When Vancouver-based design and branding agency smashLAB was conducting research in fall 2013, Eric Shelkie used IRC (Internet Relay Chat) during the process. Like a bulletin board or chat room, IRC allows users to communicate with one another through text messages that become a flowing conversation thread.
Today’s analog to IRC would be the comment threads we see under Facebook posts, where one comment follows another, oftentimes resulting in a back and forth discussion. Text messages are similar, although in a closed system. And while Twitter also allows for public replies and discussions, you’re limited to 140 characters. When Shelkie shared his IRC findings with Eric Karjaluoto, the two of them decided to create their own IRC, and named it Chapp.
And why not? smashLAB is in the business of designing digital experiences not only for their clients, but also for research and development, as they explain at their site. A Twitter client named Guuda, a top-five tool (appropriately) called MakeFive, and their own CMS deemed shiftCMS are just some of their self-driven projects.
As an easy-to-use tool for carrying out conversations about any topic, Chapp was soft-launched in May, with a limited number of people invited to use it. Although I received an early invite, I’ll admit I was hesitant. Between Twitter, Facebook, Google+ (yes, I use it), and a range of other social media tools including Linkedin, I felt that the last thing I needed was one more social media outlet to manage.
But having (finally) used Chapp to conduct an interview with Karjaluoto this summer, I now see how wrong I was. Chapp is less about social media, and more about carrying out conversations–and you’re able to do so within an elegant and user-friendly interface.
Eric Karjaluoto on Creating Chapp, A Conversation Tool
|Meet Eric Karjaluoto and experience his presentation on “The Design Process” at the 2014 HOW Interactive Design Conference series in Washington D.C., September 3-5.|
The full interview with Karjaluoto about Chapp, which was conducted as a private channel at Chapp, resides below.
jasontselentis When you began creating Chapp, what were some of the first steps you took during the initial research phase? This can be technical, user related, functional or aesthetic.
karj Chapp started because my business partner, Eric Shelkie, was using IRC to ask questions about building cryptocurrency miners. As he did so, he told me how useful IRC was, and how active certain channels were. What surprised him, though was that IRC had hardly changed since he first used it, many years ago.
I’d never used IRC before, so my first step was to learn about it. I found a client, located some channels, and just played. Doing so wasn’t particularly easy. Although I could sense there was utility to IRC, having to read how-to articles just to set it up seemed sort of bonkers. Meanwhile, the UX of most IRC clients felt unintuitive.
jasontselentis How have you gone about making Chapp’s UX more intuitive? For me, the UX has a lot of familiarity: there’s a typing/image input that’s similar to what I’ve used in apps on my phone and tablet; and I can follow the thread above, to see what’s been said. Chapp isn’t cluttered like the old bulletin boards I used to use. It’s also easy on the eyes. Are these the kinds of things you built into the UX, to make it more intuitive? What else?
karj I think we’re making the UX intuitive by concentrating on essential functions. We generally take an “as little design as possible” approach to our work. This is why you see very few options, icons, and treatments in the current iteration of Chapp. (In fact, even the logo is essentially hidden.)
Upon getting a working prototype in place, we launched, and added elements as we better understood what might benefit the application’s functionality. For example, we’ve added flashes to notification alerts, and now highlight comments made on a channel since the user’s last visit. These don’t sound like big things, but getting them right makes the app feel better for our users.
The current iteration of Chapp is still rather primitive, as it’s essentially a mobile web application that can be used on the desktop. We wanted to develop this first, so we could get something in users’ hands, and learn from it. Now that we’ve seen it in use, we’ve been able to build out a more comprehensive roadmap for the mobile product. We’re also working on a desktop version of Chapp that makes better use of the large-screen environment.
That said, even as we add new elements, we’re trying to do so as efficiently as possible. That’s why you won’t find any tooltips or instructions on how to use the application. We’re forcing ourselves to make the UI intuitive enough that new users can make sense of it without needing it explained to them.
jasontselentis What have usage numbers looked like so far? Are the initial users returning daily, weekly? How many new users have you adopted?
karj It’s still quite slow. That said, we have a small handful of regular users who come back every day, and seem to be getting something positive from the experience. (This feels like a positive sign, and the repeat activity makes me happy.)
Community-building takes time, and I think folks are still unclear on the many ways they could use this application. It’s my job to change that, but Chapp’s something we work on in between client projects. As such, we never have as much time as we’d like to reach out to new audiences.
jasontselentis What do you hope to gain and learn from all of the work you’ve put into Chapp? Are you monitoring the research and development, looking at KPIs (key performance indicators) to apply what you’ve done here to other work in the future, other client work? Or is this all strictly done out of curiosity? Maybe just a great self-promo tool?
karj We’re curious and like making things. As such, I think the main reason we build applications like Chapp is because doing so is fun. When you’re working directly with users, you can push out new changes quickly, and focus on making the experience as good as you can. We like that immediacy, and opportunity to hone our skills.
There’s typically less room to experiment with client work, as such groups tend to look to studios like ours for more assured results. As such, it’s with projects like Chapp—and the others that have preceded it—that we build our skill-set, gain applied UX knowledge, and are able to test new technologies and approaches. I think all of this affords our clients a lot of value—because we’ve often seen for ourselves how certain approaches have worked, before we propose them to clients.
That said, Chapp is intended as a serious undertaking of its own. Private channels offer a great way for working groups and corporations to interact with one another. These channels are equally useful for families, teams, and clubs that want an easy way to share notes, without having to flood one another’s email boxes. Additionally, public channels hold a lot of utility for groups to access peers and other knowledgable individuals, and get answers to their questions.
We feel there’s a lot of opportunity with this application to service a number of different use cases. That said, we have a long way to go still, and we’re working hard to advance the platform, as there are still many core functions that need to be built out.
jasontselentis When it comes to those functions you hope to add, what kind of A/B Testing have you done, and where?
karj Chapp doesn’t yet see the kind of volume we’d require in order to make A/B testing worthwhile. Actually, I think this is an important note for many UX people to remember: without a large enough pool of users (and use cases) A/B testing can do more damage than good.
In my mind, A/B testing works best in granular settings, in which outcomes are clearly determined. (e.g. Does our landing page have better conversion rates when the Buy button is at the top or bottom of the page?) When you’re building something less defined, though, it’s difficult to narrow down a test in that granular a fashion.
So, our process is much more akin to the way you’d build something out of physical materials: Determine needs; Outline a plan; Sketch elements; Build a working model; Use the prototype; Learn what doesn’t work; Correct; Repeat.
I think part of why this works is that Shelkie and I use Chapp so much. As a result, we keep seeing (for ourselves) the real functional mistakes/oversights we’ve made, while we’re working with it. For example, I know that navicon in the top right of the screen needs a bigger target area, because I often find it unresponsive. There’s still plenty broken here that we’re well aware of; however, we also have to prioritize which elements most deserve our attention.
Similarly, we really listen to our users. You can see this for yourself if you visit http://chapp.is/chapp and http://chapp.is/chappbugs, or even on sites like: http://www.producthunt.com/posts/chapp.
We’re lucky in this way: Our users effectively play double-duty as high-quality beta testers. They help us spot issues that have slipped through our fingers, and we work with them to advance the UX.
Some requests don’t fit with where we want Chapp to go, or they have larger implications that make them impractical for us to address. However, for the most part the feedback we receive is very useful. Some changes take us time to implement, just due to their scope; meanwhile, we’ve also implemented some smaller user-requested changes in as little as 45 minutes.
jasontselentis What have been the reactions to some of those changes? Have users noticed them? Do they comment on them? Being so active on Chapp, I suspect that you’d see the compliments as well as the complaints.
karj It all depends on the nature of the change. Some revisions and new features go by largely unnoticed, whereas, others tend to be received with more notable reactions. For the most part, I think UX design should be about getting out of users’ way—and when we do so, we see the most positive responses.
For example, when we started the service, we looked upon it like a real-world conversation. In such a context, you can’t change the nature of what’s already been said, you can only add to it. Thus, we chose to not make comments editable. However, we soon realized the problems with this sort of approach: most notably when users would accidentally bump Return/Enter, having posted their comment prematurely. Many asked us to change this by adding an Edit option, and they were quite happy once we acted on their requests.
This is what I mean when I say, “getting out of the way.” Our visual design approach wasn’t impeding their ability to act, however, the absence of this function was frustrating for them. Now, an edit button isn’t a particularly sexy feature that most would even notice; however, it’s in making usage a little easier that we get closer to building something that might eventually become a very useful tool.
Similarly, some were telling us that they wanted a way to acknowledge someone’s comment without needing to write a response. They explained that they wanted something similar to Facebook’s Like button, but without being so much a thumbs-up. That led us to add the Nod, which is an easy way to indicate that you’ve received the message, or that you appreciate the sentiment.
Although we’ve had feedback on things that might be missing from Chapp, so far, we haven’t had much criticism for new features/approaches. I’d expect that to come with time, though, when the time comes to push through bigger changes.
jasontselentis In what ways have you seen Chapp used that you did not expect?
karj When we first envisioned it, we thought comments would be really brief, much like they are in standard chat applications. However, it’s not uncommon to see very long, detailed replies here. I find this sort of exciting, as I think it’s indicative of a good community in the making (as opposed to a whole bunch of “Hi” comments tossed in by random visitors).
In a way, I wonder if Chapp is less like IRC, and more like an alternate approach to forums—just given the kind of dialogue we see happening on it. I suppose this is endemic of just where Chapp is right now: it’s a little like this, and a little like that, but sort of something different—even though we haven’t quite put our finger on how.
Regardless, the length of posts has resulted in UX changes. The comment field was quite shallow, initially, and has since been made much larger, in order to give users more room to compose those longer remarks. In the next iteration, you’ll see this area simplify even more, to buy back more screen space for commenting.
The use of animated GIFs here has also been a bit of a surprise. (That said, I’m surprised by how that format has become a bit of a thing on the web in general.) I like how they add some playfulness to the community, however, they’re a bit of a pain to work with, as most of them have pretty substantial file sizes, and therefore they’re troublesome to load without performance issues.
jasontselentis Playing devil’s advocate: why not disable image uploading?
karj Because image uploads add utility to Chapp, and that’s what we’re most concerned with: making a good tool for our users. The fact that some features are challenging to implement doesn’t mean we should discard them; instead, we just have to keep working to integrate them smoothly.
Additionally, most images are fine. The challenge we face is on channels like http://chapp.is/AnimatedGIFs, where we often see animated GIF files running between 5 and 10 MB a piece. Loading all of those rapidly—particularly when scrolling through many posts—is still not working the way we want it to. Shelkie has implemented some lazy loading to make it feel better, but we still have work to do.
jasontselentis Seeing Chapp evolve must be exciting. And I can understand why you need to and want to let certain features stay put and become useful or more useful, or used in new and exciting ways. And adding features where they’re needed, or where they’re requested, makes sense to a degree.
But at what point would Chapp become something like Tumblr’s waterfall of imagery plus Twitter’s conversation thread full of replies and mentions? Or are you more interested in just seeing where Chapp goes, no matter where it goes?
karj It’s really hard to say where it will end up, as we haven’t seen it run at scale. As volume increases, I think we’ll see use cases change, and perhaps involve ones we’ve never anticipated. I’m not that worried about this turning into Tumblr or Twitter, though, just because they’re such different things.
At first glance, many web applications seem similar to one another. For example, I first thought Twitter was essentially the status update feature from Facebook, living on its own. In some respects, I suppose this notion isn’t entirely inaccurate, however, it is a very different tool.
What’s important to keep in mind is that minor changes to the UI and overall approach can shift the nature of a community/application almost entirely. (At one point, Posterous and Tumblr seemed like they were doing the exact same thing.)
I think the utility in Chapp has a lot to do with smaller groups. For example, Shelkie is friends with a bunch of guys who like to go on mountain bike rides. For them, Chapp is a way to quickly figure out dates, times, and locations, without needing to inundate one another with email. For my family (me, my brother, and our parents), it’s a way to stay in touch on a daily basis (in a private setting) when we have a moment to write a few words or post an image. And then there are discussions like this one: http://chapp.is/GDCNational in which a number of designers started talking about the GDC and their feelings on how it should evolve.
My point is that the potential uses of Chapp are many more than I can properly wrap my head around (which makes marketing more difficult).
Similarly, each service has certain unique characteristics. I’m still not entirely clear on what Chapp’s are, but I’m starting to see certain behaviors that indicate that this is a distinct tool. So, we’ll push forward, and keep an eye on the ways people use it. As they do, we’ll just try to keep up and make it as good for them as we can.