What makes a Slack bot outstanding? Is it its technology, accuracy, or perhaps its wittiness? All are important — but above all else are the creative minds of their developers. This is precisely who we had in mind when we coined the Brilliant Bot category of our App Directory.
Today’s story features two of our friends in Moscow: Artyom and Mike, the talented developers behind one of our most popular apps, Statsbot, the analytical companion for Slack. Among your everyday team messaging, it’s now easier than ever for you and your teammates to make informed, data-driven decisions right in Slack. With Statsbot, answers to important queries are just a question away!
The inception of Statsbot
Both Artyom and Mike used Slack at their previous companies. “Leading my remote development team across different time zones was easy once we started using Slack”, says Artyom. “It became our main operating system for work, the place where all of our decisions were made.” But in order to make these informed decisions, someone from the team was responsible for fetching necessary data. Artyom began to notice a pattern where one team member would switch over to their analytics tool, take a screenshot, and copy and paste the image back into Slack. Since data was integral in driving the team’s collaboration, metrics had to first be discussed before a proper decision could be made. There was no way to streamline this workflow. Or so they thought…
A bot in progress
Among the earliest adopters on the Slack Platform was the dynamic duo at Statsbot. In November 2015, Artyom and Mike took on a hobby project, and successfully built a custom Slack integration, which later became known as Statsbot.
The first version of Statsbot was built for Artyom’s former team — it was able to fetch data from other analytics tools, allowing any team member to ask a question in Slack, receive an answer, and make informed decisions. It also allowed them to schedule daily, weekly, or even monthly summaries of important company or team-specific stats — so everyone was continuously informed about key metrics. Teams could also specify target metrics they wanted to achieve, like reaching X number of sign-ups, or receive notifications in channel when goals were met, so everyone could celebrate together.
“It started to work well at my old company, so I thought, this could definitely be useful for other teams as well!”
Artyom called Mike and told him about this idea. The rest is history.
Statsbot enters beta
First, the team had to validate their hypothesis with other people (not bots), so they called up a bunch of their hacker friends who were already using Slack at work. The goal was to make sure their new bot was solving real-world problems, rather than creating new ones. Within a few weeks, 15–20 teams were successfully running Statsbot. “People loved it!” they recall. They especially enjoyed how quickly they were able to make decisions as a team. Early feedback was overwhelmingly positive and they received a handful of bug reports and feature requests to work through.
The beta program expanded to over 100 teams and became integral in learning more about why teams were relying on Statsbot, how they were using it, the problems it solved, and what could be done to make their jobs even easier. Plus, it became apparent that the team had built something valuable for others.
“Once we realized that we were building a useful and valuable tool we decided that we wanted to focus on this full-time to prepare for a big launch. And so we quit our jobs!”
Getting ready for the big day
News of the Slack Platform Launch was quite timely for the team at Stasbot. Soon after Mike and Artyom left their full-time jobs, they heard about the Platform announcement, and put final touches on their bot. Within just two days of the platform news, they were ready to submit Statsbot to the App Directory.
Artyom distinctly recalls how helpful the team at Slack was during this process — they provided invaluable UX input on the optimal placement for the ‘Add to Slack’ button (which later brought them ~10% growth in conversion), and product advice on how to best implement multi-account support.
“Any cool product starts with a team. And I have to say that Slack Platform has a very strong one. Slack was not the first one to create a messaging App Directory, but they’re the first place where you can really feel the quality of the platform, the applications, and the people behind the scenes.”
@statsbot: How is Statsbot performing?
Once Statsbot was published in the Slack App Directory, the team was excited about amplifying the launch and promoted their app on Product Hunt, where it ended up being quite popular, collecting over 1,100 upvotes. Each day they closely watched as their install and engagement numbers grew — and soon Statsbot was featured in the Top Apps, New & Noteworthy, and in the Brilliant Bots categories of Slack’s directory.
“The App Directory is by far our biggest channel right now and it was an incredible experience to be featured on the main page. It drove tremendous traffic to our site!”
In just two months, over 5,000 teams had installed Statsbot, with 40% of those finding it in the Directory listings. At 500 installs per week, the team has achieved a total of ~9,000 total sign ups since their launch in January, including teams at Y-Combinator, Product Hunt, Khan Academy, Vimeo, and even NASA ADS.
Mike and Artyom tell us that they’ve barely scratched the surface of the opportunities that lie ahead for Statsbot. They’ve recently added a cofounder, Pavel, whose expertise in Natural Language Processing (NLP) will help advance the core technology that powers Statsbot, allowing them to expand capabilities even further.
With this additional brainpower driving Statsbot forward, Artyom and Mike have also concentrated their efforts on content marketing. To gain additional traffic, they’ve recently published two posts on Medium — one about their experience building on the Slack platform, “How to Release a Slackbot” and one thought leadership piece on the topic of “Data-Driven Collaboration”.
In the coming months, Statsbot is hoping to connect to additional tools, like MixPanel (which launches today on Product Hunt), offer even deeper integrations with existing solutions,and continue to support SQL databases. The team has also started to offer customized experiences for select customers and will further explore this area as a future business opportunity.
Brilliant tips for building a brilliant bot
For other bot enthusiasts out there, we asked the Statsbot team to share some helpful tips from their experience building a bot on Slack.
Here are just a few:
- Try it out first: Before releasing your app to the Directory, figure out if your product offers true value to customers. Consider running a beta program to test if users are engaged and submitting positive feedback. Before Statsbot’s big launch, they piloted their bot with over 100 teams, to make sure they were solving real problems for Slack teams.
- Ask for help from the Slack team: Our team is here to help make you and your app successful on our platform. Check out our many resources, including a new Ideaboard– a list of useful ideas that could be built into Slack apps, and reach out to our platform team with questions whenever they come up. During the app submission process, Artyom and Mike received valuable advice from the platform team about onboarding, UX, and product implementation. They continue to stay in close contact with the team as they further develop their app.
- Learn from other developers: There are over 400 apps published in the App Directory — which means there are hundreds of developers working on Slack apps! Over the past few months, many developers in our community have published articles with helpful tips and how-to’s for developing on the Slack platform as well as in-depth technical pieces. The team at Statsbot learned a lot from the developers at Kifi, Meekan, and Kip while drawing on the support of the helpful Dev4Slack developer community.
This is the second post in our new series of interviews with the people and companies behind many Slack apps and services, which we’re calling our Developer Stories. We’ll continue to add more stories like this very soon.