The most obvious change that happens after hiring a graphic designer is that the app/website stops looking like shit, and adopts a pleasing color palette and set of fonts. There is real value in this, and the median graphic designer definitely chooses these better than the median engineer.
But UX is a broader umbrella which encompasses interaction flows at the large end, and single function widgets at the small end. For whatever reason, the median human is very bad at predicting the overall UX of a system. It's rare that you have someone who can look at a spec for a system they've never seen before and accurately predict what will be easy to use vs. hard to use. Graphic designers are not meaningfully better at this vs. engineers either, it's just uncommon.
For that reason, UX is usually developed by copying existing solutions, or using the guess and check method to try out novel things. It's very difficult to create good UX by design because evaluating the system by imagination is much harder than with an implementation. Contrast this to backend system design where entire categories of error can be predicted and avoided through basic principles and reasoning.
Where this can go wrong is when you think that you can hire for something which is actually rare in the talent pool. If you have a graphic designer or engineer who has demonstrated an excellent gut feel for UX, then that's incredibly valuable. But you can't wait around to find such a person, or pretend that you will be able to hire someone like that.
Here's the best tool for finding usability issues: https://aistudio.google.com/live
You share the screen with Gemini, and tell it (using your voice) what you are trying to do. Gemini will look at your UI and try to figure out how to accomplish the task, then tell you (using its voice) what to click.
If Gemini can't figure it out you have usability issues. Now you know what to fix!
> If every product in your space does something the same way, there’s probably a good reason. If one company does something different, ask yourself: Is this intentional, or just a mistake?
To add to this, often you can come up with a lower friction UI idea for your specific use case (e.g. that requires less clicks) if you think hard about it, but if you stray too far from what people are used to seeing and interacting with it creates its own kind of friction, and you'll get feedback like "I found this unintuitive" or "it took me a moment to figure out how to use it". So you need to balance using familiar patterns vs new ideas.
E.g. maybe you think you can improve on the Amazon checkout experience for your own site, but by doing something different, you're tossing out the familiarity bonus you get for free. Similarly, by preferring checkboxes, radio buttons, dropdowns and text fields, over custom widgets, you get so much for free like user familiarity in reading the current state, and knowing how to change the state.
"Unintuitive" can often mean "I'm not used to this pattern" even if it might be a good pattern once people get used to it.
That was a lot of words to say "Jakob's Law"[0]
"Users spend most of their time on other sites. This means that users prefer your site to work the same way as all the other sites they already know."
"1. Users will transfer expectations they have built around one familiar product to another that appears similar."
"2. By leveraging existing mental models, we can create superior user experiences in which the users can focus on their tasks rather than on learning new models."
"3. When making changes, minimize discord by empowering users to continue using a familiar version for a limited time."
I recommend focusing on general design principles and mindset.
- Read "The Design of Everyday Things" by Donald Norman - once you understand what makes a good (or bad) door handle, you'll start seeing design patterns everywhere.
- Read "The Art of Game Design" by Jesse Schell. It discusses how to create engaging experiences, and games are particularly unforgiving. While people might tolerate an annoying tax app because they have to use it, they'll immediately abandon a game that's even slightly too frustrating, confusing, or boring.
Surviving without a designer? More life thriving without one.
Honestly I think many startups at an early stage don’t need designers at all and they just want to look trendy (which admittedly is critical for consumer-focused companies that live and die on becoming trendy, but that’s not what I do).
The best designer for an early stage startup these days is someone with not awful visual/UX taste (no need for a specialist or a visual prodigy; most teams I’ve met have at least somebody who isn’t hopeless in this respect); but with a strong understanding of their target users and what they’re familiar with, so you can just copy prior art as is relevant. Much better than taking a designer who knows nothing about the space and needing to transfer your learned experience into their brain somehow.
Most startups aren’t doing anything truly novel from a UI standpoint so why reinvent the wheel? Designers become more useful when you stop having strong communication and trust in your team since it just got too big for that.
Doing something because all the big companies do it also leads to cargo cult mentality. You should know exactly why you are building every little part of your system. “Oh Google used a really annoying captcha on that page, I better do that as Google knows best”.
Have some confidence and don’t assume that other bigger companies are smarter than you are, think about what you can improve. Most of what Google have to offer, they bought from smaller companies that had the confidence to do just this.
From experience I will say that you can hire a UX designer even if bootstrapped and low on cash, and that it's a very valuable investment.
Just don't hire them full time as the article seems to suggest is the only choice.
Getting a small firm to go through a design sprint with you with, e.g. designing 3 concepts, letting you run a couple of UX workshops with your potential users, then picking one of the options to flesh out into a clickable prototype, then workshop again, then final prototype, can come out within a $5k-$10k budget.
That's 100% worth cutting $5k from your front-end dev budget, and will definitely translate into way more than $5k in user retention gains within the first year.
This is what we did before coding the MVP, and we're doing it again now (at Seed stage) before shipping our biggest upgrade to the product.
> Use AI to spot blind spots
> Tools like ChatGPT can highlight UX issues you might miss. It’s a quick sanity check—not perfect, but better than guessing.
> Tools like ChatGPT can highlight UX issues you might miss. It’s not perfect, but it’s better than guessing. Some prompts to try:
Was this an intentional joke?
Can't remember last time I worked with a dedicated designer, someone who actually knew anything worth knowing about UX.
Devops seem to be going down the same path, it's like they expect coders to do it while the code is compiling.
Next up seems to be coders.
And I get it, hiring professionals is very inconvenient.
We need a part 2 for this, the stage after you've picked up a bag of VC money:
1) Add an AI chatbot to every screen, add "AI-powered" to the homepage headline
2) On the pricing page, title the top tier as "Call our Sales Team"
3) Make buttons with undesirable actions hard to see and click ("Cancel Plan")
Here’s some typical reasons why a startup can fail:
1) it failed to communicate and market it’s product
2) it’s product didn’t fit the user’s needs
3) it’s technology strategy made development too expensive
4) it’s product technical quality was too low
5) it’s product did not look appealing to potential new users
Developers are responsible for 3 and 4, sales and marketing for 1 and finally designers for 2 and 5.
With competent developers you can start a startup and make sure 3 and 4 never come to pass, but lack of good product designer will eventually kill it.
Here I use the broader sense of user-centered designer, which includes:
- research
- testing
- prototyping
- validation
- UI/UX design
- visual design
- …
The first four being the most important for a product market fit.
This is especially important for B2B products, because there understanding the needs of the business and their processes is key to making sure the product fits the user’s day-to-day work but the businesses’ future needs as well.
It may not be common, but you can and should use extended UCD research methods on the customer business processes itself instead of relying on PMs and sales just asking customer’s what they want. (This is often called Business Design or Service Design around here.)
Highly recommend Ryan Singer's (original Basecamp designer) work. His blog [1] is filled with great stuff. In particular this [2] post has given me over a decade of value re: designing UI and UX flows.
[2] https://www.ryansinger.co/thinking-of-interfaces-as-sets-of-...
I have a different approach. I look for a theme on Theme Forrest which has most of the layouts and components I think I'll need and I lean on those VERY heavily. And for logos I use icons from Font Awesome or Bootstrap.
Most of the time the project doesn't take off and when it does I can hire a designer.
Some examples of both a theme based app using an icon as the logo :).
If you're not hiring a designer, someone else ends up doing their job instead. Someone, whose pay is probably 5-10x higher.
You don't even need them on the team permanently, commission a design or a review of the existing one.
Perhaps I'm biased. Graphic designers are dirt cheap. From our perspective, UX crowd is full of underqualified people looking for easy tech money. I can see how it can make hiring far more complicated.
> Common patterns—like password strength indicators—usually exist for good reasons.
NIST does not recommend password strength indicators. Make sure the form is compatible with password managers.
I like the article because it's very similar to the workflow that I developed while working on my first project after quitting my last job. One thing that I would add to "Be Explicit About Your Goal" is that it should be about both your goal and the user's goal. Otherwise it's very easy to hit blind spots that both you, another person, or AI will miss because of bias.
A small related rant is that many large companies with a dedicated team that works on design system don't seem to actually care much about UX. It feels demoralizing because sometimes it feels like many people are just using UX to push some other agenda.
Design must flow from customer demand/desires.
And 90% of design is just "correctly assigning priority" to elements and actions.
If you know what is important (and what is less important) you use...
- white space (more whitespacce = more important)
- dimension (larger = more important)
- contrast (higher = more distinct)
- color (brighter = more important)
... to practically implement the decided priority.
How to validate you have implemented priority correctly?
Just ask a few people what do they see first, second, third, etc in a page.
If you designed it right - their eyes will see things exactly in the order you expected them to.
In short - "design is guiding user's senses in the most prioritized manner to the user in achieving their goals"
In our startup - we call this the "PNDCC" system (priority, negative space, dimension, contrast, color).
There are a few more tricks to make it even more powerful - but as I said - just getting these right puts you in the top 10%
Like... sorry to be That Guy, but maybe just employ / have a team with the right people from the start?
If a designer had a tech startup with no engineers, we'd all be (rightly) sniffy about it.
Imagine this post the other way round: "Hey, designers, here's everything you need to know about engineering and devops and coding in ONE PAGE!". Think about the HN response to this - it'd go down like a ton of hot shit. And rightly so.
I used to work at an educational tech company and they would - without even breaking into a sweat - take one of their engineering team and put them in charge of marketing. Same rule applies - imagine it the other way round, a marketing person with no knowledge of engineering setting the strategy for engineering. (And yes, I know it happens, but when it does we all moan on about it, no...?!)
Long and short: a startup / product without a UX person or designer in it right from the start is likely to be a clusterfuck.
My startup is a non-standard domain targeting the iPad.
Our need for a designer with functional knowledge and creativity is very high.
I’ve muddled through with iterative designs of my own, but this challenge has delayed our beta.
Designers don’t work for equity where nearly every other kind of talent will.
This is a real blocker for startups.
Tailwind + daisyui can get you pretty far. My thinking is, if your start up takes off a real designer can remove all of the daisyui stuff and re-design with only tailwind.
Where do startups typically get their branding done? I'm assuming the VCs usually refer their cohort to the same group of branding agencies? Who are the quick and dirty ones? Do they ever hire direct freelancers? Possibly to save money?
I’m hardly a UI, well, anything.
But I want to point out the application that is used at the Quest Diagnostics lab. The one they use to enter your demographic, lab info, etc. The daily driver for their legion of phlebotomists, at least here in So Cal.
One word. It’s fast. Oh man is it fast. Bunches of fields, bunches of dialog boxes, looks like a Windows back office application. Prints all of the vial tracking labels. These people are trained and know what is coming when.
But, it seems to be in a web browser tab. Perhaps they’re running a RDP tab to a nearby server running some Windows based desktop app, but, again, this app flies. I’ve not seen modern app this fast.
I’d love to know how they are pulling it off.
It takes a designing engineer to startup a touch minefield that User X practically cannot survive without triggering a popup.
As a quick alternative, why not use a freelancer?
Practical UI from Tailwind is a very good guide for this.
Were em-dashes this popular before LLMs?
Or, just hire a good UX designer. Seriously, nobody would think about "practical coding for startups surviving without a programmer."
() Well, with AI coding... who knows...
eh. For who want's to get deeper/broader view, there's a "bible" on UX.. called "Usability engineering" by Jakob Nielsen ~1993 [0].
IMO The first 2 (two) chapters are, like, mandatory - the what and the why. The rest, is mostly details - how. (but it has even how to organise random-street-user-tests and what to look for in those, etc)
Although, looking at recent 10 years of (software-or-car) UXs, i feel noone reads these anymore :/ Sadly.
[0] http://www.amazon.com/Usability-Engineering-Interactive-Tech...
This is good practical advice
1qq
Great try this and see how far it goes ! None of this matters if u don’t find pmf and u don’t need a designer for this. Totally disagree with this. Article started great but then niched out too small with login flows. No startup is reinventing this.
AI is not designing good? How? The tech bros hate artistic minded people.
They hate design because it is the engine of growth and change. Tech bros like sameness and libraries. They love "design patterns".
But change is inevitable and design is not in the colors, fonts, whitespace.
Design is human centered practice for solving not only functional tasks but emotional. And for this there is no Tailwind, React, AI solution, or design patterns template.
PS: I am a designer and frontend engineer with 20+ years practice with my own company and plethora of delivered solutions for my clients. The divide between engineers and designers is the most persistent thing that I have seen in my life.
It is beyond my understanding.
To me, peak usability was 25 years ago, when most applications had a toolbar and a menu that followed a standard pattern. If you're a frequent, non-power-user, you use the toolbar (e.g. "insert row" button). If you're an infrequent non-power-user, you go through the menu (Insert > Row Above). If you're a power user, you remember the shortcuts indicated through underlined letters in menu labels (e.g. Alt, I, A).
If you want to change settings, you open the settings dialog (Tools > Settings), and it as tabs like "General", "Fonts and colors" etc.
Most people were a lot less computer literate back then, but they were able to use most applications with little help. If they really needed help, the help system was built into the application.
The goal back then was to have the user get the work done as efficiently as possible, in effect, minimizing the time the user pends on the application. Modern UX doctrine aims for the opposite goal -- to keep people "engaged" as much as possible. This might be okay for consumer apps, but maddeningly, the same doctrine gets applied to enterprise applications as well. I've literally heard non-techie employees of a Fortune 100 company ask for their legacy green screen terminals back because the new, flashy SPA was slowing them down.