The first phase of my User Research explored human motivation and behaviors when it comes to communication in the workplace as it relates to collaboration, as well as general behaviors. This second phase explores usability issues with Slack’s desktop and web app.
To narrow my focus, I explored usability issues with Slack’s file sharing feature since collaboration was a huge pain point for participants during the user interviews, and also a reason why email and services such as Dropbox/Google Drive are ubiquitous in the workplace. This ties into the third hypothesis I made in Phase 1, which was as follows:
For Slack to be a far more valuable tool that extends far beyond the engineering/product teams in an organization, the experience of file sharing needs to be rethought in the current product.
Overview
Goal: explore the search and file sharing features of Slack, and understand its weaknesses to determine what can be done to make it a more impactful and empowering feature.
Hypothesis
Better usability on Search and File Sharing will empower users to be more successful at their jobs thereby creating more value, while also paving the way for Slack to be seen as a tool that goes beyond the Engineering/Product (includes Design) teams of an organization.
Background
I conducted 6 usability studies on existing and former Slack users, using the platforms that each participant was comfortable using. In this case, most of them used the desktop app, and two used the web app. No one used the mobile app. Half of the tests were done in person, and the other half were done remotely via Google Hangouts’ video and screenshare features.
Usability Tests
I gave participants a number of tasks to conduct. The objective of my usability studies was to test three primary actions. Testing for the primary actions were given in a way so as to minimize any biases and exclude leading the users. The three primary actions I was testing for were as follows:
- Sharing a file
- Interacting with that file (either by altering it or collaborating on it with others)
- Retrieving a file
Reviewing Notes
I first organized my notes by picking out the usability issues participants faced (see image above), as well as any points of confusion and pain. I organized these notes by color-coding each user on post-its.
Next, I took the post-its color-coded by user and organized them by theme to look for commonality. The themes were as follows: search and information retrieval & sharing files, interacting and collaborating with documents, right-side flexpane, previewing information, first-use onboarding experience on a new team, and miscellaneous. For the purposes of this post, since I’m exploring the deep relationship between communication and collaboration, I will do a deep dive into search and information retrieval & sharing files.
Interacting and collaborating with documents was another huge pain point where there was a desire to be able to mark-up documents in Slack, as well as view and track changes. As Slack doesn’t currently offer those features, I’ve chosen to hold off on sharing my findings on interacting and collaborating (the vast majority of behaviors in the user studies involve using Dropbox or Google Drive).
Theme 1: Search and Information Retrieval & Sharing Files
For the majority of participants, it was frustrating to attempt to retrieve information. One user discovered the calendar feature in search, but he also happened to be an engineer and perhaps the most technologically savvy out of the group — this discovery occurred serendipitously. Below are a few notable behaviors:
- The first instinct for most of the users was to scroll up in a channel (chat room) to find the information they were searching for. It took awhile to notice the right flexpane and search text box.
- For the 2 users whose instinct it was to go to search, it was difficult to retrieve information or a specific file in channels with a lot of noise.
- The preferred method for every single participant, when it came to searching and retrieving a shared document was to go into Dropbox or Google Drive. Additionally, files that a participant gave feedback on would often be done in Dropbox or Google Drive, and then Slack would be used to notify the receiver that there’s feedback on the document.
- Files that were work related or text files were typically shared in Dropbox or Google Drive with only the link shared on Slack. Often, the type of files that were shared on Slack but not on Dropbox or Google Drive were images. Most participants ignored Slackbot’s offer to convert or import text files. (I would be curious to know if this is because of a lack of awareness at what that does, or if it’s because it’s friction towards sharing).
- One participant who’s an avid user of Slack’s desktop app went to Slack’s web app to retrieve information he was searching for. In order to get to the web app, he went to the chevron drop-down where the team name is, and clicked on Configure Integrations. Next, he searched through Message Archives. There are two things to note here that are of interest. One: searching via the web app after using the desktop app is a behavior that has a bit of friction. Two: the user clicked on Configure Integrations, which has absolutely nothing to do with searching nor an intuitive term letting the user know that will take them to the web app.
“Oh I’ve searched for things before (in Slack). It’s a pain.”
-Account Manager, Enterprise
- When searching for a file, there was an expectation to be able to cross-filter by keywords. It wasn’t obvious to most of the participants that they could click on the Message tab or File tab to narrow by message or file, as well as narrow by file-type to find a specific type of file.
- The entire experience of sharing a file for one participant was confusing and frustrating. For files shared in a private channel, there’s an expectation that the comments wouldn’t be shared as well when the same file was shared in a public channel.
- Participants expected a file that is shared and collaborated on, and then shared again, to be considered one file and not two. Slack treats these files as two. A few of the participants were expecting to at least be able to somehow link files in Slack to indicate that they are tied together. There currently is no way to know what changes were made. One participant cited confluence which shows the delta between two versions of the same document.
- Two participants tried dragging and dropping a file to share with a colleague by hovering over the name in a private conversation. This is likely a pattern that was picked up from iMessage for desktop as both participants were avid desktop iMessage users (this interaction has the added benefit of improving the user experience on a micro level by eliminating the step of clicking on a colleague’s name to drop a file to share in a conversation).
- Slack was viewed as the more delightful communication tool to its competitors by the participants, but they felt that search and file sharing capabilities were severely lacking thereby keeping many participants in email and other modes of communication to discuss files, rather than using Slack.
Three Suggestions
- Create better search functionality by improving semantic search — this is a difficult (but not impossible) problem to solve as the complexity in designing the experience is challenging in addition to it being technically complex.
- UX macro-level improvement: The entire experience of sharing and retrieving files and information needs to be rethought and improved. This is a feature that could truly empower users and make Slack a far more valuable part of a person’s workplace toolset. This could take Slack from being a tool to casually communicate with colleagues, to a tool where more formal productivity occurs.
- UX micro-level improvement: Create better visual hierarchy and organization of the information in the flexpane by redesigning the flexpane to make filtering functionality more powerful, which would significantly improve the search experience. One solution could be to add two pill buttons with the search box — one for files and the other for messages. As a fairly common pattern it would visually draw attention to the search box, as well as communicate to users that they can filter by files and messages. The tabbed approach for filtering by files or messages is often missed — creating frustration with searching for a file. While I don’t typically do medium fidelity mocks since lo-fi is better for quickly getting ideas down, for the purposes of this post I have chosen to do this example in med-fi:
First-use Onboarding Experience on New Team
Although the focus of this post is communication and collaboration in Slack, It’s necessary to briefly call out the observed behaviors and frustrations with the first-use onboarding experience on a new team. One participant, who had used Slack at his old job, and now uses HipChat at his new job, went through this during the study and it was one of the most painful usability sessions I’ve ever observed in my 5+ years of being a designer.
This experience was frustrating for both the participant and I as we both spent about 15 minutes trying to figure out how new members could be invited, and the experience of signing in after receiving an invite was incredibly confusing. During the 15 minutes, he wanted to create a new team from the desktop app, but nothing happened, even after entering the code that was emailed to him. Eventually, it was so frustrating, that he quit and gave up.
“There were 20 f*cking steps to create a new team. This wouldn’t be easy at all for someone that doesn’t work in tech, and I work in tech. There’s so much f*cking lingo they have to learn. There are teams, channels, accounts, and private channels. The most frustrating thing was that I had to use a combination of this desktop app and my browser: having to juggle between the browser, this desktop app, and go to my email for a code that they’re asking for. I don’t understand why my email is involved.”
-Sales Engineer, Enterprise
One Suggestion
The entire first-use onboarding experience on a new team needs to be mapped out. One way to achieve this is to map each step of this experience on a note card, then lay it all out as a user journey map from signup to interacting with others. Once a user’s journey has been mapped and visually laid out, it makes it easier to understand what steps can be combined or eliminated altogether, which experiences can be modified or changed to be better, and where friction from signing up to using Slack can be eliminated.
Not all is bad with the first-use onboarding experience. Slackbot’s automated welcoming messages and having Slackbot onboard information for users was delightful to experience, and as one participant put it, it was nice and useful.
Conclusion
If the search & information retrieval and file sharing experiences were remapped, Slack would truly empower users to be successful at their jobs. Slack is great for workplace communication, but could be more valuable if it tied into how people collaborate with each other at work.
The goal of both the user research and the usability studies is to 1. understand human behavior as it relates to what’s being studied, and 2. understand the shortcomings of Slack and what can be improved on a micro and macro user experience level. This should all then inform the design with the goal of empowering the user rather than influencing them.
[Read Part 1 → A User Research Case Study on Slack: Phase 1, User Interviews.]
Disclaimer: I do not work for Slack. I’m a product designer that just really really really loves Slack and am an avid user and admirer of their product. If anyone from Slack is reading this, how do I get my hands on a pair of those really cool socks you guys have?
