This is not a generally true statement. Big (aka successful) companies with a poor UX are likely focused on selling or some core feature set that is hard to create, and if they're big then it means they are or were making that tradeoff correctly.

In all likelihood, most startups would benefit from focusing less on UX concerns and more on selling, selling, selling.

You can have:

- products with horrible UIs and horrible graphic design, but awesome user experience (UXs) -- because they solve a problem perfectly while enabling a good workflow, or are extensible and adaptable to new unanticipated workflows

- products with horrible graphic design but great UIs -- if the user flow is intuitive and productive, the interface responsive and discoverable, it really doesn't matter how shit looks

- products with great UIs and awesome graphic design but horrible UX -- the UI may be both intuitive and awesomly designed and responsive... but if it enables wrong or sub-optimal workflows more than the right ones, nudging users towards bad mindsets/perspectives/workflows, it lowers everyone's productivity and sooner or later people will realize that they are dragged down or disabled by that product with the "great" UI

So yeah... if what I wrote above ain't obvious to you, then don't focus on UX, because you'll actually do it wrong anyway. Focus on the problem you solve, on the product you develop to solve it, and on selling the solution. If the UX is not extremely bad or you're not in a "fashion/fad" driven niche, it will probable be ok despite suboptimal UX if you do the rest right.

Also, can you cite some examples of the list of products?

I can think of

- products with great UIs and awesome graphic design but horrible UX: Shopping cart checkouts that if you click something, you have to write the CC/Name-details again. Multi-city flight search on any major travel site is a pain in the ass.

Yeah, getting to a good definition of UX vs. UI that everyone could agree on would be great ;)

I'd start with UX defined as "what the users actually ends up doing while using your UI, the actually followed user journeys, the workflows they lead to, all this from the user's subjective perspective (hence the 'experience' part)". Also relating all this to goals and talking about "conversion" etc. would be about UX. And the general feeling of "this really helped me get the job done" that the user gets after using something. Heck, even customer-service can be part of UX if your product does something so complex that having users regularly and often interact with customer service is part of the "regular usage pattern" (think stuffs where CS is not just where you complain, but where you may have talks with reps about "special shipping and assembly arrangements" etc.)

And UI would be "what and how can the user possibly do with your interface, what are the actual interface elements that actually dictate these "how" and "what" of the user's actions, things like responsiveness, intuitiveness (matching user's previous expectations), information density, discoverability". So you can talk about "user satisfaction/enjoyment" here, about "intuitiveness" etc. (but these don't necessarily mean 'higher conversion' or making the user more productive!) Also accessibility comes here. And think things like "is this button big enough to be tapable" or "it that option in hover-only menu that is impossible to access on mobile devices" or "is that other thing a keyboard only shortcut that tablet users will likely miss" etc. These are obviously independent form how things look: the design can be horrible, but everything can be easy to find/tap/click etc.

Graphic design is obviously "how things look" and this also affect the "how easy is to find things", to it does affect both UX and UI... but, there is a big "but" here: designs that receive awwards and are perceived as beautiful tend to have functionality harder to find, and more confusing user journeys. Think about the "ghost buttons" that people used and still use... but have been proven to decrease conversion quite a lot! ...but now the trend might have reverse because users are finally getting used to them. Like, do you really want to be playing this fashion game that can hurt conversions when you don't expect it in a resource constrained startup?! If you're Apple, you can "reeducate users' aesthetic tastes" and stuff which is cool, but you're not Apple. So better under-design a bit just to be safe :)

Thank you for articulating this point so well. I think websites like HN, 4Chan to an extent Reddit are best examples for this category

I LOVE HN's design.

That's one of my pet peeves. This literally is goes in the opposite direction most of the time - the prettier the graphs presented, the more useless they are. They show less, take more space, omit units, scales, or well... even useful data. Not to mention connecting things that should not be connected, or drawing areas where they shouldn't be, etc.

It's almost like if some people wanted to screw over their customers on purpose. But, I guess, it's more of people who don't know what they're doing selling things to people who don't know how to use them.

The intersection of those too groups is too small to make any difference, thus tools will be heavily optimized for one of those.

http://www.darkhorseanalytics.com/blog/data-looks-better-nak...

Wikipedia, Reddit, even HN are great examples of successful projects with good enough UI. Even Gmail or Facebook had terrible UI when launched.

If you are solving a unique problem, UI is not so important.

HN/Wiki/Reddit don't: break the back button, take forever to load 1000 JS/CSS files, have broken "infini-scrolling", or even break when JS is disabled (well, reddit kinda does now).

Compare that to the majority of over-engineered SPA-style webpages that employed many, many front-end engineers and UX designers, and Wikipedia, Reddit, HN as well as old-school Gmail and FB start looking amazing.

Man oh man do I miss old Gmail... that was a great interface. I exclusively use FB on mobile, no-js mode.

Web brutalism should be the new hotness.

I'm quite tired of not being able to use sites on slow connections.

I guess Gmail brought it to the masses, along with a decent amount of free storage.

In case of Reddit and HN, I started to use them many years after I learned about them – mostly because of the weird UX. I still prefer hckrnews.com for HN.

You can't launch Craiglist in 2017 though. I know that there are numerous exceptions from the 00s and even from the App Store 5-6 years ago. But it's not the same anymore because blue oceans are rare nowadays.

That's what I've used in the UK and Australia, an ebay own it too.

However, for all these projects, the UI may be relatively primitive, but the UX is outstanding because it's stripped to exactly the things the user needs to accomplish what they're trying to do.

LastPass killed it because they had Enterprise grade features people wanted. I totally agree with you. It's about building a product for your customers

All the comments disagreeing with this seem to be talking about "big" companies like Reddit. Think more IBM. OP's comment is very accurate.

And the 'core feature set' can be intangible features of the company, such as brand.

Why is that fact relevant? Care to elaborate?

Isn't this what "their clients often have no other choice" means?

I just left a job where the founder (who didn't have any software background) deeply ingrained himself with this philosophy and wound making life a living hell for engineering.

While building a full-stack prototype, we slapped bootstrap 4 on the front end because who has time to roll CSS when the project is months behind schedule?

The founder with no sofware experience decided that bootstrap was terrible and that he'd roll an entirely custom app style. So he went into the codebase on a weekend and ripped out all the views/controllers/assets and replaced them with his own, which massively broke the test suite on Monday morning (which of course he didn't know how to operate or fix.) After a couple of other "executive audibles" like this, I took the next good offer and ran.

It's weird to see 'visual design' as a mainstay of an article titled 'What you should know as a founder of a software company', as if managing a team and building solid engineering fundamentals and culture are either assumed or unimportant.

That could have saved you from weeks of hell and would also increase your influence by a great amount.

I thought this too until recently -- got fired after trying to protect my employer from a toxic exec.

There are some that do not accept anything but yesmans or silent codemonkeys. It sucks but it pays. If you are not willing to lose your position, just let it go.

Am I naive to think that in a business setting, problems should be constructively confronted proportionally to the extent of their damage potential directly with their author? Wouldn't the business be able to grow more efficiently and be able to keep a peaceful environment free of backstabbers if people could accept that everyone makes mistakes, that they don't really matter as long as they are fixed quickly, and that they are all learning opportunities that can become valuable assets? I feel like am I throwing buzzwords around and that to some extent I have always heard this in a form or another but for some reason, never seen it put into practice.

A) The company is bootstrapped, profitable, gets positive cash flow from paying customers and is investing profits into adding new features that will cover larger markets. Its competing with other bootstrapped companies on how fast it can add features.

B) The company is VC-backed, runs at a loss, gets it cash from rounds of investments. A competitive factor for getting more capital is showing rapid growth and creating an impression that the company is sexy, has a great future, etc. A competitive factor in getting rapid growth is offering services subsidized by the venture capital and getting high-profile salespeople through the CEO's network.

In the first scenario it's all about the product. In the second it's all about the CEO's ability to sell the right impression to the venture capitalists.

Things get more complicated once you bring in psychology. The people that are good at confidently overselling a product and giving a perfect impression are usually people-oriented. They can give a good impression, and they draw their motivation from impressing others and glorifying their social status. If a person you're talking to starts dropping hints about how much of a big guy he/she is, that's the personality type I'm talking about and you don't want to confront that person because you'll be seen as a personal threat regardless the long-term effects. This doesn't benefit the long-term product quality and efficiency, but will benefit their self-confidence (hey, I'm a big boss, nobody dares object me) and will help them attract more venture capital and get more high-profile sales people aid their course.

So yes, in many real-world scenarios everyone is talking about results, learning opportunity, positive environment, growth, etc. because this is a part of the impression you want to make, but in reality everyone is backstabbing each other because if you don't do this, someone will backstab you.

What is also another good way to say that, shutting up or not, they deserve to be out of an engineer, so go immediately looking for another boss.

The goal is to develop a sense of good design, so you can pick a good template or hire good designers later and argue about the results of their work.

I'm still working on a way to stop commits that drop the coverage below a threshold so even new code can't be added if it isn't tested.

#1: Solve an important enough pain point that companies or consumers will pay you for it.

Sorry it's not more catchy.

The list provides a very good insight into specifics of making a product/company work. These aren't sufficient conditions, but they are often necessary. Product packaging matters. Customer relations matters. Resource management matters. As does the non-technical side of things. All of these are integral to making "companies or consumers willing to pay you for it."

I can think of quite a few products that look nice and the workflows are quite easy, but at the end of the day they're just not solving a problem I'd be willing to pay money for (and, often, they disappear later).

It's very hard to admit that you were wrong about the fundamentals of your product, and very easy to think that re-designing the homepage will fix everything.

Most start-ups aren't solving any huge problems, and sometimes market inefficiencies are there for a reason (see all the start-ups involving the field of law, for example).

Even for the ones that are solving a big problem (that people are willing to pay for), it's very uncommon to have no competitors.

So typically the way to make a successful company is to be better than your competitors, and the way you do that is through customer support, UX, etc.

So to your credit, I think for most start-up founders this is the way to move forward. But it's definitely a bit of a sunk cost fallacy - it's really hard to admit your idea will never be successful and to start over from scratch.

It's easier, and has more obvious ways to measure success, to adjust relatively minor things.

Look at twitter. Fundamentally people do not want to pay for it. They still haven't figured it out.

"Have I solved a problem serious enough for people to bother to become my customers?"

Thanks for the added clarity and article.

I never mentioned it in Pinboard marketing, but I moved heaven and earth to keep the site fast for everyone, and it was speed that gave the site its first toehold.

That said, I think the leap from 100 to 10,000 users is very hard, and I don't know of any advice about how to cross that gap.

Also, when we started to model pedestrian traffic, we also began to start on the long road to releasing software so that we could sell the software and the companies could model their own traffic. (Many companies consider such data to be proprietary and it can be difficult working for them because of their protective nature.)

We were able to get consultation through Jacob Nielson. He was still fairly new to the game (early 2000s) and things like eye tracking were just coming into the fray. I wasn't able to attend any of his meetings and didn't have much to do with the specifics.

But, my employees were happy with him and our customers were happy with the results. He's mentioned in the article by NN Group, which is Nielsen Norman Group.

The employees were happy and, as mentioned, so weren't our clients. They were also able to take what they learned and apply it to our internal applications and processes. (UX is much more than UI.) This, in turn, helped us more than what we'd expected. I distinctly recall QA being happy with a number of changes.

But, it points to the value of good UX and the benefit of getting a professional to consult or, perhaps, directly on staff.

Because of my lack of direct involvement, I'm not going to specifically recommend them - I don't feel as though I'm qualified to do so. I will say that the results were very good.

We'd chosen them because one of our employees had been following their site and was very impressed with the work they were doing with regards to web sites. It was that same employee who had recommended we get a UX consultant in, in the first place. Given the results, I am happy to have listened.

:p

I run a rake task to scrape all of the landing page content from the Rails app, then copy in the static docs site, and upload everything to S3. So it's like I'm using Rails as a static site generator.

Basically my root domain is all static files, served from S3 / Cloudflare. And the Rails server is running on an "app" subdomain. I think it's been working pretty well. In the future I might make the whole app static, and just use Rails as an API.

They do not solve a customer problem.

Solve their problem, make their life better and you will do well.

Then a few tech startups doing analytics I had some experience with, probably overpromised (?? perhaps it was my fault) what their engine could do. Requests are often in backlog and can take many months to roll out, which by then contract renewal is up.

Markets form around problems, not solutions.

If instead you're doing Yet Another Pinterest, then good luck, you're playing a very different game.

A tough one. Asking myself what I absolutely need, I come up with air+water+meat+housing. And I'm a coder! =) but even for softer interpretations of "absolutely", haven't ran into any such non-niches (we're talking "people", not obscure-exotic-niche-use-case-prospects) in a long time lately.

This is a hard truth.

But we are doing very well, we are now over 40 employees and have several of biggest names in the nordic countries as clients and our software is being interacted with hundreds of thousands of individuals every day.

Good UX is important, but it doesn't have to be "perfect". The most important thing by far for a startup is getting out of your office and talk with potential clients.

If you are solving real problems and you are constantly learning and improving based on your understand why your customers hire/fire you then nothing else will matter.

If you are really customer-value-centric then your UX/Visual design, PR, marketing, SEO won't matter since all of your customers will be telling more & more people why you are awesome and they'll keep paying you for more.

While reading the article, I had no idea you weren't a native speaker until I reached that line. So: congrats.

If you're B2B the strategy may be somewhat different.

> One of the best ways to do it is to watch people interacting with your prototypes without giving them any hints.

This is so true. My product has a lot of very complex features, but after watching some early users, I realized that people were struggling to figure out some of the most basic things - even just adding and removing fields. I had made the 'delete' icon only show when you hovered over the field in the sidebar, but I made it more obvious by always showing it when the field was selected. I also added another 'Delete' link to the field option in the right sidebar. And then I also added a welcome modal that explain how to add and delete fields with animated GIFs. (Including the "delete" and "backspace" keyboard shortcuts.) Now that there's 3 separate ways to delete fields, people figure it out almost immediately.

Even adding fields was hard to get right. My initial version just had click-and-drag, so you had to first click, then drag to adjust the field width. I saw that new customers were just clicking on the page, and nothing would happen. So I added support for single clicks as well, and now it adds a field at that position with a default width.

It's scary that I might not have realized these UX problems if I wasn't paying attention.

> Nowadays it’s easy to get amazing stock templates for less than $50 instead of spending thousands on a custom design.

I strongly recommend https://pixelarity.com. I've used 3 of their templates now, and I think they're very good. Some of them even have jekyll versions that you can download. http://unsplash.com is incredible for stock photos, and I have a subscription to https://thenounproject.com for vector icons. There are a lot of free icons out there, but the Noun Project has a ton of variety, and lots of things you can't find anywhere else.

> When doing the development yourself, it is easy to slip into a false productivity when you increase the amount of code without achieving your business goals. The most infamous examples are building tools instead of a product or overthinking a complicated architecture "for the future."

This is really tough, but I'm actively trying to avoid this. Not just the fun tools and side-projects you want to build, but even some feature requests from customers. Certain might end up taking weeks or months, and they'll over-complicate your product, or they just won't be a good investment of your time.

> New and trendy technology usually means bugs, breaking changes, immature tooling and lack of documentation. "Boring" mature tech will allow you to achieve the same goals much faster and make it easy to find developers for hire.

I used Rails and React, since I'm very familiar and productive with those. I was tempted to try Elixir, but I think it would have taken me so much longer to build an MVP. I have a friend who started building their startup with Elixir, but switched back to Rails for the better productivity.

> Depending on the country, your experience with these business aspects can range from mildly unpleasant to the absolute worst.

I've almost finished setting up my company with Stripe Atlas, and I can say that it has been the absolute best experience. I can't recommend it strongly enough. I'm going through the post-incorporation stuff now with UpCounsel, and everything is just so easy.