App licensing multiplier

Rodrigo Saiani

Hey guys, so here's a new question on the business end of fonts.

So from similar or adapted InDesign files, you can generate a PDF and a file for app publishing (iOS/Android for example). This kind of licensing is known to be different and more expensive. For example fontspring charges 10x base price for that. What I'd like to know is how do you guys approach this on your foundries.

I for one feel like it should be dealt on a case by case basis, with variables such as: is it for Zynga, which will make millions from an app, or is it a free app that appeals to a more local audience.

What do you think?

Thanks!

Ralf

I for one feel like it should be dealt on a case by case basis, with variables such as: is it for Zynga, which will make millions from an app, or is it a free app that appeals to a more local audience.

That was the dilemma of the last years. No foundry wanted to publish pricing, so when Zynga would call, the foundry could ask for much more.
But from the perspective of the users this is terrible. Apps and eBooks are so common today, we just need to offer these licenses directly in our web shops. And currently more and more foundries are doing it. The models and the pricing vary greatly though.

I tried to set up our model as simple as possible.
eBooks license: 2 times the desktop price
app license: 3 times the desktop price

And just like the desktop fonts, the eBook and app licenses include 5 eBooks/apps. So if you create an app for iOS/Android/Windows Phone you don't need to pay 3 times.

Our app license does include a limit of 500 000 app downloads though. So the few big players would have to negotiate a custom license.

Mostafa Hajizadeh

Are these talks about number of app downloads and web visits brought up just because they can be measured easily? Before these things, foundries wouldn’t say clients have to pay more if there are printing a million copies, instead of a thousand. Or would they?

(The case differs for hosted web fonts, because they can reflect hosting costs.)

Craig Eliason

Before these things, foundries wouldn’t say clients have to pay more if there are printing a million copies, instead of a thousand.

I think "couldn't" is the word you want there.

Ralf

Before these things, foundries wouldn’t say clients have to pay more if there are printing a million copies, instead of a thousand.

The difference is in the definition of font use. Font use happens when the original font software (!) as TTF/OTF is used to render text. It doesn't matter how many copies of a book are printed from a static PDF file. Printing the pages is not the font use. It's just duplicating the results of design work done with the font software. You need the license to do the design work, not to replicate it later.

For websites, apps and ebooks you actually mass distribute the font software itself to be used on each and every device. It's like an InDesign user with one desktop license would expect to be allowed to burn InDesign on 10.000 CDs so every user of that CD can open the included InDesign files. Distributing eBooks with bundled fonts is the same thing.
It needs a special license that reflects this distribution in contrast to the desktop use on 1 to 5 computers.

Ben Blom

Our app license does include a limit of 500 000 app downloads though. So the few big players would have to negotiate a custom license.

I do something like this for desktop fonts. End users for which the added value of the use of a desktop font, is much higher than the standard license price of that font, have to purchase a custom license. (Here “use of a desktop font” does not mean the number of users who have a computer with the font, but what those users — perhaps only one user — produce with the font.)

Mostafa Hajizadeh

For websites, apps and ebooks you actually mass distribute the font software itself to be used on each and every device.

This makes more sense now. But I still think most websites and apps (and almost all ebooks) do not let users do anything but read, and so are a lot like replications of a single design.

Mostafa Hajizadeh

I think "couldn't" is the word you want there.

It should not have been that hard, specially for large publishers which are the most important clients in this case. But I think Ralf’s point is the main reason for this: “you need to license to do the design work, not to replicate it.”

Ralf

But I still think most websites and apps (and almost all ebooks) do not let users do anything but read, and so are a lot like replications of a single design.

Yes, but it is still “font use”. The font is rendered on that device as if it is installed.
The different purpose (e.g. reading an ebook vs. designing commercial print work) is reflected in the price. While using the font for print design might cost $30, reading an ebook might just cost $0,003 per user.
If the design is “replicated” without the actual font software on the target device, e.g. as a GIF, then there would be no additional fee at all, even if it is published to millions of devices.

Ben Blom

But I still think most websites and apps (and almost all ebooks) do not let users do anything but read, and so are a lot like replications of a single design.

Yes, but it is still “font use”. The font is rendered on that device as if it is installed.

According to this logic, an additional license should be purchased for the embedding of a font in a PDF file (not as a bitmap image, but as a font which is being rendered). However, many desktop licenses allow the embedding of a font in a PDF file without an extra fee.

Ralf

an additional license should be purchased for the embedding of a font in a PDF file

And several foundries do exactly that.
PDF is a grey area according to “that logic” and that's why there are so many different restrictions in EULAs when it comes to PDFs.
A typical subsetted PDF is not like a font embedded in a website/ebook or app. There is no real-time text flow. The position of the glyphs is fixed and the full font is usually not embedded. Just the parts that are necessary to make it work. It's basically a simple vector representation of texts that uses parts of font files to keep the PDF size smaller.
The fonts could be fully embedded (e.g. for PDF forms), giving the reader of the PDF the option to create any text on that page and actually becoming a “font user” again. And that's exactly the reason fully embedded fonts in PDFs are prohibited in most EULAs.

Ben Blom

Although PDF files are in general meant to be read, not to be edited, it is possible to edit them with a program like “Infix PDF Editor”. I guess few people do edit PDF files. But those who do, are in fact “font users”. When an embedded font is subsetted, these users can only use the glyphs available in the PDF, so then the use is limited. When the complete font is embedded in the PDF, these users can use the complete font. All in all, when a font is embedded in a PDF file, it can be used. (Whether this use of fonts in PDFs by a limited number of people, should be a reason to change a font license, or not, is a different story.)

Ralf

All in all, when a font is embedded in a PDF file, it can be used.

So?
Your talking about what can be technically done, I am talking about what licenses allow to do and on what grounds this is priced.

Ben Blom

Ralf, I do agree with your general point. You suggested that one who creates text in a PDF, is a “font user”. I suggested that anyone can become such a “font user” (with a PDF in which a font is partly or fully embedded). The more complete a font is embedded in a PDF, the more full such a “font user” can use the font.

“So?” You suggested that what can be done technically, may be relevant for a font license. You suggested a relationship between being able to create text in a PDF with a fully embedded font, and the EULA of a font (which prohibits full embedding). I am sorry if I misunderstood you.

Jack Jennings

TL;DR: price like an insurance salesman / at least stop calling your license "Desktop"

---

Pricing "Desktop" and digital* fonts differently mixes two different models that I'm not sure are compatible. I have a feeling that this is why font license pricing feel arbitrary and confusing to so many potential licensees.

Per User Model: users vs. "users"

This model casts the font software as a "tool" for a given set of users. The number of people who can use the tool is limited by the license (with the cost rising accordingly). 5 CPUs = 5 "tool" users, etc.

Pricing "Desktop"** licenses this way works fine, but this doesn't allow your pricing to scale with the application of the typeface into digital contexts. Your mental model has to shift from users to "users". You are no longer dealing with users of your tool, but also of unwitting end-users who get the font software bundled with other software. The former are valuable, because they actually buy fonts, but the latter are to a high degree less valuable because most of them don't know what a font is. Users then start to have various subjective values, you have to decide how much each of those users are worth (even if they never even know they have your font software). Yikes.

Per Render Model: "uses" vs. uses

Here, the pricing model casts the font as a "renderer"-software. The number of times that the software is used to render output on a different device counts towards the total price.

We can easily apply this model to digital uses. How many times is your app downloaded? How many unique visitors do you get per month? Each of these uses (downloads) counts towards the total cost of the license. We can't move this model backwards and apply it to our "desktop" licenses, if not only because they've rarely-if-ever been priced this way, then because the question of what is a print "use" isn't really answerable (book units? pages? reprintings?).

So…

So this puts the type designer in the place of pricing two (or more) licenses, using two different conceptual pricing models. Since the models are no longer linked by a common unit, the relative price points start to feel arbitrary. Maybe you feel x font is worth y as a tool for a user, but has an average use value of z. Unavoidable subjectivity starts creeping in. Your paranoia about cracked iphone app downloads makes inflates your per user average price, etc. etc.

My hunch is that potential licensees sense that disconnect between models. When there are more options it gets worse:

Desktop, $35
Webfont, $35
Mobile App, $350
eBook, $70
Server, $350

…the potential licensee has many questions. Why are Desktop and Webfont priced the same? If I'm hosting a webfont am I required to get a server license? Why is an eBook twice as much as a webfont license? If I already have a desktop license, why do I need a webfont license if I'm making my website on my desktop? etc. etc.

In the end, none of these prices seem bound together in a way that is communicable to the buyer.

Jack Jennings

Risk-based Model: risky vs. very risky

Some of this seems to be solvable by pricing on "user risk".

You will have to assume for a moment, bear with me, that type designers are selling digital assets. (Knowing full well that the license is what the type designer sells, at least part of the value of the license comes from keeping the digital assets exclusive until someone ponies up [or at least prevents the type designer from employing a cadre of lawyers to actually police EULA infractions].)

Part of selling digital stuff of any kind is the risk of it being replicated, and the loss of revenue from that replication. Given that the context in which the type software is going to be used is known, we can come up with general estimates of how risky it would be to sell a license to a given customer based on how much of future sales would be lost from the usable-digital-asset-sans-license finding its way onto other devices.

Pricing on "risk" assumes the "tool"-software identity of typefaces, but only counts users as licensee organizations, not all "users" that might end up with your software unwittingly (app bundles, @font-face, etc).

With these assumptions, pricing per unit actually possible on a sliding scale across media:

0 < λ < ∞

λ = How many times will this given user, within their use-case, transmit (wittingly or unwittingly) the font files in a usable form to a third-party (who would otherwise buy the font)?

λ = 0 = The user is you. The file is on your hard drive.
λ = ∞ = The user intends to include the font software in an Open Source Software project that is bundled with every OS.

u = price you would need to sell x units tomorrow for, in order to recoup production costs

price = u * λ

So, given the base price of $5/weight/style:

"Print" User = λ6 (interns, sneaky freelancers, etc)
  -> $30
"Ebook" User = λ12 (using sub-setting as a deterrent, etc)
  -> $60
"@font-face" User = λ100
  -> $500

I'm just pulling λ numbers out of nowhere since I have no data to base them on, but the idea is there.

This allows you to find the per unit value of the license. Other pricing schemes can be derived from that point. That @ff user's license might be a monthly fee (a la a hosted service), in which case $500 unit fee might translate not-obscenely into monthly royalties from a service.†

The second parenthetical to λ is important here as well. If the use is high-profile enough to catapult the popularity of your typeface into both sales and torrent trackers, this might affect how willing you are do deal with this "risky" user for a smaller licensing fee. So then:

"Apple's New UI Typeface" User = λ100,000
  -> $500,000

App Users might actually be less "risky" than Webfont Users given the degree of difficulty needed to extract a font from an app on a smartphone (only developers are going to know how to do this, seriously), but still have their λ value increased through popular exposure.

Of course, there are strange bits to this system as well (or problems that it doesn't solve). How would you price student licenses, for example, knowing that they might end up "leaking" to more users in academia?

I'm not saying that this isn't how people are already pricing their font licenses. It probably is to some degree. You might still end up with the same justifications:

500k page views = λ1k
  -> $5000

But at least this is a way that you can base digital prices against established "Desktop"‡ license fees (the original question) given:

u = price / λ

And by pricing against risk you can tell your customers: "I think that I'm going to potentially lose λ licensees based on the way you are planing on distributing my font in your software, therefore...". Then they can walk away if they think you are wrong, otherwise you've justified a price tag that they might have otherwise walked away from to begin with.

* web, ebook, app, etc.; for lack of a better all-encompassing term
** now just a bad euphemism for "print"
† also a wild guess…
‡ please, please, please, stop calling your print license this

[Deleted User]

The user and all related content has been deleted.

Jack Jennings

In my defense, James, I did offer the predigested version at the top.

[Deleted User]

The user and all related content has been deleted.

James Puckett

Please knock off the trolling, James. Not everyone has had all this stuff worked out as long as you and some people would like to discuss it in detail.

George Thomas

@JamesM; the "Insightful" was accidental although in a certain way it is true. I couldn't figure out a way to delete it.

[Deleted User]

The user and all related content has been deleted.

[Deleted User]

The user and all related content has been deleted.

[Deleted User]

The user and all related content has been deleted.

Ralf

Pricing "Desktop" and digital* fonts differently mixes two different models that I'm not sure are compatible.

Not at all. That’s my point. “Font use” is always the same, only the purpose of that font use changes and therefore the price per single user.

If your desktop license specifies "CPUs“ or “electronic devices” this is just because the font needs to be installed on such a device. But you assume that there is always only one user working on such a device. And that's what you charge for. Should the device serve more than one user, then it's a server and needs a special license.

Same with webfonts: Pricing is (almost always) based on page views. This might sound very different than "CPUs” but it's not. It counts the number of users of the font software.

The problem with app and ebook licenses is that neither the foundry nor the customer really knows how many times the ebook/app will be downloaded. That's why the foundry might charge 10 times the desktop price. If I don't know if you have 100 or 1 million downloads, I better set up my price as high as possible.

Deleted Account

... and that makes it very hard to publish prices.

Rodrigo Saiani

I guess so David

[Deleted User]

The user and all related content has been deleted.