One of the people involved in the project and answering questions in the article. Happy to answer more questions.
We talk about reasoning for Cereal in the article and in my case study too[1]. In large organizations, you often have tens or hundreds of designers, support multiple platforms (ios, android, web(mac, windows, linux) and do ad campaigns around the world. When it comes to typefaces, there really isn't any universal system fonts (other than Arial maybe), that are available on all platforms. If you use different fonts on each platform, the design will be slightly different, trying to manage a cross-platform design system (like we do [2]) becomes challenging. Every time, designer has to pick the right font, potentially do 2-3 designs with different font to make sure they all look how they want them to. Also then each time user or anyone in the company looks at the design on their device, it look will slightly different. If the marketing then also uses different type, it will look slightly off if you include product shots (one reason why Apple now uses SF everywhere). Short answer: having multiple fonts causes additional work and can confuse people, especially inside the company.
Since there isn't universal cross-platform fonts, it means you often have to embed fonts, and in some cases license them. Licensing fonts is different than just buying a font file. Even for small developers quality fonts can run up to $50,000 per app (imagine what the licensing costs are for a global company with large userbase!). With licensing you never get exactly what you want in terms of expression and functionality, and other brands can easily buy the same thing.
As the last point, what I find really exciting that we can now consider font kind of like a software. We change it whenever we find problems or just want to improve something. That is usually not possible with licensing. Obviously, you would and should only do your own typeface if you have these kind of issues and care enough wanting to solve them.
Congrats to your team on creating what looks like a very versatile font. I can appreciate the benefits that creating a custom font will have.
One of the other commenters mentioned that AirBnB's Cereal and Netflix Sans (both overseen by Dalton Maag) are almost identical. Could you speak to the differences, technical or otherwise between the results and how that relates to differentiating the brand aesthetic?
I wasn't part of the Netflix Sans or have insight into the process or have access to their fonts, so cannot really speak to the differences per say.
I can say that modern(since 15th century) latin based type, is usually based one of the traditional type families: Humanist (Venetian pen-writing, eg. Palantino), Grotesques (sans-serifs) and Neo-grotesques (more geometric grotesques like Helvetica). In terms of creating a usable typeface, you pretty much follow one or two of those traditional families, and then adjust terminals, proportions and other features. If you go too crazy, it might just look strange and not work well anymore. You can create something very distinct for a very special purpose, say a movie title score for a scifi movie, but for an UI, you want something that doesn't distract the user. Its like most visual design, you want it be pleasant, but user should eventually forget the design and focus on the task at hand.
We started this project a while back (18mo ago which is timeline not an active development time), and received and tested probably around 30 different font file iterations during that process. These included several directions of the font families based on our initial ideation workshop. Even when you're in the process, it can be pretty hard to see the differences, just looking some of the letters, unless you really focus and try to tease out the differences. Something like 'o' is pretty much the same in all same family typefaces, and most letters have very similar shape. In our case, we did a lot of testing around actually designing existing or new UIs and where is becomes much more apparent how all the letters and proportions work together. (If anyone wants to test it, then pick 5 or so very similar looking fonts like Roboto, SF, Helvetica, and design few screens with it with varying sizes, amount of content and weights, compare and see how the differences feel.)
Our goal since the beginning was to find something that works well in UI, and is expressive for marketing. Some weights like Book, Medium and Bold are meant for UI where as ExtraBold and Black are more towards marketing. We find that where we are now is definitely an improvement for what we had before. Both our brand and design is quite minimal and heavily relies on type, we think that it's important to control it and make sure it feels like us.
This is a brilliant answer, and not terribly far from what I was expecting. Thanks for taking the time.
Do you have any specific insight into where testing didn't succeed, or where other typefaces reached the limits of your constraints before deciding to go this route? For example, if a low-res phone screen couldn't legibly display type at small sizes.
Fonts are a very subliminal thing. Sure, you may not be able to point out the differences between certain fonts, but many can tell the difference between different corporate identities.
>If you use different fonts on each platform, the design will be slightly different, trying to manage a cross-platform design system ... becomes challenging. Every time, designer has to pick the right font, potentially do 2-3 designs with different font to make sure they all look how they want them to. Also then each time user or anyone in the company looks at the design on their device, it look will slightly different.
Where simply commissioning this font to be used globally for everything from product to marketing will reduce the amount of time spent on creating design materials (ultimately saving the company money or allowing their employees to focus on more important tasks)?
Or his other point where commissioning this font incurs a one off cost and reduced spending in licensing.
> Since there isn't universal cross-platform fonts, it means you often have to embed fonts, and in some cases license them. Licensing fonts is different than just buying a font file. Even for small developers quality fonts can run up to $50,000 per app (imagine what the licensing costs are for a global company with large userbase!). With licensing you never get exactly what you want in terms of expression and functionality, and other brands can easily buy the same thing.
> If you use different fonts on each platform, the design will be slightly different
Oh no, the fonts will look different on different devices! Better commission a custom font.
Back in the day, we just accepted that was how the web worked.
Jokes aside, I see why you'd want consistency across devices and print. But what's wrong with using an open source typeface, or modifying one for your purposes?
> and other brands can easily buy the same things.
Why do you think a unique font is essential for branding? Some well known fonts are in use by many companies.
I think people are perhaps downvoting you for the joke but it is a really important question.
A or B?
A - Design a custom typeface in 3 months for a project
B - Use a proven typeface from one of the world's best designers, say Adrian Frutiger's Univers that has been tested and updated for digital medium?
I am a bit old fashioned I guess. There is so much beauty in Didot, Univers or DIN than today's quickly churned out fonts. Technology has changed, but the methodology hasn't and an impeccable font takes a better half of the decade to perfect.
We talk about reasoning for Cereal in the article and in my case study too[1]. In large organizations, you often have tens or hundreds of designers, support multiple platforms (ios, android, web(mac, windows, linux) and do ad campaigns around the world. When it comes to typefaces, there really isn't any universal system fonts (other than Arial maybe), that are available on all platforms. If you use different fonts on each platform, the design will be slightly different, trying to manage a cross-platform design system (like we do [2]) becomes challenging. Every time, designer has to pick the right font, potentially do 2-3 designs with different font to make sure they all look how they want them to. Also then each time user or anyone in the company looks at the design on their device, it look will slightly different. If the marketing then also uses different type, it will look slightly off if you include product shots (one reason why Apple now uses SF everywhere). Short answer: having multiple fonts causes additional work and can confuse people, especially inside the company.
Since there isn't universal cross-platform fonts, it means you often have to embed fonts, and in some cases license them. Licensing fonts is different than just buying a font file. Even for small developers quality fonts can run up to $50,000 per app (imagine what the licensing costs are for a global company with large userbase!). With licensing you never get exactly what you want in terms of expression and functionality, and other brands can easily buy the same thing.
As the last point, what I find really exciting that we can now consider font kind of like a software. We change it whenever we find problems or just want to improve something. That is usually not possible with licensing. Obviously, you would and should only do your own typeface if you have these kind of issues and care enough wanting to solve them.
[1]: https://airbnb.design/working-type/
[2]: https://airbnb.design/building-a-visual-language/