Before I start I want to thank Sahil for being my inspiration for writing this blog. I would’ve never had the guts to let it all out there if it wasn’t for him. Sometimes it just takes one person, one thought, one book to make your life spin in a different direction, and to me, telling our story in an honest and open way feels like a better direction. Thanks Sahil.
Datagran was founded in 2017 in a small country in South America called Colombia by a Turkish guy as our head of AI and two Colombians, our CTO and me. The three of us each came from modest backgrounds where our main goal was not to become billionaires, but instead dreamt of the possibility of one day being able to live with our families in the United States of America. For some Americans this may sound odd, but for many of us is a childhood objective. Pretty soon our dreams started to become a reality.
After applying to YC, Techstars and others, a newly created acceleration program in New York called Quake capital accepted us for their first batch out of New York City. Their general partner and founder, Glenn Argenbright, a sharp ex operator and his team welcomed us, and by August of 2017 our lives were completely changed. We were living in the greatest city in the world.
During the Quake summer batch, Datagran was always the best Startup. Everyone talked about our rapid growth and everyone thought it was just a matter of time before we became a unicorn. We closed deals with Subway, Discovery channel and many more. Soon we had angel investors in line wanting to participate in our pre seed round and we almost immediately before Demo Day, closed a 1.5 million dollars round while most of our batch colleagues didn’t even have a term sheet yet. We couldn’t believe it and we felt blessed.
We continued our growth and by 2018 we raised our Pre-Seed round of 2.5 million dollars from top angel investors and we felt unstoppable. With money in the bank and revenue flowing we started spending rapidly to accelerate our growth and scaled the team to about 60 people prioritizing revenue with the next objective in mind to raise from top VCs. We welcomed amazing people to the company like for example a top executive from Microsoft who believed in our mission to make Machine Learning easier and more accessible. He left everything, a big paycheck and his stability to build along with us.
In 2019 we felt invincible and started knocking on doors of top VCs. We were approaching $300k in Revenue which back then was the sweet spot to raise a nice A round. Over and over we got rejected. After doing due diligence and reviewing our financials they all concluded the same, we were heavily concentrated on a few clients like Google, Subway, among others. Google alone represented around $110k of our total revenue. We met with more than a hundred VCs. Probably I know them all at this point. All of them were impressed with our growth and revenue, but disappointed when digging deeper into our financials.
In any case I thought that once we hit the $300k mark raising would be a breeze, it would be unequivocal. I was wrong. We hit the $300k mark or close to it, it was now the end of 2019 and we still had zero new funding in the bank. And then, without any notice, the pandemic hit. It was the beginning of a nightmare but the start of our salvation.
Our revenue went from almost $300k to $90k in less than 3 months. At that point we were targeting our product to marketing people who saw a dramatic cut in their marketing budgets. Most of our revenue came from a Google division targeting SMB’s, small businesses suffered the worst part. Soon we were helping our clients weather the storm and giving our product for free to several of them.
At that point we realized that our business was not sustainable. A few months back we were bouncing back and forth with an idea based on all the feedback we’d gotten from the more than 3,000 clients we had at that point and from VCs who turned out to hate the Martech industry. We decided to pivot. We needed to make a business that was structurally better, less dependent on a few clients, but not only that, a business that could scale faster (some thought our growth was scaling already) and that could make a real impact. Without no new capital and only about 1 year of runway, making a pivot would mean reducing the team, letting go of our legacy clients and changing our team completely. Our current team was specialized in marketing and we wanted to focus now on CTO’s, Data Scientists, Data Analysts and Data Engineers.
We laid off 70% of our workforce, asked the remaining employees to lower their salaries and had to let go of friends that we truly cherished. I cried many times and felt helpless. We also had to tell our Microsoft hire and friend that we couldn’t pay for his services anymore. That destroyed me. I cried like a baby when I had to tell him on a zoom call. What a hell of a fighter he is. Even in the hardest of times he always had something positive to say that lifted my spirit. Nobody would guess that he had a girlfriend with Cancer and and a young boy with a discapacity. I celebrated like never before when I heard after a few months from him that he got an offer from a FAANG to be a product manager. Well deserved.
Back to Datagran, we managed to increase our runway at that point to a year and a half with our goal in mind to completely change the product, a product that will allow companies with one or two data scientists to be able to deploy ML models blazingly fast, without the need of having to scale machines, build APIs or schedulers. We liked calling it the zapier.com for Machine Learning. It would be a game changer having in mind that this is one of the biggest problems in Data Science to date according to Gartner and to every single Data Scientist we spoke to.
It was June 2020 and we set ourselves a target. To have an MVP in just 3 months! Sounds crazy even now taking in mind this would be a huge product, but we did it building on top of some of the features we had in the previous product. In 3 months we were testing the product with a few clients, and bit by bit we got more and more confident that this was the right path. Internally, some thought it was stupid to let go of legacy clients that were giving us $50k in revenue. But for most of us, it made perfect sense. We wanted a product that was truly SaaS, that people could use by themselves and a product that was purchased by many and not by a few, and for that, we needed focus, we couldn’t afford distractions caused by consultancy work. So we started with a blank slate in terms of revenue and looked forward. More and more early employees and co-founders left, a process that drained our energy but gave us a huge financial breath. Probably today one of the most heroic acts on their part. Some of them knew we needed it.
As with any new product or service things don’t go as smoothly as you would wish. Some months we would bring a ton of revenue and others we wouldn't bring any. With the added nuance that we didn’t have any extra money to spend. All of our cash was in the bank trying to preserve one year of runway with the expectation that by December of 2021 we would be able to raise funds.
So many great people along the way have made our path easier. Heroes like Suhail, founder of Mighty and Austen Allred from Lambda School noticed our product and made us winners of one of Suhail's twitter activities called shipping Fridays, bringing a ton of attention to our product. Not many people realize what Suhail does for the startup community and the amount of lives he touches without expecting anything back. There are several others that have helped and continue to help us without asking for anything in return like Amjed from Repl.it. Real heroes that do their work silently and that are rewarded by the cosmos for their work in life.
.jpg)
With renewed confidence we went on to apply for a 7th time to the prestigious YCombinator acceleration program. With a new product, a new vision and believers like Suhail and Austen we thought that everything was on our side. Even Stripe with their amazing Stripe Atlas program helped us craft our application in a way that we wouldn’t be able to do by ourselves. They helped us coin the concept “The Zapier for ML”. We will be forever grateful! And indeed this time something happened that we’ve never believed possible. We were called for an interview. With our new product, new clients and emotions overflowing my Turkish co-founder (at that point our CTO had left the company) and I attended the interview with a glass of wine on our desks. Although we didn’t cross the finish line yet it hell felt like it! The wine helped me not only to remind me this was one of my greatest achievements, it also helped me relax to confront one of the most defining moments in my life, or at least I thought it was. Of course the 10 minute interview went by incredibly fast and we felt amazing. By the end of it we knew we would be accepted. Then, several days went by and we got an email from one of YC Partners. This partner knew us well. She reviewed each one of our applications since our first one. In that email she mentioned that before they made their decision they wanted a full due diligence on Datagran. A full diligence we said? That’s strange. Well, we didn’t want to put too much thought into it, with the pandemic things had changed so we immediately sent contracts, financials and more.
Unfortunately one of the contracts from one of our clients was a contract that specified that we were hired as an Advertising Agency. The YC Partner was suspicious and asked why. We said the truth, the client was the marketing unit within a very important brand in the US and their contracts were standard, they would not change the objective of the contract even though we pledged our client to do it. Our client said that doing so could take years in the organization and that it would put at risk the contract itself. We didn’t want to push them anymore. YC didn’t believe us. Not only that, we specified in the application that AbInbev was a client, which it was true since we got accepted into their accelerator program and after a successful proof of concept they committed to hire us as a provider, but we didn’t have any revenue from them yet. YC sent us an email a few days later letting us know that after reviewing our documents there were discrepancies between our application and the due diligence and that for that reason we were not accepted. We were destroyed.
2020 went by and our revenue kept growing. Very far from the previous $300k in MRR though, but we felt confident and people we really admired gave us the motivation to keep going.
2021 arrived and everything on the pandemic’s side started to look brighter. Not only that, VC money started flowing to Startups with an added catch, data startups where extremely hot. We thought that if we could maintain our growth we would easily raise money. Many data startups were raising without even having a product nor revenue. By June 2021 our revenue was now $11k MRR and we were projecting that we would peak at about $20k for the first quarter or 2022.
We were ready to approach VCs again, although let me tell you, it was the scariest moment of my life. I didn’t want to feel like a total failure again. To my surprise there was surely an appetite. We quickly met with all the major VCs. All interested in our product. About the same time we received crushing news. We learned that YC had accepted to their 2021 batch a company called getbasis.com that applied with exactly our same claim, the same claim that the Stripe Atlas team helped us craft (“The Zapier for ML). What were the odds? Basis was founded in 2021 and had absolutely zero revenue. Thankfully, a few months back we had decided to patent our idea. We thought it was worth it to spend some of the money we had in the bank to protect what seemed the only thing we could hold on to at that point. Basis went on to raise more than 2.5 million dollars from top VCs and Angels in the Valley after Demo Day.
As always, we moved on. We met with Kleiner Perkins, Andreesen Horowitz, Softbank and many more. Most of them claimed on our calls that the idea was great and most started digging in. With some we made it to the final stages. What is hard about our idea is that even technical people have a hard time understanding it because it is novel. It is a problem that people in Data Science are starting to face and hence VCs still don’t understand it 100%. All the technical advisors that VCs invited to evaluate our product fell in love with the idea. Unfortunately some claimed that sooner or later Google would copy it, others that they didn’t understand the space very well and others just wanted to see more revenue from us. Why? Well because probably we need to prove something a bit more than the others. A little reminder, Latin and Black founders received 2.5% of all VC money since 2015 according to Crunchbase (Source: http://about.crunchbase.com/wp-content/uploads/2020/10/2020_crunchbase_diversity_report.pdf). Latino and black communities in the US account for 31.9% of the total population according to the 2019 census.
We didn’t understand how companies that don’t even have a product or any revenue at all raise millions of dollars and extremely high valuations.
By the end of the process there was light at the end of the tunnel, a Corporate venture in Latin America committed to invest 1 million dollars in our company. All we needed. I was joyful and celebrated with my wife and 6 year old. 2 weeks after I was able to relax again. Then we got an email. They would not be able to invest due to some complexities regarding bonds they would soon issue. Another disastrous set back.
We decided to stop raising. Again. And we failed, again. But, this will sound a cliche, we are richer than ever. Why? All the advisors from the VCs that did due diligence stayed in touch, offered support and that boosted our confidence even more.
After all of these setbacks and process much has been learned so we started planning our next move. We needed to focus on our vision, execute, stay lean and achieve profitability fast to be able to control our future. Thanks to our exponential growth in the past 2 quarters (Q4 2021 and Q1 2022. See revenue chart above) and with clients like Starbucks, Subway, Discovery channel we finally got the opportunity to receive funding from Republic a Crowdfunding solution, Compusoluciones a Corporate Venture and a C-Level executive from Uber. The main stream of funding will come from Republic which will allow us to raise from our community and from the people that believe that anyone and anything is possible. This funding will extend our runway to about A year and a half, time where we will become profitable and the rest will be history.
No matter the hardships and the ups and downs, anything is possible, just keep fighting.
.jpg)
