Lessons learned from building an enterprise bot
In 2018, we launched a Ai-infused Bot, a enterprise bot to help our customers manage their project and portfolios. We launched with a single use case — the ability for a project manager to get a pulse on their project status in Slack. Over the past year, we systematically grew the MVP and expanded the product’s value. Reflecting over the past year, I wanted to share three insights we learned — two we learned along the way, and one that we got right — on the journey to product market fit.
Insight 1: Find your evergreen product values
Admittedly, we got caught up in scripting the bot instead of understanding the user’s goal when interacting with the bot. We had dreams of our users having conversations and our bot telling the user what they needed to do next. This resulted in user-bot conversations being step-by-step.
When I visited one of our early adopters, I noticed that the champion after a day of using the bot, was trying to type in power commands. After a cursory experience with the Bot, users wanted to get the information quickly, without having to go through the steps.
We got our product values wrong. As Jeff Bezos describes:
You can build a business strategy around the things that are stable in time…When you have something that you know is true, even over the long term, you can afford to put a lot of energy into it.”
We ended up retooling our product strategy to focus on evergreen product values. We ended with three core values:
- Faster: Our users want to reach their goals faster
- Easier: Learning a new app is always hard. Reduce the friction from learning how to use a product by making it more intuitive
- More Proactive: Our users wanted to make sure that nothing fell through the cracks.
For example, our users wanted the experience to be faster. The responsibility of the bot is to let the user accomplish their intent as quickly as possible. This manifests in two ways: user takes less steps, and speed up bot’s reaction time. The team spent hours agonizing how to reduce steps from three to two steps.
To speed up the bot’s reaction time, we starting tracking performance to prioritize where our optimization efforts should be.
Figure out what your customer will always want from the product experience and work to achieve that lofty goal.
Insight 2: Use a different lens
Another missed assumption was who our growth user was. We had assumed that our core user was the Project Manager persona — after all we are a project management app.
However, while Project Managers drive strong usage, we found that our growth user was a wholly different persona. We had made the mistake of assuming that our core users were exactly the same as our web product.
Users that find more value from the Bot have a more causal interaction with our app. They were either managers that wanted some lightweight reporting to see how their teams were doing, or individual contributors who were responsible for logging their time on a project.
Turns out our sweet spot was less about being in the middle, but being at the edges:
Don’t assume that your core users in one platform are the same when you extend to a new channel. Apply a different lens starting from the benefits of the channel to find your core audience.
Insight 3: Meet your users where they are
We’ve been able to grow the product with the help and advice from our users. What we did right was to continually communicate with our users — even those that said ‘No’ during the early stages. When someone ‘No’ to your product, its because the cost of adoption outweighs perceived value. To get them to say ‘Yes’ you’ll need to demonstrate that the value is 9x the cost.
Customers partnered with us because of a shared vision,
but left because we failed to execute.
Find users who want to love you
When searching for traction, your best users are ones that once said ‘Yes’ but now say ‘No’. Every month, we’d continue to update all users (including those who churned) in a simple monthly note on our progress. In the update, I’d include new improvements and ask a simple question. The update made them aware of what’s going on, but the question made them invested into what’s going on.
The constant communication served as feedback loops for us to recover from missed opportunities. Ultimately, we were able to find hot problems to solve and successfully execute against it.
Openly and consistently communicate with former and current customers
Over the last year we’ve learned a lot about how to build better team chat experiences for our enterprise user-base. This year, we’re expanding the Bot from Slack to Microsoft Teams. Our ability to find the growth market have been largely due to the following insights:
- Find your evergreen product values
- Use a different lens
- Meet your users where they are