What we learned doing three pivots
Last week, we announced the launch of our new product named Plato. Plato is on a mission to help Engineers become great Engineering Leaders thanks to a community of mentors who are experienced engineering leaders from top tech companies, including Segment, Lyft, Facebook, Google, Spotify, Trello, Slack, Uber, Netflix…
Plato started as Birdly in 2015 as a very different product. These last 18 months, we went to YC, NUMA, raised a seed round with prestigious investors and pivoted three times. We pivoted a few times, learning each time something new, that helped us improve the next product. Yet, we went through thick and thin this year, made many mistakes, failed at a few things, and mostly, didn’t take time to understand why.
A few months back we say down and thought about where we’ve come from, where we are what we’ve learned along the way and where we want to go.
What happened at Birdly/Plato in the last 12 months?
During the last 18 months, we’ve worked on 5 different products:
- A mobile app for expenses
- Birdly expenses — A Slack Bot that does your expense reports
- Birdly info — An intermediary layer between Slack and customer data (SFDC, Zendesk, Intercom, Stripe)
- Birdly culture —A Slack bot focused on improving employee engagement, especially for sales team
- Plato — A slack-based interface to connect engineers with mentors to help them become great Engineering Leaders
We went through a lot and learned a lot. We understood our individual drivers and determined our “idea framework” to find the next right thing to build.
YC’s Sam Altman told us to find the right startup idea, you wanna look for the intersection of 1) what you’re good at, 2) what you enjoy and 3) what way you can create a value for the world.
Ours is not much different:
1) We need to build something that is solving a real problem for companies
2) We need to be excited to solve this problem
3) We need to solve this problem in an innovative way, better than competitors
Birdly in 2015
Our first product was a mobile app for expenses. We had this idea of building the same product on top of Slack, right before all the recent wave of bots, we explained here why we did that.
This bot for expenses was working pretty well, but we felt something was wrong. We didn’t really know why at the time. But a year later, we figured out it wasn’t fulfilling the three points of our “idea framework”
<✓> — Solving a Problem
<✗> — Exciting
<✓> —Innovative
It’s very important to be excited about the problem you’re solving. Not only because it gives you a sense of purpose at work (I’m a Millenial, it’s important for me :) but because you need to be super excited to do the extra-mile to understand deeply this problem. Did we try to have lunches with as many accountants, CFOs, office managers as possible? Have we been to this conference about new expenses and travel solutions? Not really…
December 2015 — Why did we change from Birdly expenses to Birdly info?
We had traction with our first bot in Slack and a compelling vision around the future of Enterprise Software in Slack. We’ve been to YC in January 2016 and began to explore all use-cases around Slack (Birdly for your business cards, Birdly for translations, …). We never really launched or went really far with any of these apps.
Then one idea came up, we could be the intermediary layer between business apps and Slack. We would rethink the way people interact with software. We would be the single bot that connects you with all your SaaS apps and processes. That was a big plan, and we needed to begin with one smaller thing.
We asked our users (from the bot for expenses) which software they wanted us to integrate first. The result was undeniable: Salesforce. We didn’t know anything about Salesforce nor how a Sales team was working. We decided to go anyway with that idea of connecting Slack and SFDC but without doing our homework… We didn’t talk that much to salespeople to identify what was their real problem in SFDC. Why? We weren’t salespeople ourselves, and actually didn’t care that much about their problems. Consequently, we were trying to solve a problem we weren’t much passionate about.
But we had lot of curiosity and traction thanks to Slack, increasing interest about chatbots and YC… and people were downloading the bot so that we went up to 3000 teams onboarded.
But… retention was not here. That is not a problem per se, you just need to work on it. But we figured out that the speed at which we were learning was not quick enough. We didn’t have the product-market fit, and week after week, we were not really closer.
We began to lose the faith of our employees who became less and less aligned with our mission. This vicious circle began to be dangerous as we were spending more time trying to explain why they needed to work hard without any arguments, faith or vision ourselves.
What about our idea framework? 2 on 3 checked:
<✓> — Solving a Problem
<✗> —Exciting
<✓> — Innovative
May 2016 — Why did we change from Birdly info to Birdly culture
Our 3 developers who have been to YC with us left the company. And at the same time, Cedric my good friend from NUMA told me that his startup (after going through TechStars NYC and raising a 1M round) was shutting down. Three reasons convinced me that I needed to do everything in my power to recruit him: 1) We had an exceptional fit about how we think and about how to build a company 2) He challenged our ideas and thoughts in a very interesting / eye-opening way 3) All developers he used to manage told me they would give everything to work for him again.
After two months of HR issues (leaving and recruiting) we got back to the product and thought about what needed to be changed. Two things:
- We wanted to build something with a strong use case (people were doing something similar manually or wrote a piece of code to do it)
- We wanted to build a bot in “push” mode VS “pull” mode
The reason for the second point was that we wanted to send information before any intent of the user to overcome the barrier for the user to think about Birdly at the moment we could be useful.
To identify the first point, we asked our users to show us their Salesforce and Slack and explain to us how they were using those two softwares together. We noticed one pattern, many companies had either built a bot to announce won deals (coming from Salesforce) or were announcing them manually. That is the whole beginning of all these things around “employee engagement” and “company culture”
Then we asked those same companies to tell us about their “company culture”. Another common pattern was that many companies had the “all-hands meeting” and the structure was pretty much similar: 1) Presentation of KPIs 2) Qualitative achievements 3) Q&A. That inspired us a lot and we decided to build a bot with similar structure. As we were already integrated to SFDC, we began with three features for Sales Team:
- The virtual gong: announcing and celebrating won deals
- The weekly update: announcing KPIs of the week
- Q&A: asking questions to the leaders
We decided to go manual first and did many things ourselves, to stop wasting our time developing features without prior discussions with our users.
October 2016 — Why did we decide to stop Birdly Culture?
Stats were:
- 26 companies installed our beta product
- 7 companies were very happy and shared heavily their successes in Slack
- None of them were ready to pay a significant price
Also, when trying to recruit teams in beta-test, We realized the three featured we offered weren’t really correlated: the value proposition and main target were too different.
Conclusion: The virtual gong was a very good “nice to have” product, a good Trojan Horse for a broader product, that we didn’t have yet… Also, it was obviously important for companies to have this celebration process, but it was not really a problem. We mixed up the two notions. Something can be important but not really a problem, and people would not pay a lot for that.
What about our idea framework? 2 on 3 checked:
<✗> — Solving a Problem
<✓> —Exciting
<✓> — Innovative
What did we learn?
- We need to solve a real problem companies will pay for
- We need to be excited to solve this problem to understand it deeply
- We need to solve it in an innovative way, because it’s our DNA
- We need to distinguish problem and importance
- We need to be rifle focus on one target and one pain point
- Don’t build a trojan horse if you don’t have the main product yet
What’s now?
A few months ago, Astrid, Cedric and Louis joined Plato.
Today we’re almost building a new startup with:
- A team of cofounders better individually and collectively (as we learned to work together)
- Cedric, Astrid and Louis who are amazing first employees, more than any founder can expect
What we took for granted:
- Respecting our idea framework
- Building on top of Slack because that’s our vision of the coming shift in Enterprise Software
- Building a product solving a problem we had ourselves
We are now building something new, more meaningful to us and hopefully, to you. Last week, we launched Plato.
