The people who live and breath art aren't being hired by tech companies to be designers, especially the large ones. They don't fit the corporate status quo, but then again they tend to be independently wealthy and never had to treat academia as a means of procuring a job anyways. At least for hackers / cybersecurity / introverted programmers types, tech companies have done a good job at assimilating people were in social outcasts in early decades. The design orgs feels much more political and feels like a grown-up version of a high school popularity contest. Google's recruitment process of emphasizing academic prestige (school, grades, advanced degrees) may work well for other areas, but it's filtering out the best candidates, and selecting for people who are frankly mediocre.

In the early days of Google, when Google was a pioneer in design, design was lead by Marissa Meyer, an engineer who had a knack for design. It she an artist? Maybe not, but Google was a pioneer in website design and it's at least in part attributable to her. Engineers who have design sensibilities are probably better for designing websites / apps than career-designers.

So that's UI designers. UX Researchers are supposed to be scientific and minor statisticians. They're supposed to glean test hypothesis and make unbiased judgements from a representative sample of the population. The UX Researchers I've worked with were some of the most unscientific and non-quantitative people I've worked in my 10+ professional years. The math stops at basic percentages. I observed some of these surveys and user studies where they prime candidates and ask leading questions towards answers they've pre-assumed, rather than actually listening to the user. That's why there is such a huge and growing disconnect between Google products' usability and the user. Most of the people who advance in this org are unsurprisingly great at talking.

OR (and I say that in the angriest way I can manage via text) the business is completely looking out for itself in the short term without consideration of its needs beyond a quarter. When a department has served its useful purpose to the very specific goals of the business then the department gets discarded. When a company needs a design for a website, it pays for a UX designer. When the designs are finished, they fire the designer. Does that mean the business never needed a designer? No, instead, they believe that they _no longer_ need a designer, and continue to profit off of the work that the designer did for them. The designer then gets to leave and write blog posts blaming themselves for the pitfalls of capitalism and the absence of design unions.

There is no way to stay around a hungry tiger and survive for long, no matter how one ingratiates themselves to the tiger. Designers, don't write blog posts claiming that getting fired is 100% your fault. You're not Jack Welch, you're the body he spent no time thinking about leaving behind.

I think that generally, UX in software is dismal. Modern software tends to be hard to use, hard to figure out how to do things in, and generally opaque and lacking in discoverability.

I tend to blame the UX field generally for this, as that seems to be where the things leading to this problem have come from.

I disagree with the author, though, in his diagnosis that the issue is a clash between capitalism and UX researchers. I think the issue is that the UX field itself tends to focus on things that don't really matter as much in terms of user experience. There's a huge emphasis on efficiency, for instance, rather than reducing cognitive load, making sure that workflows all follow a common logic, and making things discoverable.

I don't know enough about the field to diagnose anything, but this smells suspiciously like a failure of UX research to focus on the most important things, and UX developers simply follow that research.

We pitched this idea to a mentor at the hackathon that worked for the sponsoring company and had worked with UX. The question he asked "how are you going to make your app sustainable, how are you going to keep a steady flow of new users?".

We replied: "because it's convenient", and he immediately replied: "normal users don't care about convenience, they won't stop using 5-10 different apps because it's convenient for them to use a new app where they can find everything they need. They need other incentives to use your app".

At that point I couldn't reply back to him saying "well I am a user, we are users of these apps and the UX is simply terrible, if you think this is okay then sorry but you're a dumbass", because we had to finalize our business pitch and we were in a tight deadline, so for nervousness I couldn't say it to the guy that he is wrong.

I don't hate UX. I hate people that think it's totally okay to hide important or common things behind 3-5 clicks AND screens. This is the golden standard for UX nowadays: making things complicated for the end user. I don't know if I am just tired of this bullshit they shove at our faces, but a LOT of new major products have a terrible UX now.

Examples? I use OneDrive for storing my files. I mainly use the web interface because there's no native Linux client and I keep forgetting about rclone. I used to access office.com, login to my uni account, click on the OneDrive icon on the sidebar and it would redirect me to my OneDrive.

Microsoft changed the layout recently, now to access 1Drive, you need to click the sidebar, click on the menu icon, select OneDrive, to a point it will redirect you to a OneDrive page, and then you have to click "My files" to get the default view where all your files are listed. I hated this update and I have no idea what Microsoft was smoking to think to hide the OneDrive button under extra actions.

Another example? FIFA 2023. You have lootboxes that contains player cars and utility cards. You can get these card packs by completing objectives. You can complete objectives by doing actions on the main menu (organizing your team, buying a player, etc) or inside a match. And guess what, there's no button to collect all of your pending rewards. And if you have lots of card packs to open, you can't just click a button to open them all, you need to open them one by one. Oh and to make things worse, if you go to any other Ultimate Team menu option, you need to go back to the main Ultimate Team menu, click Store, Packs and then you can open your packs. The game costs $70 and has terrible UX.

Only organisations with a specific cultural focus on user centrism will have great products.

Apple at least has some of this, but not always.

Have faith: there is a payoff in the long run for many product focused companies. Not all though.