Opinion: The State of Keyboard Design

The initial goal with the blog was to model what Hodinkee has done with watches. Hodinkee doesn’t write hit pieces ever. Instead, their philosophy for pushing the watch industry forward is by celebrating the best parts of it.

It’s easier for Hodinkee to focus on celebrating greatness because the watch world is rich with great product, but I don’t have that same luxury with keyboards. That’s cuz keyboard design kinda sucks right now.

The 2-part case with seam

In the last few months, I’ve seen the same design tropes repeated to the point of being stale. The 2-part case with seam is ever-present. Deviations from that case style are limited to throwbacks to vintage keyboards, repackaged OTD features (the Koala curve has been particularly hot as of late) and/or variations of a side profile recess.

Quantrik’s QXP — one of the many recent releases that sports the Koala side curve

It’s not that the community only wants simpler designs, either. We’ve actually been supportive of wackier designs. Gray Studio’s SPACE65 keyboard comes to mind as one that received positive feedback in its Interest Check phase and is now doing decently in group buy. Also, an ‘out there’ design from TGR, Quantrik, LZ or (insert literally any other popular designer) will sell well regardless of how it looks. Designers just need to make the leap and actually expand the frontiers of keyboard design.

Gray Studio’s SPACE65 and its creative LED usage

In addition to case aesthetics, improvements to performance design have also stagnated. Top-mounted, 1.5mm, 8-holed brass plates (a 2016 innovation) are great, sure, but do we know for sure that they’re the best if we have not been experimenting? And if we can’t even improve on the major stuff like plate mounting, there is little hope that we perfect smaller details like the number of mounting points, the position of mounting points, the PCB-to-base distance, case thickness or weight implementations. Heck, the Tokyo60 couldn’t even get keycap-to-bezel spacing right and that should be a given.

My grievances with keyboard design go beyond the case. PCB programming is a defining feature of custom keyboards, yet it is mostly a shitshow. QMK is the best thing we’ve got for programming PCBs, and I much prefer it over the competition, but it still lacks the polish and ease-of-use of a truly great product.

Screen Shot 2019-01-10 at 12.09.11 PM.png
The QMK ‘Complete Newbs Guide’ that’s detailed and helpful, but could be simplified with a double-click-to-install application that’s actually easier for newbies

I can’t help but compare the state of keyboards to the other hobbies I’m interested in. In high-end clothing, fountain pens and watches, designers take every detail into account. For example, the attention to detail in making clothes is staggering. Construction wise, the drape, fit, fabric, stitching, etc are all rationalized and designed. Performance wise, there has always been a heavy emphasis on flattering the body and maximizing mobility. This obsession with design is a far cry from the mere 20 hours on Autodesk Fusion 360 that some designers put into a project. (EDIT: ‘some designers’ and not ‘designers’)

This slowing down of improvement has had and will have serious consequences on the health of the hobby. Because of the lack of both depth and variety in design, some of the most passionate of collectors have moved on to other hobbies. Furthermore, we are just not attracting new members like we used to. I (very irrationally) love this hobby we have here, and the last thing I want is to see it shrink.

So, what can we do?

Expect more prototyping. It’s no surprise that the most successful keyboard designers have been the ones who prototyped the most. Notably, yuktsi of TGR fame went through 9 prototypes for his TGR Alice keyboard before settling on a final design. The extended prototyping phase helped yuktsi nail down the measurements, surface finish and typing feel of the Alice, so much so that many people consider it THE keyboard of 2018.

yuktsi’s TGR Alice

Yes, repeated prototyping is prohibitively expensive for many independent designers. However, I (and many others) would gladly pay more for a keyboard if I know that a keyboard has gone through multiple iterations before going to market. Not to mention the fact that many designers make serious bank from group buy and B-stocks.

Expect innovation. To push the hobby forward, we not only need creativity from our designers, but also selectivity from us as consumers to support projects that actually matter. With every new interest check, ask yourself: ‘What is this keyboard bringing to the table that doesn’t exist already?’ It could be a new feature, different implementations of an existing feature, or similar features at a much lower price.

Seek inspiration from outside the hobby. For too long (10 years!) have we stuck to the design language established by OTD keyboards. Instead of looking within the hobby for inspiration, keyboard designers should look outward. There is so much out there to draw from in product design, architecture and visual arts that we shouldn’t limit ourselves to the two-part case with seam.

With its glass-and-concrete construction and exposed pillars, UCSD’s Brutalist Geisel Library could be a source of inspiration

From architecture, we can call upon International Style, Brutalism, high-tech architecture, or even the avant-garde for new ideas on case shapes and construction. From product design, we can look to the Bauhaus and Braun for creative color usage and positioning. By tapping into existing fields of design, we not only inject fresh ideas into the hobby but also draw new people into the fold.

Keyboard design and innovation has slowed down recently, but I believe we can and should do better. 


Published by Brian

Reviewer Extraordinaire

14 thoughts on “Opinion: The State of Keyboard Design

  1. Hi Brian, QMK has progressed past solely using a text based solution to create keymaps. One of QMK’s newest tools is the QMK Configurator located at https://config.qmk.fm. While you’ll still need to do the text based approach to fully harness QMK’s most powerful features, the QMK Configurator should handle most cases. In addition, Olivia/Wilba have released a GUI based on QMK for boards such as the Zeal60, Rama M60-A, etc etc that make the keymap changing process even simpler. You can find it on https://rama.works/configurator

    1. Hey MechMerlin,

      I just watched your video on the VIA Configurator and it looks like everything I wanted in PCB programmability. If QMK as a whole were to get to that point, my argument about QMK not being polished would be moot.

  2. Can’t lie, I think this is entirely out of touch with what is actually happening.

    A few years ago we had a few custom 60% and TKL boards, but really nothing too outlandish. The JD40 was around, but in very small numbers with very rudimentary build quality.

    The first huge mountain we have overocme as a community is the adoption of new layouts. This is extremely difficult, and the number of new layouts we have accepted as “standard” in such a short period is nothing short of mind boggling. I’m considering layouts like 65%, 75%, 96 key, and even ortho to be essentially fully adopted at this point, with the proof being that muliple makers are creating these layouts, and custom keycap sends tend to fully cover these sets now. Hell GMK’s very first stock set covered all of these layouts, including ortho! There is a ton of thought that has to go into a layout to be successfully adopted, but people have come up with some incredibly unique layouts.

    Case design actually has a vast ammount of variation now, even compared to 2 years ago. The days of
    just having the sandwitch customs from GON/Winkeyless and full customs from OTD/KMAC/DUCK/LZ are over. These days there is some new design coming out on what seems like a weekly basis. We have bat shit crazy things like LZ Iron, EV.A, TMO50. We have vintage inspired boards like the CA66, ALF G60, the 60% Dolch board, and the the numerous M0110 boards.

    There is also the hugely overlooked Deskthority crowd that is just always making something bizzare and amazing. Brand new Model F’s, converters for just about anything (much easier said than done usually), incredible single builds, etc. The innovation I find on DT blows my mind on a weekly basis.

    Frankly, there number of designs based around those of OTD, and how fast they sell, proves there is still a huge market for such classic design work. Sure, we have plenty of takes on the classics these days (like 60% and TKL) but there’s nothign wrong with that. Design variations within those layouts are vast. The Time 80% was just hilariously crazy, and there are more build styles of 60% cases that I ever thought would be possible. I’m not sure if you just want something that looks totally alien (if so, just go look for any gaming keyboard) or don’t see anything that isn’t on the first page of GH, but the speed at which this community has grown, innovated, and continues to push the boundaries is incredible.

    1. Hey livingspeedbump,

      I disagree with your assessment of my article being out of touch, because many of the examples you put forth don’t dispute the arguments put forth in my article.

      1. I agree that there has been a lot of innovation in the layouts we have now. However, with many of those keyboards in new layouts, they’re not doing enough to improve the performance or case design aspects of it.

      2. Sure, there are new designs coming out often and the field isn’t dominated by Korean makers, but many of the boards fall within one of three categories: 2-part case with seam and variation of a side recess, vintage keyboard inspired. My argument is not that there is no innovation, but that what innovation we do have is limited to very definable themes. When we do run out of layouts, side profiles, vintage boards to copy from, etc, what next?

      3. The stuff happening at Deskthority is great.

      4. I agree there’s nothing wrong with putting out iterations of classic designs. Unfortunately, what I do see coming out from the more established makers have mostly been those classic designs. They can exist, but they can’t be most of what exists.

      5. Keyboards that really push the boundary of case design like the LZ Irons of the world are only a small percentage (and an increasingly small percentage) of the keyboards that do come out to market. I’m not necessary looking for alien shit, but just stuff that doesn’t look like anything that exists currently.

      Thanks for such an engaged reply!

  3. I have to agree with Merlin and LSB, this opinion seems out of touch. My latest build has an aluminum unibody case, solderless hotswap sockets, per-switch rgb, and usb-c. I never had to look at a line of text, because it was programmed live (no flashing or compiling) with the VIA Configurator. We are leaps ahead of where we were two years ago.

  4. I will start this off by saying that I respect your opinion entirely. It’s yours and yours alone.

    I’ve derived the main points from what I’ve read in your piece and I would like to state those right now. Please feel free to correct any misinterpretations of your opinion.

    – This blog is trying to highlight innovation for keyboards, but I am having a hard time because of the lack of interesting ideas in keyboards. Recently, board designs are becoming homogenous and therefore boring.

    I disagree with this sentiment. The trends of 2018, such as a shift towards hot-swap, polycarbonate as a material, unique anodization, gaskets, new takes on sandwich cases, and even the explosion in popularity of thick plates have incorporated themselves into the foundation of the hobby where they will branch out in the next coming year to become their own thing. Sure these things may have existed before but the spread and refinement of these have taken off this year and people are doing new and creative things with them. Innovation doesn’t have to be something new and flashy. It can be a subtle refinement that improves the board slightly towards a better experience for everyone involved in the hobby.

    – Seemingly the community agrees with me as well because of more outlandish designs. Popular designers move units regardless of quirk so they should make the board more quirky.

    In addition to being hobbyists that design many designers also design for the money, as a business. Therefore in a subjective and opinionated area such as this, a safe answer is often reliable and will move units for less work a majority of the time. If they’re going to move anyways, why invest the work?

    – Boards have also been technically the same for a long time. We need to improve on these so we can improve them further later.

    Well yes, but there has been innovation in this space too. Some boards have started using 5mm plates. Polycarbonate as a material has sprung up. Hell, there was even a board with a PBT plate this year. Gasket mount is and was a thing and will only grow more popular, and new takes on old styles like the zeal65’s take on the sandwich mount continue to push ideas that already exist.

    – I also dislike PCB programming as it is unpolished.

    QMK configurator came out this year! I don’t think you’ve done any research into the state of QMK. You neglect all of the other progress that was made with PCBs this year as well! USB-C! Daughterboards! AIT! Like, C’mon, brother! It’s easy to use. Easier than ever! There are new configurators out this year from other community members as well. Maybe a year ago, the state of the configurator may have been unpolished but to call it unpolished would be dismissive of all of the efforts put into the project.

    – Compared to my other hobbies, keyboards seem underdesigned.

    Do you want overdesign keyboards? That’s really damn expensive, Brian. I would go as far as to say many first boards are overdesigned. THESE. BOARDS. ARE. NOT. FINANCIALLY. FEASIBLE. They never make it to market as a result.

    – The lack of innovation is causing collectors to leave this hobby and so it will shrink as a result.

    I think collectors are leaving the hobby because they with the recent refinements they’re finding their endgames easier and are done. Now they’re spending their money on audio equipment and anime tiddies.

    – Multiple Prototypes = Quality

    Getting lucky on the manufacturer roulette = Quality.

    – Repeated prototyping is expensive. I would pay more for a keyboard that has had more prototypes.

    Group buy runners don’t have that money, though. A certain amount needs to be put aside for the contingency of the buy. People won’t be interested in my board if it’s $600 and another person who got lucky on the factory roulette can get their board out for $300.

    – We need to innovate more and be more scrutinous of interest checks.

    I think we’re innovating plenty and we definitely have enough people to nag on interest checks.

    – The same designs have been used for years. We should branch out.

    As stated before, they’re safe. People like the fundamentals, so the market likes the fundamentals. The market dictates demand, so if the market gets the basics (and the units move accordingly) you’re not going to get the innovation you want. Come back another time when the markets saturate.

    1. Hey Krelbit.

      Thanks for the comment. I agree with most of what you just said, and you and a few others who have made good points that have changed my mind on the state of keyboards.

      1. My expectations for what keyboards that are designed should look like are much higher than what our market can support. While I do still think that we could do much better, I concede that realistically it’s not feasible.

      2. Aside from the specific case design and performance examples I mentioned in the article, there has been innovation in the hobby. I’ve ignored those (to my detriment) because of my bias towards the things I find important. All the things you’ve mentioned are innovations, so I concede that point.

      3. I brought up QMK not to bash it, but to make an example of how bad the field in general is if even the best isn’t as polished as it should be. I agree that there has been a lot of improvement in PCB programmability, but it’s still not polished. The VIA Configurator seems great, though.

  5. I have to agree with MechMerlin, LivingSpeedbump, and Krelbit. I don’t want to restate what they have already said so much better, so I will just add two things:

    1. If you do not think we are currently living in the Golden Age of Custom Keyboards, then you aren’t paying attention.
    2. If you want a better keyboard, make it.

  6. I’m in this hobby for only two years now and i think you did write an nice article.

    And i think you’re partially right, from watching top clack every week it really feels like designs are repeating more manufacturers are flooding the market and take inspiration from bestseller boards.
    and designer boards are sold out very fast in a limited amount because those designers care about QC and want to keep their name.

    But if i’ḿ taking a step back and take a look at what has changed over the last year that seems like a lot to me.
    And to me who currently owns 8 boards i still feel like i have much to explore, i did try some pcb design and currently am designing a case for the quefrency.

    And to me as someone who is interested in computers qmk is pretty easy even when you have to copy and change the keymap from another user and have to compile it.
    I would like to see some more fancy features under the hood instead of a noob friendly gui.

    Also i would not like the alice on my desk because she looks overdesigned to me.
    i prefer a more boring fjell or hhkb.

    The community is growing, that brings in more people that just want an good or even overpriced great keyboard.
    But it also brings in new people that want to contribute, some by remixing existing designs and some by bringing in inspiration form the outside.

    Have a nice day.

  7. Really out of touch with the state of the hobby. That’s the trouble with the internet. Everyone thinks that their opinion should matter…and be heard.

Leave a Reply to mikethetiger Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: