How dot Learn became Vectorly
Given that this is the first official blog post of Vectorly (or the last of dot Learn?), I apologize for the highly personal nature of this post. At the earliest stages any startup is a story about it’s founders and their efforts to build a new company from scratch, and ours is no different.
This is the story of how dot Learn became Vectorly — about how we pivoted from building a video learning app in West Africa to developing a new video compression technology. I apologize it’s taken us so long to talk about this, but it really hasn’t been until the last few weeks that I can say we’ve completed our pivot. With our pre-seed round of funding behind us, and with a new mandate to develop our video compression technology, I finally feel ready to talk about what has been the most difficult year of my entire life.
dot Learn was born when I met my co-founder Tunde Alawode at MIT, at a class called “Development Ventures” in the fall of 2015. We both had a passion for online education in emerging markets, and during that semester we decided to build a online learning platform for emerging markets, and we decided to initially focus on West Africa (where Tunde is from).
With some initial funding from MIT we spent January 2016 in Ghana doing market research with high school students. It was after one set of student interviews, realizing that bandwidth was such a barrier to watching online video in Ghana, that we came up with the idea for “vector-graphics” based video, which enabled us to make videos that used 100x less data/bandwidth than regular video. It seemed like such a good idea that over the following semester our pitch became “A data-light learning app for West Africa”.
After graduating in 2016(and participating in MIT’s accelerator) I moved to Ghana to begin working on it full time — building the app with 2000+ “Khan Academy” style video lessons for the West African test-prep market and growing to 5,000 students in a year through school-to-school guerrilla marketing.
In 2017, Tunde graduated as well, and we got a bit more funding, so we decided to expand to Nigeria with the goal of growing our app to 100,000 students in a year. We set up an office and expanded our team of teachers to accelerate our content development.
By October 2017 we had a team of amazing teachers and brilliant coders, a real product with genuinely innovative technology, real customers who we really were helping, not to mention a few awards — we were a real startup.
Everything seemed to be going so well, so what happened? Given that that was our last major blog post for dot Learn, you can probably guess, this is where our pivot story starts.
Things were going well actually — or at least, they were going “good enough”. From when Tunde and I began our market research at MIT in January 2016 up until our expansion into Nigeria in 2017, thinks were more or less okay.
We had a few things going for us:
- We had identified a real pain point (studying for exams in West Africa)
- We had an innovative solution (a genuinely new technology that enabled an online learning platform where otherwise bandwidth made it impractical)
- We had an amazing team (incredible co-founders, amazing teachers and brilliant coders)
- We had real organic growth of our platform in Ghana
- We had real user stories of real students we were helping, by providing a cheaper alternative to online textbooks
We had some real challenges though:
- We never really figured out product market fit in Ghana. Honestly, I think it’s because we never properly involved the parents (the economic end buyer). This seems obvious in retrospect — parents are the economic buyers for every successful ed-tech product for the K-12 demographic, and I just don’t know why we didn’t spend time on it. This is why only 2% of students ever ended up subscribing for a course.
- Payment was a challenge. Even for the few students who did want to pay, only about 20% of students we interviewed had the means to digitally pay for something (mobile money)
- The majority of high-school students in Ghana (our target market) lived in boarding schools where phones were banned for the entire year. 80% of students studying for the exam couldn’t use our app because they didn’t have phones.
We eventually did grow to 5K to 10K monthly active users, but our revenue never exceeded $100/month in Ghana. This was fine when we were starting out, but we knew we had to 100x that in 1 to 2 years to even reach profitability.
By summer 2017 (1 year after launching our app in Ghana), we had decided to expand to Nigeria. We figured that students in Nigeria took the same exact exam as in Ghana, but the market was 10x bigger. Most of our team was already in Nigeria, and we really hoped that moving to Nigeria would turn things around and accelerate our growth, so making the move made entire sense.
We were wrong on every front. From September 2017 entering Nigeria, we came across very big problems:
- Nigerian students reacted very differently to our app. Our analytics indicators showed lower engagement across the board for Nigerian students on our app. The market was sufficiently different that we’d need to build a different product to cater to Nigerian students.
- Payment systems were even worse in Nigeria. Unlike in Ghana, where 20% of students had the means to pay digitally, exactly 0% of Nigerian students did.
By October 2017, it was becoming obvious that we needed to do something. We had the following choices ahead of us:
- Go back to Ghana
- Adapt our product for the Nigerian market
- Do something else
- Give up
In retrospect, I think we could have app made our app work. We definitely shouldn’t have expanded to Nigeria without getting to product-market fit in Ghana first, and we definitely shouldn’t have treated Nigeria and Ghana as part of the same market.
But hindsight is 20/20, and in October 2017 we had a decision to make. Needless to say, we didn’t didn’t go back to Ghana, we didn’t adapt our product for the Nigerian market, and we didn’t give up. We decided to do something else.
In October 2017, we were invited to participate in Tech Crunch’s first Battlefield Africa in Nairobi, where we would compete with the continent’s top seed-stage startups for glory and $25,000.
We flew to Nairobi, spending 2 weeks preparing for our pitch with the help of Tech Crunch’s staff. In the days leading up to the pitch event, Tunde and I started talking a lot about the direction of our company, and the problems we were facing. At the time I felt we could make our app work by adapting it to the Nigerian market, while Tunde felt we should consider a B2B approach.
I knew Tunde was right, but I was afraid of even considering a B2B approach because I knew that it might involve very big changes to our company that I didn’t want to consider at the time. I eventually conceded, and at the end of our Tech Crunch pitch, which focused on our app, we mentioned we were open to licensing our technology to other video platforms, to see if there was an opportunity in B2B.
Our Tech Crunch pitch came and went, and after getting back to Nigeria, it’s as if nothing had changed for our company. In the few days after our pitch, we only had about 10 people reach out to us about using our technology, and we kept working on mSchool as usual.
After about a week though, we were mentioned on Tech Crunch’s website itself, and within a month we had over 80 different requests, which made us feel like B2B was indeed the right track to go with.
In early November we announced to the team that we were deprioritizing our app, and we began our first steps into B2B. We started by talking with the other major study apps in Nigeria to provide our videos on their apps in a revenue share model. As an example, we worked with Pass.ng, Nigeria’s largest ed-tech app, to integrate our entire video library into their test-prep study app.
By December our videos were being licensed to 10 platforms, and we were getting our videos into the hands of thousands of students, but this never exceeded mSchool’s own user-base, nor did it translate into significant revenue.
Stressed all around, our company took a 2 -week break for the Holidays, and I was able to fly to Mexico to spend Christmas with my wife’s family.
On January 4th 2018, returning to Nigeria Tunde picked me up from the airport and we began the 2 hour long drive across Lagos to our office.
As usual, we got to talking about the state of the company and the problems we’d been facing — from difficulties in our licensing model to upcoming concerns about fundraising. Just as we were about to arrive home, Tunde asked me a question that completely threw me off guard:
“What would you do if you could start dot Learn again, with a clean slate?”
Consider hearing that question in context. I had been working with Tunde on this company for almost 3 years at this point. It took years of effort to build something that was very special to us. The following video captures the sentiment very well.
We truly had something special going on, that simple question brought up the possibility of all of that — our team, our office, our product, our customers, our story — just going away.
It’s painful to write this blog post one year later, because I know that that’s exactly what happened. I knew then that that’s what was going to happen.
If we started the company again from scratch, this time focused on B2B, we would need to shut down our Nigeria office and let everyone else go, potentially even Tunde.
I didn’t want to think about it back in October, but I started seriously thinking the unthinkable.
From January to February, while Tunde worked to see if we could still make our content licensing strategy work and keep our educational content team alive, I went back to square one on a new strategy — What if we really did start from scratch? What would a B2B company look like?
We interviewed over 50 ed-tech companies in emerging, including many of those who had reached out about licensing our technology, and realized that most already had libraries of video content and just wanted it compressed. At the time, our software couldn’t do this — we could only make data-light videos using our own video creation software.
Taking a companies’ existing videos and converting it to our vector-graphics format would have required incredible amounts of computer vision expertise, and we had for the longest time considered this a non-starter.
By January though, with most other options closed, we decided we might as well go for it and focus our company around developing the technology to vectorize other companies’ video content. Tunde had so much faith that the B2B path is the right way to go that he started working on the early code for the video compression even though he had zero computer vision experience and had a thousand things to do as the Nigerian business was to be wound down. By April, he had built a basic version of a program to identify transitions in slideshow videos and distinguish, in real-time, points corresponding to written lines in Khan Academy videos from the cursor, with ~90% accuracy. These are crucial first steps in vectorizing these kinds of videos.”
Over the next 2 months, we pitched a “video compression” product based on our new vectorization technology (which of course we hadn’t built yet). I had gotten commitments from 15 companies to test our vectorization technology when ready, and we collected sample videos to begin playing around with. I had no idea at the time how we would actually vectorize their content or if it were even possible, so we just told them we were in development and personally just hoped we would figure it out eventually.
In parallel, we had been spending $15k per month to maintain our full content creation staff and Nigeria office. We had never made more than $50/mo on our video licensing model, so by February we made the decision to shut down all Nigerian operations and pool our remaining capital to pursue our vectorization direction.
We had notified the team of the change in direction, and set an end date of March 31st as the official closing date for the office. This was the worst possible timing, as Kemi(one of our core team members) had just gotten married.
Closing the office was more than just about business. Many of us lived in the office — we celebrated thanksgiving together, went on trips together, went to our team member’s weddings.
Within a month, it was all gone. By April 1st, we had gone down from 20 employees to just one.
On April 2nd, I packed all of dot Learn’s possessions in my suitcase and boarded a plane headed for New York, the only employee that dot Learn had left.
I’m going to pause the story here and talk about Dubai. If it seems like an inconvenient time for a pause in the story, well that’s what it actually felt like it real life.
Just before leaving Nigeria (seriously, the week before) I was actually invited to conference called the Global Education and Skills Forum, a yearly conference held in Dubai.
In 2018, they decided to launch a new initiative — the Next Billion Ed-Tech prize, where they invited 40 companies around the world dedicated to making ed-tech solutions that could impact that Next Billion.
Our company was obviously a good fit, and after seeing our Tech Crunch performance in Nairobi, they invited us to Dubai to participate. I actually didn’t want to go at first (given all the stuff happening to our company) but they sponsored the airfare and hotel, so I figured I would give it a shot.
We were among 40 companies selected to pitch over 2 days in front of small panels of judges.
The top 6 companies would be selected to present in front of the entire conference at the closing key-note, 3 of whom would be awarded a $25k award.
I figured it was a long shot. I was scheduled to pitch on the 1st day of the 3 day conference, and though I practiced the hell out of my pitch, but I didn’t really have high hopes. After my pitch, I treated the conference as an opportunity to get to know some other ed-tech companies in the field (some of whom might be potential customers?).
On the afternoon of the 2nd day, when they were due to announce the top 6 finalists, I didn’t even bother going to the announcement ceremony, instead spending the time catching up with another entrepreneur / friend outside the conference hall.
It was only until I got a text-message from a friend congratulating me on making the top 6 that I realized we were actually selected as a finalist.
I spent that night prepping, delivering our pitch over and over again to the hotel bathroom mirror. The next morning we were due to pitch first-thing in the morning to the entire conference of 10,000 people, including education ministers from around the world, as well as heads of state and celebrities like Al Gore.
After an initial intro, the finalists went one by one pitching. When it was my turn, I got up and pitched the hell out of our company
By some miracle, we ended up winning the $25k prize.
I think most people would have been ecstatic about winning — and of course, I’m more than happy and grateful that we did, but more than anything, I just felt surreal
I’m still not sure how we won. It was a long shot. I just knew we had to win. As I got on that stage, it was as if years of work on dot Learn — every pitch, every late night working, every student interview, every team meeting — were on the line, and I was pitching for the survival of the company.
In retrospect, that sentiment was spot on. 9 months later, in December 2018 our bank account ran below $25k. If our company hadn’t won this competition, our company wouldn’t be alive right now.
Returning to real life, I landed in New York on April 3rd with all of dot Learn’s positions in my suitcase, and I crashed with my parents for a few days.
For the next few months, the company was just me. We had enough funding to pay myself a basic salary, but we had no office, no employees, no product, no customers — nothing.
What I did have was a fiancée who lived in Mexico. My fiancée and I had managed a long-distance relationship for 5 years at this point, so I figured now was a better time than any to actually go live together.
After a brief weekend at my parent’s, I packed everything with me and moved in with my fiancée in Querétaro, Mexico.
I found a local co-working studio in the city, and used it as a base to build dot Learn again from scratch.
If you remember, about 1–2 months prior to this point we had talked to a bunch of ed-tech companies who expressed interest in using our “vectorization” service, which we promised would compress their video content tenfold without quality loss. I had no idea how we would accomplish then, and I still had no idea when I got to Mexico in April.
A number of them were now sending follow up emails asking for updates, and I had to make up something about R&D and “still in development”. By April 8th, I made my way to the co-working studio and spent my first day thinking about how we would actually go about even doing this in the first place.
I really had no background in Computer Vision and little experience in Python, but I figured now was as good a time as any to pick it up, so I started taking a few courses on Udacity. I picked it up quite quickly, and within a week I began actually developing the beginnings of our vectorization algorithm.
I identified “Slideshow videos”, like those from Coursera, edX or Udemy, as the easiest kind of content to begin to “vectorize”. Building off Tunde’s code, I started by putting together a simple algorithm that would take in a typical slideshow video, identify which parts were “real-life” content and which parts were “slides”, save the slide portions of the video as images, and reconstruct the video using our special “.lrn” format. Surprisingly, this obtained 20% to 50% file size savings for some videos for very little effort
By May I had begun working on actually vectorizing the slide images (the hardest part). If I could vectorize each slide image, we would essentially have a real MVP for vectorization, and we could begin getting back to customers. I worked furiously, 14 hours a day to get it done, and by the middle of May, I had actually gotten a simple working version of this for a very basic slideshow video!
The problem was: none of our client videos were that simple. As an example: Here’s how our algorithm ‘vectorized’ a real client ‘powerpoint’ video:
Our vectorization actually blurred the text, and handled the gradient at the bottom poorly. When I tried running our algorithm on other videos, I realized we would need months more of work to handle things like fade transitions, cursors on the screen and the myriad of other complications that real clients had in their videos.
At this point — in May 2018, I had a comfortable rhythm going in Mexico. I was finally with my fiancée, my $1.6k/mo salary stretched quite far in Mexico, and I could have kept working for another 2 years at that run-rate.
As comfortable as it was though, I knew that this wasn’t the way to run a scale-able business or achieve the mission we originally set out to achieve, and I knew at some point I would need to find new co-founders and build a new team.
I had no idea where to start. I had been out of MIT for several years, and had lost a lot of contacts on campus. Querétaro wasn’t exactly the best place to build this kind of company. The question then became — where do we build a team?
Being a citizen of no-where, I really did consider the entire world, looking for locations with
- A very large talent pool for high-tech computer vision talent
- Access to funding
- Low costs
- English-speaking talent pool
- Low legal barriers for us to set up and operate our company
I really did consider the entire world — Nigeria, France, China, Argentina, Singapore, Mexico etc.. Some countries didn’t have the talent, some were too expensive, and for most I couldn’t just move and work there without a residence permit.
I ended up considering India and the US — not because I feel a particularly strong affinity towards either country — (I sincerely don’t), but because they tick the right boxes and because I didn’t need visas to go live there. If visas weren’t an issue in China, I would have looked there too.
With locations settled, I put out job postings for computer vision engineers, and started circulating them between friends, Facebook, LinkedIn, Angel List, Co-Founders Lab and any network I could find.
Some networks were more helpful than others: Facebook led nowhere, and Co-Founder’s Lab only got me messages from other would-be entrepreneurs with “the next billion dollar idea” looking for someone to code up their app idea.
LinkedIn and Angel List were far more helpful. Through Angel List I was able to source over 60 okay to good candidates between the US and India. LinkedIn got fewer candidates, but because they were through my networks, I was able to source better candidates (including, it turns out, most of the team members we ended up on-boarding).
I spent the next few weeks interviewing non-stop- from early morning talking to candidates in India, then US candidates during the day, and then Indian candidates again late at night. I talked to a lot of good candidates, but if I was going to work with someone for the next 5 years on this, they had to be excellent, they had to be at the right stage in their career, and they had to be excited about this idea. It felt it was a lot like panning for gold — what you’re looking for is certainly out there, but you have to spend a really long time finding real nuggets.
In June, after having talked to 50 good candidates, I asked Savant (our CTO) if he wanted to join and help set up our office in India, and he said yes.The following week, we brought on Avi, and 2 other engineers and we decided to set up our office in Bangalore. We had had our first team meeting in June — the first one since our office shut down in March, but now with an entirely new team.
Every year Cisco’s CSR team runs a program called the “Global Problem Solvers Challenge”, where they recognize and support social entrepreneurs using technology to solve global problems. In 2017 we entered the competition and were awarded #2nd place, an award which helped us set up our office in Nigeria and grow our app to 100,000 students.
In early May 2018, as we were beginning the search for a new team, I got an e-mail from Cisco’s CSR department asking if we would be able to present our work at Cisco Live, Cisco’s annual conference in Orlando in June. They offered 2 round-trip tickets and lodging in Orlando, so we went for it — bringing Tunde and myself to Orlando in June.
It was really refreshing to see Tunde again after a few months. The conference itself was huge, with 10,000 people from around the world to hear talks about networking, etc… For the 3 days of the conference, we were given a booth and grouped with 2 other social enterprises, to explain to attendees about our company.
As an additional incentive, attendees could vote for their favorite company, which would win $25K. For our company, it had been a difficult few months already, so there was nothing else to do but to put our pitch-face on and pitch the hell out of our company to conference goers. We came early, skipped lunch (well, I ate protient bars) to make sure we had time to pitch conference goers about our company. We pitched because we knew the survival of our company depended on it.
By the end of the 3 days, at the closing keynote, they announced that we had won the $25K award.
I was extremely happy and grateful, but more than anything, just relieved. Given the circumstances of our company though, the only thing I could think about was “I hope this helps us make it through our pivot”.
Literally right after the keynote, I had to take a taxi to the airport, check in hand, to fly home. I spent the entire flight looking out the plane window. Things were tough, but we would make it through.
That flight wasn’t headed back to Mexico, but rather to New Jersey because, as if my life weren’t eventful enough at this point, I was about to get married. Twice.
I had proposed to my fiancée back in December 2016, and we had already gotten legally married in a civil wedding in August 2017. Now, in 2018 we had been planning 2 weddings — one Hindu wedding in New Jersey on June 24th, and one Catholic wedding in Mexico on July 7th.
Arriving in New Jersey a week before my wedding, I worked part-time, interviewing business development candidates and meeting with a few investors in New York while also organizing last-minute logistics for the wedding.
After a hectic week of preparation, my wife and her family arrived in New Jersey, and before long the big day arrived.
After the wedding, and 2 days of showing Nadia’s family around New York, we all headed back to Mexico to prepare for the next wedding.
It was in the week between our weddings, literally as I was in the hotel putting up decorations for the Catholic wedding in Mexico, that I was on the phone with different candidates for our business development position.
After discussing with Tunde, we extended an offer to Michelle, who accepted and joined our company full time.
On July 7th, we finally got married for the final time.
In the days after our Mexican wedding, we had planned a short honeymoon in Cancún since it was close to Querétaro and cheap. When we got to the airport to check-in for our flight to Cancún, Nadia tried to check-in and she realized there was no flights to Cancún that morning. Whoops. In the middle of all this, I planned a secret Honeymoon in Dubai.
The month of the US wedding, our Mexican wedding, and our honeymoon was by far the high point, of what was otherwise turning out to be a very difficult year.
I’m extremely happy and grateful to have celebrated two authentic weddings with the people we love, and a full 2-week vacation with the woman I love.
It was a good decision, because coming back to real-life, I would enter what would the the absolute worst phase of the entire pivot.
Fundraising is brutal, and even more-so when you don’t know what you’re getting into.
Until this point, dot Learn had lived entirely on grants from MIT, Cisco and others. This had served well for our social enterprise model where the risk was subsidized and impact was part of the goal.
As we pivoted to a deep-tech model though, we knew that if we wanted to build technology and get it into the hands of a billion people, we would need to raise funds from traditional Venture Capital sources.
For some reason, I imagined that at this point we were in a position to raise money from VCs. I thought that all you needed to raise an initial “seed round” was an idea, a team, an MVP (a simple but working product) and some early adopters. In retrospect, this is just plain false, and the bar is substantially higher.
Either way, given that “seed rounds” are on the order of $1m or $2m, I thought it would be fairly easy to raise just $750k in 2–3 months, and then move to India to hire 10 engineers to build out our vectorization technology. I had even bought plane tickets to move to India, and had a list of apartments we wanted to move into. Reading this now, I can’t help but cringe at how naïve I was.
After returning to New York on July 25th, I started working from my parent’s house while my wife went back to Mexico to finish her last few months at her job. The goal for the next 2–3 months was clear: raise our seed round.
We had an idea and we had a team, but we still had 2 clear milestones to achieve before we could raise funds: a product and some early traction, so making this happen became priority #1. Luckily, we had already started this process in June when I on-boarded our team.
If you remember back in May, I had actually already built a demo which could take in a simple powerpoint slide video, and spit out a “vectorized video”. Between June and August, our new team had cleaned this up and put everything together into a single script which could work on a good range of powerpoint style videos (and not just a single demo video).
We now worked on getting back to the 150 companies who had originally reached out to us about our technology. At this point, all we had was an algorithm written in Python to “vectorize videos”, and since many of those companies had given us sample videos to play with, we just ran their videos through our algorithms, uploaded them to the web, and got back to each company with samples of their compressed videos.
For the companies interested, we then followed up with pilots where they gave us their entire library of videos, and we manually ran our vectorization algorithms on all of them. We would then make sure the results were good before uploading them and sharing spreadsheets with links to code which the clients could use to embed their videos in their website. Within a few weeks we already had a few companies using us, so it seemed like it was a good enough milestone to begin fundraising.
For the first half of 2018, we had gotten in contact with a number of VCs. I won’t give names, but there was a mix of African and Africa-focused VCs that we had gotten to know while working in Nigeria, as well as 2 regular California VCs, a few social-impact VCs and some random global VCs without much of a specific geographic or industry focus.
By mid August I started getting back to all of them, giving them updates and letting them know that we were launching our seed round. This set off a flurry of emails and calls, and over the next 4 weeks or so, I went through a marathon of investor meetings and it was just absolutely brutal.
Some were genuinely interested, some were lukewarm and some literally just scoffed at me. I understand that many entrepreneurs go through hundreds of “no(s)” before they get to a “yes”, but 20 rejections was pretty disheartening. Nevertheless, I tried to be rational and began putting the dots together.
We were way to early to raise a seed round. As one honest investor brutally put it “Look — you guys don’t have revenue, you’ve known each other for less than a year — what are you guys doing?”
Luckily, many of my MIT classmates had since gone into Venture Capital, so I was able to have a few honest discussion with impartial investors about the state of our company, which helped corroborate what I was seeing first-hand in our investor meetings:
- We were too early
- We needed actual sustainable, growing revenue before we could raise a seed round
- It would take way longer than 2 months to actually raise a seed round
By late September, things got really low. I remember just sitting in my bed one morning crying, and not wanting to get up and face the world.
Eventually clearer thoughts prevailed, and I decided to take a time out on fundraising and re-evaluate our strategy.
Our original goal of raising a seed round was to just build our vectorization technology fully, but it became clear that we couldn’t just raise $1m to do that. Talking with my investor friends though, I learned that companies now raise “pre-seed” rounds of $200K to $500K to get them to a point where they can raise a seed round, and that our traction to date was in the right ballpark for a pre-seed round.
We therefore decided to scale back our ambitions to raising just $150K. This should have given us about a year of time with our current team to make the milestones needed to close a seed round. We already had a few contacts / potential angel investors who were interested in joining our round, so we focused on our closest contacts, and by October we got to to commit an initial $30k of our $150k pre-seed round.
Everything dramatically changed in end of October / early November. On October 26th, our patent (which we first applied for in July 2017) was finally issued by the USPTO. From here-on, we actually had a bona-fide patent, and our technology was no longer “patent-pending”.
Next, in early November, we were invited to a pitch competition run by TiE New York, a group of South-Asian diaspora investors and entrepreneurs. It was just one of those pitch opportunities we heard about from a mailing list and applied to. We submitted a pitch deck, and we were accepted to pitch.
Significantly, this was the first time we used our new brand name “Vectorly” in our pitches (we hadn’t even set up a website for vectorly at this point, it was just an idea).
Unlike most “startup pitch competitions”, this one had a good deal of actual, active investors in the room, and some well-known investors in the judging panel. Unexpectedly, our company won first place in the competition, and all of the sudden we got a huge flood of interest from various angels and VCs at the event.
Over the next few weeks we were talked to 20 to 30 investors who were actively interested. While this was great, what I didn’t realize was how long it would take, or how emotionally draining it would be to convert this interest into actual investment.
The actual mechanics of doing this isn’t hard. Investors asked straightforward, reasonable questions to understand our business, such as our team and our market, and ask for key documents such as our financials, etc… While some investors can decide in a few days, many want to go through 4 to 5 meetings and spend weeks before making a decision.
The worst part is the complete uncertainty, as you see your own company’s bank account draining towards zero, and knowing that your current conversations could materialize into anywhere between 0 and $1M in funding (even if you only need $150K).
I can only assume that good fundraisers are great at creating an idea of scarcity and collapsing $1M of interest into $150K in actual investment. I am not a good fundraiser, and this process dragged on for 4 months — from November to February, during which time our company sat in funding limbo.
I felt powerless to control the process, but there really wasn’t anything I could do but keep following up with investors while working on our business as best as we could at the same time.
After our initial trials with clients in October, the general feedback was essentially “We loved what you pitched us, but your product has issues”. By “issues”, I specifically mean:
- Our first version of vector-based videos weren’t streamable, so when you watched one, it would load the entire video (20s to several minutes) before you could even begin watching it
- We asked clients to give us videos over google drive links, and we returned them spreadsheets with video embed links
- We had issues with the audio quality
- Our video player had different random bugs on literally every platform
Rather than being disheartened, I was actually happy with this feedback. The biggest fear in my mind since we began our pivot in February was, “are we building something people actually want to use? “
After continuous client calls, I think we’ve pretty solidly identified a real problem — bandwidth for videos in emerging markets, and a solution that seemed genuinely attractive to clients.
At this point, the main issue was fixing the inevitable issues and building a product that delivers on the pitch (lossless video compression) that we had made to clients.
Over the next few months we worked, one by one, on each of these issues. The most grueling part was bug fixing, where we had issues such as our video player just wouldn’t play on iPad 4 on Safari (even if it played just fine on iPad 5 and iPad 3). It tooks weeks of repeatedly testing our videos on every device & browser combination under the sun, before we could consider our service “production-ready”.
Either way, we got to something that was generally used by our clients, and something they were happy with, and we even got several of them to commit to paying for our compression service once we worked through all of these issues.
At the same time, we kept getting organic press, as well as inbound requests from larger potential clients and partners like Samsung, about our technology. The only problem was, we just weren’t ready for any of it, which put the onus on us to complete our fundraising so we could move faster and actually build something that would work at scale.
In February, fundraising took a dramatic turn, and by the end of the month we had reached not only $150K, but over $250K in commitments for our pre-seed round. Also, randomly, we were accepted to an accelerator program that we never actually applied for, and we got to a point where we had to start turning down investors as we wanted to avoid too much dilution at such an early stage.
By March we had signed documents for $250K in investment, even if the actual wire transfer of money to our account took a little longer. Fundraising is not over until money is in the bank, but I could begin to feel the end in sight.
Our original pitches from 2016 have always included our mission of making online video learning more accessible for 50 million students around the world by 2022.Fundraising had dragged on so long at this point (8 months) that my wife and I just abandoned our plans to move to India, and started my wife’s greencard process to move to the US (it’s own set of headaches). With signed investment docs though, I finally felt comfortable enough to make good on our initial intention of going to India, if only for a few weeks. I booked my flights for Bangalore and set of for India for the first time since 2016.
Arriving in Bangalore was surreal. I obviously know we had been working as a team remotely for close to a year at this point, but the day I landed in Bangalore and headed to our office (in WeWork) I was surprised to see our company’s logo on one of the offices.
I was finally able to sit and work, side by side, with the rest of our new team in the same office. The pace of work just accelerated enormously, and it was incredibly valuable to have the time for real team bonding.
I spent the better part of March in Bangalore and I was finally able to really gel with our team in person for the first time. For the first time in a while, I started genuinely feeling happy and optimistic about how our company was going.
It was only today that the last of the wire-transfers for our pre-seed round went through. A year after I left Lagos, it’s only now that I can actually say that we’ve pivoted our company and come out the other side.
We are no longer dot Learn. We have a new team, a new product, a new vision, and the funds to go make it a reality.
A year on, I don’t view our vision as any less meaningful. By focusing on developing pure video compression technology, we may lack the direct impact on individual students that we had when working on our own app, but we do now realistically have a shot at impacting billions of people.
Our original pitches from 2016 have always included our mission of making online video learning more accessible for 50 million students around the world by 2022.
It’s precisely because of that mission that it made sense to focus on video compression technology. It’s precisely because I care about this so much, that I didn’t want that slogan to just be just another idealistic college startup pitch. With Vectorly, there’s a real and honest shot of making that happen.
If my goal was to make money, my god there would have been way easier ways of doing that, most of which wouldn’t have entailed moving to Africa for 2 years, going without salary for most of that time, spending late nights coding to fix bugs, early mornings preparing for pitches, starting my married life by hopping from one AirBnB to another for 8 months, firing my best friends and straining my personal relationships.
This company, and this mission has been my entire life for the last few years, and I really do intend to make online video more accessible for 50 million students by 2022.
It’s just that now I think there’s a better strategy for going about it. If we’re able to develop this technology, I want us to get it into the hands of organizations like Coursera, edX, Khan Academy and even YouTube itself. I know it’s possible, and if we do achieve it, we’ll already be reaching a billion people.
I’m sorry it’s taken so long to write about this. I didn’t want to talk about any of this until our fundraising was over, in order to avoid risking our funding round. It’s also just been a very difficult year, and perhaps the hardest in my entire life.
But now that it’s over, I can finally talk about what the last year has been like, and it feels cathartic. I’m genuinely happy and optimistic, that we can plan ahead for more than a few months.
This week, we begin the NVP Labs accelerator program. Tomorrow, I’m going to head to the NVP office, and without a doubt we have a long road ahead of us. But I'm happy now, because I know we have a real shot of making dot Learn/Vectorly's vision a reality.
Thanks again for all the support
-Sam
