This past August, I sold my first SaaS company closing the final chapter on a 2.5 year journey into developing a user-focused technology platform, bootstrapping a business, and all the trials and tribulations that brought with it.
Now that I’ve had a few weeks to digest and reflect on this first foray into entrepreneurship I wanted to put my thoughts down. Hopefully the the process and learnings I went through can be helpful for future founders that are going through a similar adventure.
The business in question was Confetti AI, an education platform we built to help students and industry practitioners level up their data science and machine learning skills through in-depth curricula, extensive question banks, and targeted self-assessments all served through our proprietary software service.
My cofounder and I conceived of the idea for Confetti in January 2020 over a dinner in Sunnyvale.
Our beginnings were particularly unremarkable. We were discussing the sorry state of the machine learning interview process over some drinks and as we played out the hypothetical of what a better solution could look like, at some point we threw out of the possibility of just building the thing we were talking about.
The next day we were already working on architecture diagrams for Confetti.
Choosing to go down the bootstrapped route was a very natural decision for us.
We both had full-time FAANG jobs throughout most of our time working on the platform, and at project genesis we were just interested in scratching our own itch with Confetti.
In other words, we wanted to build something we ourselves would use and then expand the ambition to something that people would find value in.
At some point we would consider monetizing, but at the outset we were aggressively focused on building something that people would find helpful.
For a consumer-focused education SaaS, that meant emphasizing user growth, usage of the platform itself, and completion of resources like practice questions and exams.
Given that we were bootstrapping the platform while maintaining full-time jobs, the bulk of development in the first 1.5 years happened in the hours of 7pm-12am, 5-6 days a week.
Maintaining such a development cadence was certainly a challenge given our other commitments, but I will also say when you’re working on something that’s truly yours and the possibilities are boundless, your energy isn’t measured in 40 hour work weeks.
We worked tirelessly to build the best machine learning education offering out there, and I know we pioneered new product offerings for building machine learning skills that no other platform had. This is something we were incredibly proud of.
After 2.5 years when Confetti was finally acquired, our active userbase had grown to 6000+ including students at top-tier educational institutions like MIT, Harvard, and Berkeley as well as industry practitioners at companies such as SAS, Palantir, Pearson, and Grafana Labs.
On top of that we had built a profitable business that required minimal upkeep and was paying for itself.
So if Confetti was profitable and we needed to spend little time to maintain it, why sell it?
After around 2 years working on Confetti actively, my cofounder and I reached a point where our time was increasingly getting sucked by other projects and commitments.
We were very happy with what we had built into Confetti, but we felt we didn’t have the time or bandwidth to really grow the platform into the next natural stage of its evolution.
To do so would require doubling down a lot more on marketing efforts, growing out the team, redesigning the offering to be more enterprise focused, and potentially even raising some outside capital. After a lot of careful deliberation, we decided we weren’t the right people to expand Confetti in the way it needed.
But we were still excited about the value the platform could offer to users around the world. This is something we had clearly validated - there was a strong market pull for what we had built.
Therefore it made sense to see if Confetti could be integrated into a broader ecosystem where it would be nurtured into being an even bigger platform that could help more people.
This is ultimately what guided our decision to get the company acquired.
Acquisitions, like all aspects of company building, never have a single way that things happen.
Sometimes your company is big and well-known and acquisition offers show up at your doorstep periodically.
That was not the case for Confetti. We had to work to sell ourselves.
In practice, this meant taking a few concrete steps.
First we reached out to colleagues and acquaintances in our space, which were primarily companies in the technical education domain. During our conversations, we would typically float the idea of selling off our business to gauge interest.
We were looking for one of two things: 1) the company would express interest in buying us unpromptedly or 2) they would suggest people or other groups we should talk to.
Importantly, during these talks we would underscore the fact that we were not in dire straits and “needed” to find a purchaser because the debt collector was knocking on our door.
Rather we were looking for a good partner to help take Confetti to the next level, which was exactly the headspace we were in. We were fine to be patient and take no deal instead of a bad deal.
Pro tip here: it’s always easier to initiate acquisition discussions if you already have existing relationships with leads.
So if you’re a startup founder, I highly suggest establishing a connection with every relevant company tackling similar or adjacent problems early on, even if you would consider them your competitors.
That makes acquisition discussions more organic because you have a “warm” intro rather than just showing up at someone’s doorstep out-of-the-blue asking if they want to buy your company. The latter scenario is also a bad position of leverage.
For every conversation with a potential acquirer, we would always think through what the main value add of Confetti would be to someone else’s business. What did they stand to gain from partnering with us?
Answering that question would always guide our discussions and framing of the Confetti business.
Like all other parts of business, making the sale requires deeply internalizing someone else’s painpoints and empathizing with their problems.
Another thing we tried was reaching out to brokerages that you pay to initiate and mediate discussions with potential acquirers, though we ended up being too small for them to consider working with us.
Finally we listed Confetti on a startup acquisition marketplace.
In this case, we went with MicroAcquire which although it’s relatively new, had received great reviews from our friends. There are other such services founders might consider such as Flippa, Empire Flippers, and Investors Club, though MicroAcquire had the best reputation in our survey.
For a great resource on the different types of sales processes you can go through depending on your revenue levels, check out this interview with Kevin Mcardle.
In our case we had a number of parties that were interested in buying us either for the SaaS platform we had built, our active user base which was seen as a good entrypoint into the data science/machine learning market, or for the expert-curated content we had designed over the course of 2 years.
Some of those interested groups were from our network outreach. Others came from MicroAcquire.
In the end, we felt the strongest resonance with Louie and his team from Towards AI who we met through MicroAcquire.
We were inspired by the global education platform and community they were building, and from our earliest conversations it just made sense where Confetti would fit into that broader picture. We also had strong conviction they could carry through with evolving our platform into its next stages.
From our initial connection on MicroAcquire, we had a handful of conversations with Towards AI diving into specifics of the business, usage numbers, and financial numbers. Then we signed an LOI, after which we did a deeper technical dive into Confetti’s architecture and setup.
After we completed this, our acquirers submitted an official offer agreement where the purchase was structured as an asset/IP transfer of everything we had built in Confetti.
We spent some time deliberating the terms with our legal counsel, who ended up being incredibly helpful in helping us think through nuances of tax implications for the business sale and other gotchas that we would not have known about.
After we all agreed on the terms, we signed the agreement and the acquisition was completed! Well, sorta…
When we first initiated the acquisition process and started learning about what that would entail, one piece of advice we kept seeing was: build your business as if you plan on selling it.
Note: this doesn’t mean you necessarily will (or want to) sell your business but what it suggests is there is a healthy amount of upfront work you can do to make your life easier should acquisition become an option.
This means being very deliberate about your bookkeeping efforts early on including maintaining contracts with vendors or customers, tracking your expenses and revenue, and keeping tally on product usage and statistics.
This is because when an acquisition finally is on the table, all of these questions and concerns will be present in discussion.
So would you rather walk into discussions prepared or desperately crunch numbers an hour before your meetings like a person looking for lost receipts on the day taxes are due?
I can’t emphasize the importance of this point enough.
In our case, we made a few early decisions around technology that made this process a bit harder. For example, using Paypal as our service for conducting sales and platform purchases ended up biting us a little, because Paypal really does not make it easy to track metrics such as monthly recurring revenue, customer lifetime value, etc. in a streamlined way like Stripe does.
Anyway, after we had signed the official purchase agreement we had to conduct a technology transfer of all code, SaaS service accounts, domain names, etc. Overall this ended up being pretty straightforward (yay modern software). When it was finally done, then we could say the acquisition was complete!
From the time we decided to begin the acquisition search to the final asset transfer, the entire process took around 4-5 months. So do be mindful that this isn’t a crank-it-out-in-a-weekend type of project. It took quite a bit of time to go through all the stages.
One caveat about the MicroAcquire service: while we had a phenomenal experience with the platform, there was a moderately low signal to noise ratio. We ended up receiving some 20 requests to connect about our business, and probably only 4-5 were really serious about what we had built.
Expect that a lot of individuals on the platform are just browsing around like people that look at homes on Redfin they never expect to buy.
In some ways that’s just part of the game so just be mindful of that if you see a similar phenomenon. Don’t get discouraged.
When all is said and done, we were incredibly proud of what we had built into Confetti and happy to see it finding a home in a broader ecosystem, helping to realize a noble vision to democratize AI knowledge globally.
After this first foray into bootstrapping a business I walk away with many strong convictions about company building.
But I will leave you with this one: entrepreneurship is a super challenging endeavor, and yet I continue to believe it is one of the most powerful vehicles for human progress we have.
If you are a founder working tirelessly to realize your vision of the world, know that you will always have my heart and I am happy to help in whatever way possible.
Don’t hesitate to reach out.