Slash commands are a quick way to perform actions in an app—without using a mouse. For example, in Slack, you can type “/leave” to leave a channel.
Dropbox Paper is an online document tool that supports slash commands, too. Type “/table,” for example, to magically insert a table into your doc.
Typing "/table" in Paper inserts a table
Although Dropbox Paper supports dozens of slash commands, we had 2 basic problems with this feature:
According to our analytics, people were more likely to become active users if they embedded content into a doc—and slash commands were an easy way to embed content. We hypothesized that if we teach people about slash commands, more people would embed content and become active Paper users.
To test this, we planned an experiment to see if we could get more people to use slash commands. We set up 3 user groups and randomly assigned users to 1 of the 3 groups:
Because this was intended to be a quick experiment, we wanted to keep the team small to move quickly. We had only 1 engineer assigned to this project.
To get inspiration for the UI and interactions, I did an audit of numerous other apps, focusing on how users can "quick add" content. Below are just some of the screenshots and notes I collected during this audit.
A small sample of my competitive audit
The main challenge (and most fun part) of this project was figuring out how to educate people about slash commands without getting in the way. Below are just some of the early ideas I came up with.
After weighing the pros and cons of each idea, we settled on a direction that felt like a nice balance between being unobtrusive while still being helpful.
When starting a new line, we displayed placeholder text that said: “Type "/" to quick add.” After typing “/”, a menu would appear where you could browse through our most popular embeds.
Educational placeholder text shown on a new line
New menu showing the different slash commands you can use
As you type, the menu auto-filters based on what you typed
More details about the new menu, which also supports keyboard interactions
This got descoped, but I also explored a way to create your own commands (called "shortcuts")
Modal where you can define your new commands
Once we had a working build, we released this feature to internal Dropboxers for one week to get their early feedback and to also do usability testing. Dropboxers loved the feature and reported no major issues with it. Now we were ready to ship the experiment!
We let this experiment run for several weeks, and here's what we discovered:
Not quite the results we were hoping for
But equally as important, we received a lot of negative feedback about the change from users:
Type "/" to quick add is super annoying and distracting. I came to Dropbox Paper because I want a distraction-free writing experience. This placeholder need not be seen at all times. It's really, really distracting. Scrap it off.
The "type / to quick add" thing showing up where the cursor is supposed to be in the document is SO distracting! It instantly makes Paper almost completely unusable. Please!
I get adding the hint once to improve discoverability, but until I'm off this test, I will have to switch over to another document editing tool.
It's incredibly distracting and prevents me from focusing on what I'm typing.
At this point, we had to make a difficult decision. We didn't see the results we wanted, so do we iterate and ship another experiment, or just move on to something else?
After discussing with our team, we ultimately decided to move on. We could have more impact by trying a different approach that didn't focus on slash commands. We ended up keeping support for slash commands, but we removed the placeholder text and menu to keep the UI clean and lightweight.
When we shipped this feature internally, Dropboxers loved it. However, we quickly learned that external users reacted very differently. Dropboxers were "pro" users of Paper, so our needs and behaviors didn't quite match those of our external users.
Our PM was very bullish about slash commands. The concept had so much potential because we saw it succeed in similar products. But the user feedback we got was clear. The concept was too distracting and heavy for Paper, where its simplicity is its strength. It was a tough call to abandon this experiment, but this allowed us to focus on building a better way to embed content months later.