What does "intuitive" even mean?

Two remarkable posts on HN today: the new release of Affinity is now free and Canva-subsidized and Free software scares normal people

There is a bit to unpack as to why these two are related: the first is about Affinity, which is a remarkable and very feature-complete suite of applications for graphics, and the other is about Handbrake having an “intimidating” UI. The two are in perfect connection, for an important reason: we often parade “intuitiveness” as a virtue, but with applications that are tools very few people take the effort to unpack what that coveted intuitiveness is. Or should be.

🧭 I am currently available for contract work. Hire me to help make your Rails app better!

Intuitiveness is playing to habit

What is called “intuition” is, often, making guesses which turn out to be right. As far as tools go, you make a guess about a tool and you turn out to be right - the tool is, thus, considered “intuitive”.

• A pistol grip on a cordless drill implies that you can point it at the object (a screw or a spot where you want to drill a hole) and squeeze the trigger to make it go
• A button on a control panel has the same keycap shape as the button on your keyboard - you therefore know that it can be pressed and it is likely a non-sticky button
• A power plug on your appliance has two (or three) prongs just like any other power plug in your house. You know that you can stick it into a power outlet and nothing bad will happen (see also - @CursedCables on X).

There is nothing particularly intuitive about those tools that is “magical” - it is rather the fact that they are predictably like other tools. That is: if the user comes with a particular set of habits, they are likely to be able to accomplish the tasks they need with a given tool using this set of habits - the tool is, therefore, intuitive.

It is not that it is designed to delight. It is not that it uses the squircle of exactly the right shape which makes it intuitive. It is the fact that most other apps on that particular platform also use squircles which does.

The user has a habit, and playing into that habit makes things intuitive for them.

“Do thing” is not intuitive

In his post Daniel posits that the workings of Handbrake can be condensed into a drop well and a button called (I digress) “Do Thing”. The app is Magicbrake - and it’s good. Really.

But it is not intuitive. Let me explain.

It is intuitive in the UI controls it uses - some people still remember how to use a macOS drop well (although you would be surprised how much less common it has become, primarily due to the Electron dominance). And there is a button - fair enough. But the button says “Make small, fast and compatible”.

Here’s the problem: these terms have no established habit behind them. To be somewhat inquisitive, what do the statements actually mean?

• “Small” - what is “small”? It fits on a floppy drive? It will be below 10MB? It will be below 1MB per second of video?
• “Fast” - Fast what? Will the conversion speed up the video? Will it decode faster on a CPU? on a GPU? on a specific device?
• “Compatible” - what is “compatible” in this case? Compatible with what? QuickTime Player? VLC? iOS? Android? Plex?

None of this is intuitive - it is curated. And it is curated in a very non-transparent manner, with very unclear definitions of what it was curated for. Is it simple? For a given definition of “small, fast and compatible” - probably. Is it intuitive? Not unless the user already has a habit of using “small, fast and compatible” buttons in other video conversion tools. To me – not really.

Moreover, using terms like “fast” where their definition is clearly not well-conceived and not applicable hints at what I call “product management feast” - invent something opaque, slap a marketing term on it, drive it through under the moniker “it just works”.

If the goal is to make something “intuitive”, having a single button to “Do thing” is not the way to do it.

“Work exactly like the other thing” is intuitive

A while ago I have experimented using the Affinity suite instead of the Adobe applications I was used to for more than 30 years. And, in many ways, they have done a marvelous job on making it intuitive.

Did they replace the 100 UI elements with a single button called “Do Thing”? No.

Did they invent a new wizard which replaces cumbersome flows like layer duplication or type fitting? Also no.

They did something else: they have implemented 85% of the controls, workflows and names exactly the same as InDesign did it. Blatantly, unabashedly copied them down to the type editing shortcuts which diverge from the standard macOS shortcuts. For example, ⌥+⬅️ decreases the tracking distance between characters instead of “Emacs-unselect-word”. There is a hundred minute details where they took the time to copy, and literally so, most of the smallest interaction modes and shortcuts.

Moreover, as far as intuitiveness goes - the spots where Affinity Publisher, for one, turned out not to be intuitive, was in the workflows where they have diverged significantly from the InDesign model of doing things - for instance, in their treatment of placed images. The workflow in Affinity Publisher is very different with respect to those - and I found it not work very well. For example:

• Instead of scaling the “image frame” inside the “bounding frame” using a bounding box, you have to use a slider and a drag hotspot
• There is auto-fit which is not very well indicated as you work your layout - you have to check a palette or a modal for its state
• It would seem that positioning the image inside a frame happens at a different scale mapping (screen cursor movement to page distance) than other interactions on the page

It might be that some PM at Affinity decided that applying auto-fit to placed graphics would be intuitive - but they have, in fact, achieved the opposite.

Non-intuitive is doing things differently than everybody else, because with your unique snowflake interaction the user can’t rely on habit.

When habit is not usable, an emotional aspect comes into play that we can’t neglect either: if you show a person that “having done X this way for decades, you are now deprived of your muscle memory and you have to relearn everything because I, the software author, know better” can be very vexing. And it is not only about being told that your skills have become obsolete - it is also about being told that the skills you possessed, or learned, have become obsolete because the “new”, “improved” way of doing things is clearly superior to the one you have memorized and relied upon.

If the software you have mastered is complex and effective in what it does, it feels incredibly unfair that you have to re-learn everything just because some talented PM on the new Team Twample at Softco has decided to “change things up a bit”.

Another example of the “be like the others” is, of course, Blender. 3D graphics apps are known for having their own, very different and very special UI paradigms. Some apps use MMB-drag. Some apps use spacebar-drag. Some use Y-up, some use Z-up. Some have customizable interfaces with floating panels, some - have only panes. Some have several modes for selection, and some have one.

Blender, for a very long time, was known for forcing its users to learn a UI which was extremely divergent from all the other apps on the market (Maya, 3DSMax and even Houdini and Softimage) in pretty much every way possible. For the longest time the authors continued persevering in their vision of “we are doing it right, and everybody else is doing it wrong” - partly because of legacy code, partly because of their own habit.

But at a certain moment the Blender foundation understood that the key to popularity - and widespread adoption - is not robbing their users of habit. It is being like everybody else. They aligned their UI with industry-standard workflows and shortcuts, making it possible for users trained on Maya or 3DSMax to transfer their habits directly. The results speak for themselves: Blender adoption is soaring, and features are being worked in it - something we could not even fathom just a decade ago.

Creating intuitive experiences requires humility

I would recommend watching this video to understand the tradeoffs involved when changing feature-full applications. The interesting aspect of it is that you would say that Tantacrul is driving the UX of both MuseScore and Audacity in the exact opposite direction of what “intuitive” means! There are more features, more workflows, and more modes!

Yet – nothing is further from the truth. Making something intuitive requires humility. It takes humility to admit “No, I can’t make this magical and dumb this down to the level of a Duplo set so that their grandma could use it - this is an application which serves a purpose, and by trying to create Delightful Magic I will make it unfit for that purpose”.

Which is exactly the deliberation that Tantacrul highlights in his videos, time and time again.

In summary

If you want to make an intuitive application - do what everybody else is doing. Most of the time it is going to be exactly what is needed. Do not dumb down the tools, do not destroy people’s workflows - but be humble in what, how and why you curate and design. If you find yourself with an urge to create something “magical” - slap yourself on the wrist immediately.

You are not the deity to decide what “Do Thing” is supposed to be. Even Apple has failed at this repeatedly - remember Final Cut Pro X’s initial release, or the iWork redesigns that stripped out features professionals relied on? They eventually learned their lesson (Final Cut regained professional respect, Pages and Numbers restored missing features), but the cost was user frustration and lost trust. You don’t want to be in that company.