Startups, this is how design works (2012) | Hacker News Reader

Startups, this is how design works (2012)

startupsthisishowdesignworks.com

78 pointsBrajeshwar94 comments

If you've ever seen those Kumon tutoring places, their windows usually have the slogan: "Math. Reading. Success." I always joke to my wife that people should skip the math and reading classes and just take a success class.

Obviously, "success" is real. We can say very clearly that some people are successful, some products are successful, some projects are successful. But "success" isn't a unified thing that you can teach. What makes it possible to be successful in a given domain is different in different domains. To be successful in a foot race and to be successful in stock trading require very different skills, such that you can't teach a "success" class that will make you successful at both a foot race and as a stock trader.

And further, in a lot of endeavors, being "successful" requires not just particular skills to that endeavor, but a lot of different skills. Making the iPhone successful was about manufacturing at scale, software, hardware, marketing, supply chain management, contract negotiation, etc.

Let me suggest that "design" is more like success than it is like math or reading. A product can be well-designed, but I'm dubious that there are unified "design" skills, especially ones that translate across a fairly wide domain. What makes a product well-designed is much more diverse than what people who talk about design typically admit, and I'm not sure that design as a field really makes any sense.

I would define "success" as you've described as the combination of cross discipline work and market forces which produce a desired outcome.

Having good design however doesn't depend on market outcome. You can have good design where noone is using it.

I would say that good design requires cross disciplinary skills or knowledge - requiring a core set of "design skills" plus an understanding of psychology, the end user and the domain it's for. The successful emergent outcome of this is "good design".

The difference between math and subjects such as design or stock trading is that math isn't judged by humans or human behaviour. Design will at the end of the day be judged as good or not by the people that interact with it. Stocks will be judged based on whether others are buying or selling. The skill in these things is predicting human behaviour, following patterns and managing risk.

That's why in design it's a good idea to meet your users, and to iterate following rounds of user testing.

It depends on how you define design. Herbert Simon defined it as “to devise courses of action aimed at changing existing situations into preferred ones.” And there’s a lot of academic work around that.

The problem is that that’s not how the word is used in practice. I think most people would call that “planning”, and certainly design is a type of planning, but not all planning is done by people who we call designers.

I think a better definition that’s actually used in practice is that design is a way of looking at a product as if it was a kind of communication. The central question is “how will people perceive this?” so any type of art skill—not just visual arts, but theatre, music, writing, etc—is potentially transferable.

Communication is a universal human skill, so it raises the question of why a separate discipline has emerged in companies to handle the communicative aspects of the product. And I think this is because many people who work in tech fields are educated in fields with very specific conventions of communication, and an expectation that the burden of understanding lies with the listener. When these norms inform the creation of a product, especially when it’s aimed at non-technical users, it leads to “bad design”.

Contrary to what people think, good design is explainable, though designers may do most of the creation in their head. It's like asking why some code should be written this way, there's always a good explanation for good code.

Good design comes from a process, UIUX design for example is a result of product design, which is a result of user needs analysis and requirements gathering in the form of understanding a problem domain deep enough to know what is wrong with the current design, or what is expected of the functionality. This is not about vague "design thinking", but actual auditable design processes that lead to a good product.

> I always joke to my wife that people should skip the math and reading classes and just take a success class.

http://www.harkavagrant.com/index.php?id=206

"Your best bet is to be better than everyone else. That works for me, so that's my advice."

I was just thinking that this feels like it's from like a decade ago - and then boom, 2012 at the bottom. Still good points though. I do believe design also has a culture/trend ebb & flow to it, so stuff we see from 2012 actually looks dated - just like music, comedy, movies (etc.) can also feel dated.

Feels very dated. They don't even have Figma in their tools section and are still referencing things like Flinto and Skitch(!)-- a nice throwback to when I first entered tech 10 years ago.

You can tell this was written when design-thinking jargon was at its peak. The retelling of Rams' "10 principles" is always funny to me because he never "set out" to write those rules. For the uninitiated, here's a summary:

Starting in the late 70s Rams talked about the design ideas that led to the success of Braun, specifically three rules: a rule of order, a rule of harmony and a rule of economy. Later he would be describe these further as "function and performance, feasibility and aesthetic quality". Years later, his reflections on his contributions to Braun became six principles: "1. Good Design is innovative, 2. Good design renders utility to a product, 3. Good design is aesthetic design, 4. Good design makes a product easy to understand, 5. Good design is unobtrusive, 6. Good design is honest". It's not until one of his lectures in 1985 where he added the last four: "Good design means durability, Good design means consistency down to the last detail, Good design means respect for the environment and Good design means as little design as possible."

By the time these principles where published in the late 80s they were referred to as "Braun design department’s Design Philosophy" and were mostly a reflection of Braun's practice than Rams'. Nevertheless it seems Rams outlived the Braun brand and became a bigger brand himself. In his retrospective exhibition at the Design Museum in 2009, many drawings and photos of braun products where printed on posters which arguably started the Rams merchandise trend.

The poster that had the ten principles also contained a paragraph by Rams didn't get as much attention. He wrote: "I have distilled the essentials of my design philosophy into ten principles. But these principles cannot be set in stone because, just as technology and culture are constantly developing, so are ideas about good design".

I was disappointed to see Pinterest and Facebook as prime examples of design. These are companies found on dark patterns, employing hostile methodology to manipulate their users.

Yes, they are popular and use carefully selected colors etc. but are horrible below the pretty paint.

I wish the page would mention the responsibility of designers to the end users of their work.

I'd recommend Mike Monteiro's "Broken by Design" to the authors of the page.

UXer with 25+ years here.

To be respectfully blunt "the responsibility of designers to the end users of their work."

No, it's our job to make the user's goals as seamless as possible - within the functional constraints of the business. If that business is to fuck users privacy over, we must operate within that overarching environment or GTFO.

To be clear, I dont understand how conscientious UXer's function at a place like Facebook, unless the plot of Servance is a reality?

> To be clear, I dont understand how conscientious UXer's function at a place like Facebook, unless the plot of Servance is a reality?

Fully ethical work that puts dinner on the table is a rare luxury, unfortunately. In the meantime, there are many rationalizations to choose from. I imagine a popular one at Facebook is something about the ends justifying the means.

You can’t put dinner on the table for 150k, you can only do that with at least 180k + 250k RSU (I have no idea what designers make at MANGA so I’m using average senior engineer comp)

Come on now!

I'm reading this as sarcasm, but while you can "put dinner on the table for even less", it's still very rare (in over 20 years of my experience in the industry) to find a company that pays more than non-profit salaries and you can't find things to feel terrible about. Heck, some of the non-profits I've worked with were some of the most terrible ones.

The only reason to work any non founding team member would aspire to go to MANGA is the compensation. Just going to be real here (no shame or anything in admitting the truth), if it was to “work on challenging problems” there are numerous ways to do that without working at MANGA.

NOBODY else pays 400k minimums for senior level roles. But 150-200k is common for non MANGA and that’s more than enough in the USA in most places. But you can’t buy both a Plaid and a big house at the same time, and fund many luxury hobbies if you go down that path.

All I’m saying is.. be real :)

Insofar as "Severance" is a metaphor for compartmentalization, yes: it's a ubiquitous reality.

Well said.

Professional ethics are a thing. "I was just following orders" not a valid excuse.

This goes for developers and data folks as well (everyone), so not like designers are being targeted.

> Professional ethics are a thing. "I was just following orders" not a valid excuse.

Let's unpack this one for a second.

First, the idea of dark patterns in UI/UX is a relatively new concept, so the ethics are naturally going to get muddy. So pretending to be handy wavy about "just be ethical" is not fair to UX/UI people.

Let's take an example dark pattern of "are you sure you want to cancel" taking 3 modals to complete. Each modal tells you something about the app that you may not have. The UX'r does an A/B test and finds that only 23% finish cancelling after implementation of the 3-step modal. So to one user this is a nuisance having jump through 3 modals, but to another user they learned on modal #2 about a product feature they didn't know was possible and decided not to cancel.

Is the UX'r morally wrong for putting in the 3-modal even though in one case it was a nuisance and in another it was beneficial?

The obvious ethical alternative is to make it an easy process to cancel and an easy process to renew and restore cancelled accounts.

> an easy process to renew and restore cancelled accounts.

What if the person was cancelling and was never given the opportunity to know there was a potential benefit? The ethical thing would also be to completely delete all of their data (including their login). They now come back and you say "sorry we can't restore all of your data, you need to start over". Didn't you just create (1) a bad UX and (2) a new dark pattern?

Again, my point isn't whether the 3-modal pattern is a dark pattern, but rather that dark patterns are incredibly difficult to define, are novel and thus require nuance when determining their ethical ramifications.

_Eventual_ total deletion is ethical, yes, but having a grace period after cancellation where data is temporarily retained is both obvious and established practice.

It's not difficult at all, when you understand the workflow and are honest with the user about the ramifications. If honesty is difficult, that is the problem.

> Didn't you just create (1) a bad UX and (2) a new dark pattern?

You still didn't address this? Care to elaborate on what the "honest" path is here?

0) Do a better job so more folks understand the benefits.

1) Have one confirmation that explains the situation clearly (≤ 3 bullets) and avoids fat fingers, such as disallow input for three seconds.

Github delete repo flow is reasonably similar.

| the idea of dark patterns in UI/UX is a relatively new concept

Relative to what? Before the web?

I left one of the first technologies companies I had ever worked for (huge name more than 20 years ago) specifically because they were using dark patterns. Not even in web UI, but in their installer for actual software (which was to be used in conjunction with the web). Specifically, swapping default and representative checkbox states. You know, the "now that we've set them up to uncheck for opting out, we'll show the next box unchecked and the user has to check it to opt-out" trick.

The term was only coined ~12 years ago...

But attempting to cheat your customer is old as time.

Yes. You've drawn a conclusion about motive without supporting evidence. You don't know why people quit trying to cancel, it could be that they became so frustrated with the cancellation process that they gave up.

Also, if you want to inform folks of features they might not know about, waiting until the are about to cancel is probably a bad design. And needing three modal boxes to inform them is absolutely a dark pattern and unethical.

> You've drawn a conclusion about motive without supporting evidence.

It's a conceptual example with an entirely plausible outcome. No need to scrutinize it any further than that.

> And needing three modal boxes to inform them is absolutely a dark pattern and unethical.

How do you conclude "absolutely a dark pattern" if even one person realizes benefit from it? My point is precisely this - there are no absolute ethics when it comes to dark patterns.

> How do you conclude "absolutely a dark pattern" if even one person realizes benefit from it?

That's the wrong way to think about Dark Patterns. It's not about whether they are ever helpful, but whether they are ever harmful. It doesn't matter what your intentions are, it matters how the end user experiences the design. 3 modal boxes to cancel is absolutely harmful when it can be done in 1 and that 1 modal can even include all the reasons you shouldn't cancel. But you get one shot to keep a customer, not three tries to keep and frustrate them.

And I have personally given up on cancellation before because it was too complicated/long/frustrating. So yea, dark patterns are real, they are not ethical, and you shouldn't defend them.

> you shouldn't defend them.

This is why we can't have a good faith discussion, because you think I'm defending dark patterns. I'm not. I'm simply asking how the designer determines what is a dark pattern that unilaterally has bad ethical ramifications.

> 3 modal boxes to cancel is absolutely harmful when it can be done in 1 and that 1 modal can even include all the reasons you shouldn't cancel. But you get one shot to keep a customer, not three tries to keep and frustrate them.

I would argue even 1 modal is too much. Why can't I just click the "CANCEL MY ACCOUNT" button? Isn't a modal that says "TYPE YOUR ORGANIZATION NAME TO CANCEL YOUR SUBSCRIPTION" in hostile to the user? How do you ensure people don't accidentally cancel and then want to reactive with all of their old data? and...and..etc.

Do you see why the interpretation of "this is a dark pattern" is almost entirely up to the user perception and has varying degrees of difficulty to the parse as the designer?

No, just use your brain and make a decision. Does that mean you'll always be correct, never mistaken? No again. That's when you drop the mea culpa.

The idea that we can never know what is truly unethical is solipsism, maybe with a little sophism thrown in. While an interesting thought experiment, it doesn't support real world needs very well.

A concrete plan: https://news.ycombinator.com/item?id=31995084

> To be clear, I dont understand how conscientious UXer's function at a place like Facebook, unless the plot of Servance is a reality?

Seems unfair to call out UX designers when everyone who contributes to Facebook's mission is as culpable.

didn't mean to only be unfair to my kind, just a UX trait to not speak on the behalf of or make assumptions about other's motivations without direct domain knowledge or research.

> To be respectfully blunt "the responsibility of designers to the end users of their work."

> No, it's our job to ...

To be respectfully more blunt, it seems naïve to assume responsibility and job overlap completely in a world with Facebooks, et Al, unless we're wilfully abandoning ethics (or have interesting political views).

Again, some context. When this was written (2012), Facebook had a relatively spare, unobtrusive design that focused on user content. In contrast to what came before (MySpace), Facebook felt modern and even human-centered.

Of course, in a remarkably short amount of time, Facebook threw away that goodwill by chasing "engagement" at any and all costs.

Personally, I would not have praised Facebook as a model for great design even in 2012, but it's more understandable in that historical context.

As oliwander says (at least in the movie), “after all, he who must not be named did great things. Terrible, to be sure, but great nonetheless.”

I would argue that Facebook and Pinterest succeeded in spite of their respective chaotic and difficult UX designs.

Similarly there is some good research about this regarding Amazon. In short, if you tried to mimic Amazon's UX as a ecommerce startup, you'd probably fail:

Paradoxically, Amazon's design may work well for Amazon itself. The company is simply so different from other ecommerce sites that what's good for Amazon is not good for normal sites.[0]

[0] https://www.nngroup.com/articles/amazon-no-e-commerce-role-m...

I’ve been helping startups design interfaces (B2B SaaS mostly) for 12+ years [1]. Here are my insights:

the big misconception is that design is important in every stage of your startup. Which is not the case.

Design becomes important after you’ve tackled the fundamentals and hit product market fit.

The companies I’ve worked with that ended up being successful, all fit in that category. Startups that would contact us before having product market fit, would see themselves innovate slower by focussing on design. It would literally slow them down. Instead of talking to customers they would ask us to change the colors of their buttons for the tenth time.

Once you have a product, that sort of works and a handful of customers love it, despite it looking like crap… that’s when design becomes a powerful fuel to your rocket.

[1] https://fairpixels.pro

But good design isn’t just the interface, good design is also about understanding the customer and discovering the product they will love. Function over form, as the mantra goes.

Maybe this site should be designed better? Meaning it is not readable on mobile. It is not any better on the desktop. Also piecharts - really? Do they know that people with even minor color blidnes do not read piecharts in same way as you think?

I 100% understand and agree with points but this site shows that achieving these points is really really hard. Maybe impossible?

Design versus UX. Totally illegible on mobile. Looks pretty though.

This “well designed” page is so hard to read on mobile

To be fair, it's from 2012, so it's missing out on a literal decade of mobile web design.

But one of the design principles is "Long Lasting"

"It avoids being fashionable and therefore never appears antiquated. Unlike fashionable design, it lasts many years – even in today’s throwaway society."

The units conversion app shown on the page is particularly pretentious and poorly aligned with user needs. Everything about it is there to say "look at me" and "see how fashionable I am in the eyes of 2008 tech fashion." None of it is there to make the app easy or fast to interact with, nor is any of it there to make the results of the units conversion easy to read or use. The fact that the page designer chose this image to explain what good design is lays bare that the emperor truly has no clothes here.

In terms of aesthetic choices, I think the design does hold up quite well.

In 2011, 6% of internet traffic was mobile, compared to 56% in 2021 (according to the first Google result I clicked)

I wouldn't say not having a mobile-responsive design back then was entirely a design failure. I recall in the early days of smartphones, the vision for mobile browsers was they were "desktop class" and would display pages exactly like how they are on desktop. Look at Steve Jobs' original Safari demo during the iPhone reveal. Everyone was blown away by the full desktop experience of The New York Times and Amazon. This website was a few years into the iPhone revolution, but it was early enough into the mobile responsiveness era that I don't think it takes away from their point. The current design could be adapted to a mobile responsive page without needing an overhaul of the existing design language.

I agree with most of what you said but the iPhone was released in 2007. This was released in 2011.

Yeah, the "mobile-first" movement didn't happen overnight—and ironically, because of Steve's influence, it took the longest for people who latched on to the ethos behind the original iPhone Safari. Just look at Gruber—Daring Fireball still isn't mobile-optimized.

4 years is not overnight.

And Gruber's CSS has more to do with his obsessive to a paralyzing degree nature, than to sticking to whatever was said in a keynote 15 years ago.

Besides, double tapping a paragraph zooms into what becomes a good reading experience on mobile. It’s that dangerous “good enough”.

He obviously should update his site and ditch Verdana, we have better fonts and screens this days. But, if anything, it’s a testament of how good the original design was that it stood the test of time and what was designed for IE5/6 on 800x600px screen is still good on an iPhone 12.

Eh, it took the mobile web until at least 2014 to get some momentum. As someone who was designing websites in this era, 2011/2012 was still the early days. "Smartphones" were still trickling out to the general population. A mobile-friendly site was considered a "nice-to-have." The screens were so much smaller back then—it was reasonable to assume people would jump to a laptop/desktop to read a longer article like this.

I don't know what's going on with Gruber—at this point it feels like he's made an almost religious choice to never add another line of CSS. LOL

I was there too, I’ve been designing for the web since 1998. I was very much selling mobile websites in 2011. Probably delivering special purpose CSS through user agent inspection, but still.

And we’re all using this orange website that was definitively not conceived as mobile first. More like mobile barely.

Something that is long lasting can be a bit dated by the time it's ten years old, given how fast tech can move.

Yes! Dieter Rams came up with those 10 principles, but I doubt he still listens to music with the wall-mounted reel-to-reel tape player he designed (though it's still beautiful).

In fact, the reel-to-reel probably felt dated within much less than 10 years.

Even then, it's over designed? Is that a thing? Kind of like an over produced song.

To be fair, it's hard on a desktop too, and with JS off the diagrams are hidden. I've clicked mostly to ensure that yet another website dedicated to design has a bad design.

I agree with some parts what is being said here, but overall I get a feeling that the content is ultimately superficial. I don't mean this as a joke or pun.

For example really good UI design is more often than not conservative and not innovative. There are reasons to break the rules, but just breaking them in order to be "innovative" often comes off as edgy and showoff-y and almost always introduces unnecessary problems. A good reason to break the rules is borne from problem _solving_ and choosing the right trade-offs.

Something I'm missing here is consistency. It is probably one of the hardest things to achieve, because in order to be consistent, a design needs to adhere to a common set of goals and restrictions. Being inconsistent and "patching over" a design or introducing new concepts is always easier. This can introduce actual problems in a design that need solving and this is where innovation can or should happen.

Sorry, but the design for reading on this site is just too bad. I hate it. Why dont you keep it simple?

Really? Is this how design works?

Classic do as I say, not as I do, made even better by this being about design and it's terribly designed.

The page is both unreadable and unnavigable on mobile, any browser.

Nothing better than being told what to do, by people who clearly are incapable of actually doing it but think they are masters of it.

This site is miles better than a wall of text. It actually made me read from top to bottom, which is something I rarely do.

At some point, we'll have to talk about the fact that designers can't explain anything about their work (and/or the vision of their work) to other non-designers professionals without being particularly condescending.

The fact that we can notice this even in the startup sector speaks even louder.

Maybe you're just hanging around annoying people?

I work at a large agency, we have to explain our design rationale every day to people that have zero design backgrounds. The better we do this, the better we're able to do our job. If we acted like assholes we wouldn't be in business very long.

>explain our design rationale every day to people that have zero design backgrounds. The better we do this, the better we're able to do our job.

Exactly, the really good and professional designers I worked with spent a lot of time to explain and work with me and show evidence for their approach.

To be condescending and dismissive is a warning sign that maybe somebody is being arbitrary and has to cover up their lack of professional development.

Also it's a two way street. When I'm implementing a UI, I try to understand the principles and ideas behind a design. Good tooling and some experience/knowledge helps, but ultimately it comes down to communication, asking questions and just having a discussion about a design.

On the other side there is also a benefit of giving technical input about a design, sometimes there are more robust solutions from a technical standpoint that gel with what the designer does.

I think both design and implementation are more efficient this way. It's also more fun.

Kinda like being ‘cool’. It’s not really formulaic, but instead subject to its environment.

Oh, yes, it is :) My favorite take on "cool" (and Albert Camus) is this existential comic: https://www.existentialcomics.com/comic/451

Have you ever tried to explain grammar to someone who does not understand grammar? It would get condescending pretty quick, especially if the said grammar ignorant insisted on asking you to put everything in one paragraph because otherwise there’s too much “white space”

A lot of this is from not actually having concrete reasons for why they designed something some way, aside from trends and fashion (which seems to be avoided admitting to at all costs)

I get that everyone on Hacker News loves to repeat that nonsense, but it really sounds like no-one here has worked with a designer before. Are you thinking of graphic designers?

> Are you thinking of graphic designers?

Honestly most ux designers I've worked with were graphic designers by training and previous experience.

“If everyone around you is the problem, maybe…”

I think to the same extent the phenomenon you describe exists, experts (perhaps even engineers) loathe accepting and abiding expertise outside their domain that impacts or determines their work.

It’s human nature to want agency. Achieving some measure of mastery can confuse one into a growing sense of entitlement to agency…and this sentiments like yours.

Have you met any developers before? As far as superiority complexes are concerned, I'm not sure designers are any worse.

So an awful page, things everywhere around the page without any logic, inconsistent highlight and colors everywhere.

I'm not epileptic, but I felt that I was so close to a seizure just trying to read a few things on this page.

Maybe it is not a real website but an ironic joke?

How Design Works (2012) - https://news.ycombinator.com/item?id=10190700 - Sept 2015 (19 comments)

Startups, This Is How Design Works - https://news.ycombinator.com/item?id=4995199 - Jan 2013 (2 comments)

Startups, this is how design works. - https://news.ycombinator.com/item?id=3793299 - April 2012 (1 comment)

How Design Works - https://news.ycombinator.com/item?id=3791427 - April 2012 (49 comments)

Show HN: Startups, this is how design works. - https://news.ycombinator.com/item?id=3789977 - April 2012 (4 comments)

Nigh unreadable on a Thinkpad with the low-end screen due to lack of contrast.

Good job designers.

> Good design is makes a product useful and understandable.

Good editing is makes all element of a lists is has the same grammatical form.

I am building a startup, and with all respect to the creators of this project, when confronted with a UI full of objects that you have no idea how the user will preceive it, these advices don't really help.. "Make a product useful" is almost like "Do the right thing".

This site is a busy mess. I would describe its design as schizophrenic.

This feels like spam. The domain name for a single article is the same as the article's title?

There's been a lot of these on HN over the last year. All very suspicious. No legitimate person is registering a domain for every single blog item they write.

In addition, it's been submitted to HN 13 times before.

Feeling like optimization of design is overkill for early stage startups. In my perspective, if product/service can express themselves clearly with simple words, they shouldn't worry about perfect design.

Good God, that page looks like someone barfed all over my screen. It's really ironic that a page that lionized Dieter Rams has such an atrocious layout and dizzying "print". It varies between highlighting and bold italics to emphasize something. The medium also switches from text, videos, and info graphics. Sometimes the text is one column but then shifts into 3 columns layout and back again. All I got from the page was "Dieter Rams". He's amazing but we already knew that if any of us only knew one thing about industrial design.

This is so condescending it’s hard to take seriously.

Well, these designers might have considered making this page responsive so the text isn’t super timy on a mobile phone.

Clicking this on mobile makes for a very ironic reading experience. The medium is the message here.

From 2012

I mean at least have good design… that was a bad use of the web medium.

Ironic that it renders terribly on iOS

Lol at the fact that this is unreadable on a phone.

Can't read on mobile, won't read.

Come on designers... do your job, instead of talking about it.

"Please don't complain about tangential annoyances—things like article or website formats, name collisions, or back-button breakage. They're too common to be interesting."

https://news.ycombinator.com/newsguidelines.html

Btw, it was someone's school project from 2012 (https://news.ycombinator.com/item?id=3790034).

Same. I guess it's an article about design by print designers. Have to print it out to consume it.

The combination of tiny text and a giant Bitmoji (or whatever they're called) face toward the end is what immediately convinced me that this page must be satire.

Reply on news.ycombinator.com