How Statsbot is building a business on Slack
Our growth journey taking users from free to paid
Hello, Slack Fans! My name is Mike and I’m the Chief Business Developer at Statsbot. This is the story of how we built a business on Slack and got hired by over 20,000 teams.
You can build a business on Slack today.
As first-time founders, we learned how SaaS products work while pioneering a new bot user experience.
Statsbot pulls Google Analytics, Salesforce, and Mixpanel insights right into Slack. Just ask @statsbot a natural language question, and it will respond with beautiful charts right into your favorite Slack channels. Statsbot also uses machine learning to inform you of statistically significant anomalies, like when user visits spike or plummet.
From the very beginning, we defined growth as our most important target. We believe startups must grow before they focus on monetization. In the beginning, your business must grow or die. For us, well-timed launches on Product Hunt, Hacker News and Slack App directory brought us a lot of organic traffic and viral spread across Twitter.
It can be tough to figure out your business plan at first.
If you’re a developer, your main goal in the beginning should be to find product market fit before thinking about charging customers.
Experimenting with freemium and trial models
SaaS companies share a couple common business models. Two popular approaches are the freemium and trial models, and we experimented with both.
With the freemium model we found it more difficult to charge people, but with the benefit of a lower churn rate. People have time to fall in love with your bot before committing to pay.
The surprising benefit of the trial model is that it motivates people to explore the product right after installation, giving users extra incentive to onboard sooner.
Freemium versus Trial: Learnings
B2B startups cannot exist without generating revenue. Once we gained a large install base of teams and figured out our main use cases, we began exploring business models.
To jump-start our efforts, we participated in the accelerators 500 Startups and Betaworks Botcamp. It was just what we needed! We rapidly experimented with different ways to charge users, and adopted best practices from other SaaS companies.
Since we are not in sales, we started by automating the process from our landing page.
We had a good amount of inbound traffic to that page and hadn’t spent any money on marketing. Traffic came mostly from search, Slack Directory and Twitter.
We started experimenting with the freemium model first. We tried to add various restrictions on the number of connected integrations (Mixpanel, Google Analytics, etc), as well as on the number of delivered reports (only for new cohorts and for new functionality).
We found out that a paywall for features that do not limit the engagement of users worked best. Think about Spotify, for example: you can listen to unlimited music on Spotify, but as soon as you want to use more advanced functionality (music on demand), you hit the paywall.
Then we decided to go for a trial model, which we found to be more fitting to B2B products. We were afraid that a trial up front would scare away new users, but surprisingly, the results were exactly the opposite. We started getting more paying users!
Don’t wait until Slack comes out with their own proprietary payment solution. You can acquire paying customers using third-party tools, such as Stripe.
We recommend experimenting with small cohorts of users as early as possible. This will help you to understand the value of your product faster, and will help you find leads who really want the product and are ready to give you a ton of feedback.
Third-party tools we recommend
To receive payments and provide customer support, we recommend using a stack of ready-made solutions like Stripe and Intercom.
In the very early stages of your development, we strongly recommend you use existing frameworks such as BotKit, and strengthen your bot with ready-to-use machine intelligence tools like API.ai. Also, startups can afford to use workarounds and accumulate technical debt in order to move forward as quickly as possible!
In addition to Statsbot, we use other bots too. We don’t use personal bots too often; instead we use a few bots for work. It is worth mentioning Zoom.ai for booking meetings, Polly for ad hoc surveys and of course, Growbot for team spirit!
What’s coming next for Statsbot
After achieving growth, we’re now focused on product excellence . That is why we are building integration with SQL databases as a new skill for Statsbot, and are ramping up a machine intelligence R&D team.
Shoot us an e-mail if you want to try the SQL integration first!
