9 years ago I was the first engineer on a b2b product that eventually sold for $1B+ in 2021. Unfortunately, I can not share the name publicly. But this is the story of the fastest-growing product I ever worked on.
Started in February 2013, it grew to be a unicorn in less than 7 years. This company is now #1 in their niche. This success story is about luck, hard work, simple technologies & great people.
Below I write about important moments that helped the company to become a unicorn.
Never give up
The product started unlike a typical startup. There were two companies who wanted to automate their business and paid a vendor to implement it. The vendor found an outsourcing company to create the product. After 5 months of work, the product was very buggy and unusable to the point that the client company wanted to stop development.
At the time, I had been working with the vendor on couple of other projects for about a year. We already had a few successful launches together and built initial trust in each other. I received this email back on March 5th, 2013. The decision was to re-build the product in one month and present it to the client.
Building products is hard and sometimes things don’t work, but you should never give up. So we didn’t.
7 days to UI prototype
I was the only engineer on the team. The first version of the product was not usable. It was crashing every 5 minutes, and navigation was also messy. Every time you click on the menu item you see a new menu and you can not find a path back.
I got the go-ahead — “Feel free to change the design to make it prettier throughout the site” — to change whatever I wanted.
So I had a white sheet of paper — this is always the most exciting moment. You have nothing and you can create anything. I spent a few days to understand the business, review existing applications and read through the code.
Since 30 days was not enough to build the full product, I decided to resolve all the UX issues and create a new UI prototype. So I just started to work on it piece by piece: signup, navigation, first page, second page. After a few pages, we would meet with the client and discuss the progress, make small adjustments and move forward. We worked hard and moved fast.
It took us 7 days to build and deploy the full product prototype. We showed it to the two companies who wanted this product. They were stunned. The prototype worked fast, never failed, had clear navigation and event test pages consisted of real data. So it felt like a real product and they wanted to start using it immediately.
In 2013, most websites rendered on the server side. We built the prototype as a single-page application, based on the backbone framework, which was among the first to support single page applications. That helped us to create a ‘wow effect’ because the app was at least 2 times faster.
$12 for design
The design and pages we built were used by thousands of users for 5+ years, almost until the point when the product became a unicorn.
The cost of the design was just $12. We simply bought a Twitter Bootstrap theme.
It’s great to have a beautiful design and perfect user experience, but for many early-stage b2b applications you can just use something simple and focus on solving problems. The design will come later. This product was eventually redesigned, but only 6 years after the launch.
SaaS product in 2 days
After launching the UI prototype quickly we immediately had the go-ahead from the first two clients to finish implementation. It took us around two months to actually build the full product after the prototype. We added one more engineer to speed up things.
The development went well. We demoed progress every week and the end clients remained excited. So at that time, the product was only for two clients. Initially, every client wanted slight customizations and the first version was developed in separate branches to achieve that. This time we decided to keep and maintain only one product, but try to find a way to fulfill the needs of both clients.
Two months into building the product we were about to finish everything we wanted. One day during a demo meeting the client says: “Andrew, let’s make it SaaS”. I didn’t quite know what it meant, but for me, that meant that we just going to put what we built into the hands of more people. Exciting, let’s do it!
The engineering team was focused on building multi-tenancy (to make sure data separation between companies) and subscriptions. The client, who became the founder right at this moment, was focused on creating the landing site. We worked long hours, but together we managed to turn the product into SaaS in just 2 days.
We officially launched on May 26th.
First customers
We had two customers from day 1. That helped us a lot, as they were very supportive and provided us with a lot of great feedback. We felt they were part of the team, who had a lot of insight into the business domain.
The founder was spending most of his time finding more first customers — he was just manually reaching out to them. He discovered 5 b2b customers in the 40 days and around 30 in the next 3 months.
That was enough for us to understand what they needed and create a better product plan.
Be a support hero
This is my favourite part. The founder is very charming. He built great connections with many of our clients and was spending 50% or even more of his time actually serving them.
The founder did his best to reply to any issues within 5 minutes of each request. He literally had intercom opened all the time. This created the impression that a big team is working behind the product. While in fact there was just the founder and three engineers.
We deployed fixes and small features the same or the next day after each request. I think this approach alone brought a lot of loyal customers in the first 6 months.
Talking to your users is way more important than the features and quality of your product, especially during the first year. Here is the quote from the founder of the product:
“In an ideal world, I believe we ALL will be involved in customer support, and I would actually rotate support among us all. I think it makes for better software development.”
Raising the seed round
Hard to believe, but in 7 months we managed to acquire 100 b2b customers, who were paying money for the product. Up until this point, the project was mostly built using customer money and 3 months of debt for the development services. Luckily, I had (and still have) a really great relationship and a lot of trust in the founder to help him to that level.
This might not work well for everyone, but for the first 6-7 months we followed our competitors a lot. We never looked into their product, but he had a list of features they had. Over the course of the first 7 months, we built the most important features competitors had and were ready to compete with them.
In terms of development speed, we were 3x faster. We moved fast and broke things, all the time. But because the founder worked as a support hero and we fixed things quickly — the product was getting great reviews.
With 100 customers and current traction it was really easy to raise small seed round. In fact, this round was enough to keep company growing and allowed us to expand a team with QA and stabilise the product.
Post where your competitors post
This is an easy, yet very powerful approach we used to create a steady stream of incoming leads. Go to every website where your competitors post something and post next to them. Customers always compare and look for better options, so we made it easy for them to find us.
We didn’t have a marketing specialist, blog or anything like that at the time. But people were able to discover us.
Key learnings
In 14 months, the product acquired 500 paying companies. That was October 2014. We didn’t fully realize how big it would grow. In 6 years, the product became a unicorn.
- Be quick to launch and build the product for the first 100 paying customers to find your product market fit.
- Be a support hero, talk to your customers and solve their problems quickly.
- Post where your competitors post.
- Use your competitors to better understand the business domain.