Shortly: the AI-powered writing assistant
How one founder used OpenAI’s famed GPT-3 to build the best writing assistant on the market
👋 Welcome to The Exit Game. I cover the stories of amazing founders to demystify startup acquisitions for founders, investors and operators alike.
Subscribe to keep updated with future posts
Estimated reading time: 13 minutes
It would be impressive if I told you a solo founder sold their first startup whilst running it as a one-man band. It would be even more impressive if I told you he worked on this part-time whilst going through medical school. At this point you would think I was pulling your leg, right? Now, let’s say this startup was deeply AI-focused, and the founder had no prior technical background but learnt Machine Learning on the job.
Meet my friend Qasim Munye.
Early projects
Qasim is a curious guy.
At fourteen, he realised you could search for voucher codes for online stores. He had a thought: "I wonder if I could sell these?" Later that week he began selling discount codes on eBay for £1 each. It wasn’t much but to a fourteen-year-old, this was BIG money - and the start of his journey.
At seventeen, he expanded his eBay business into iPhone cases. When a customer bought a case, he would buy the same case from a lesser-known seller for cheaper. He used his higher seller ranking to out-sell his competition, and drop-shipped their products to his customers.
By the time he hit 60 orders a day, it was tiring to copy his orders over manually. He downloaded a bot to run the entire order-filling process on autopilot. Each night, he would turn on his Playstation, and start the bot, playing games whilst supervising his sales robot. No virtual assistants, no part-time hires, he found the quickest and cheapest solution for the job. The income from this would later pay off his first year’s medical school fees.
At nineteen, there was a change in the atmosphere - apps were everywhere. Big companies like Google and Facebook had begun massive pushes to teach their developers app development. Mobile developers were rare and they could charge through the roof.
Apps were the rage, and Qasim wanted in on it. He opened Android Studio to get started and hit his first stumbling block - he didn't know how to code. A few YouTube searches started him down the right path.
He decided to make a game: Beat 30. It tested the speed and accuracy of your fingers. Like most founders who launch a product he realised it had hardly any downloads. Qasim's guerilla marketing tactics kicked in. He wrote hilarious reviews about the game and posted them on different subreddits. One post went viral and his downloads skyrocketed to 10,000.
His next project was opportune. At the height of the outrage over Trump’s election results, he made a website: HowLongUntilTrumpLeaves.com. A countdown timer to the end of the presidency. He discovered Twitter and Facebook ordered post responses in chronological order at the time. So, he made a bot that would post the website as the first response whenever a popular political commentator posted. The website went viral and was picked up by media outlets. He sold it at the peak of its fame and paid off part of his tuition fees with it.
Shortly: the short stories app
During his second year of medical school, he discovered the subreddit Writing Prompts and began reading stories written by the 10 million-strong community. To make reading stories easier, he made an app that let you read them based on how much reading time you had. He kept the name simple: Shortly.
After honing his homegrown viral marketing tactics on his previous projects he knew how to get his products in front of people. Users were coming in droves. This was the most successful project of his so far and he was unsure what the next steps should be.
He entered Shortly into the Kings20 accelerator (run by his university), thinking it would give some structure to his thinking and process. Having a dedicated space to work from and surrounding himself with other founders made focusing on the app easy.
Within two years of running Shortly, his app:
reached 50,000 downloads
had 1 million stories read
was featured as App of the day in the Apple App Store
was in the App Store’s top 100 storytelling apps
Shortly: your AI writing partner
Shortly, being a side project, worked out great for him. It didn’t need too much time, which allowed him to spend time on his other interests. On top of reading short stories, Qasim had been devouring startup-related content. Paul Graham’s essays, startup podcasts, Y-Combinator videos - you name it, Qasim was studying it.
One of the interests which uniquely stood out was Machine Learning. Particularly, getting computers to write language naturally - as humans do. Qasim, still a medical student, would go through Machine Learning courses on his morning commutes to hospital placements.
The app was growing - not explosively, but reliably. It was a great place to read stories, but what if you could write stories? The first big feature after a while, allowed users to write and submit stories for others to read. A small community was slowly forming around the app, passionate about both reading and writing stories.
Shortly had officially been a startup for a year.
In June 2020, OpenAI launched its revolutionary machine learning model GPT-3. If used correctly, it could automatically produce amazing written English - better than most humans. It was in beta and access was limited. Qasim had already read the research papers for its predecessors as well as other similar projects. He had to have it.
Shortly’s application was accepted and Qasim wasted no time in integrating it. Shortly already had a feature that allowed users to write and submit their own stories. Using GPT-3 he introduced a new feature: Write for me. If a user pressed this button, the app would look at what the user had previously written and write the next paragraph or page of their story for them. It truly worked like magic, and it was blowing his users’ minds.
He made Reddit posts about it, and his site exploded with traffic. This time it was more than ever before, driving 100,000 curious people to his product. The hype from the post didn’t last but it made one thing clear. Writing short stories quickly became his main use case. Writer’s block was a thing of the past.
It was at this point that he would keep going back to Paul Graham’s essays for advice. He had this product that was taking off like a rocket and he had to make sure it went in the right direction. He would speak to users constantly and iterate quickly. He focused the whole product on the core problem that was being solved. He was so deeply immersed in the AI research and GPT-3s inner workings that he found it easy to create new AI-powered features for Shortly.
The pitch was simple: imagine whilst you are writing an article you have a really smart friend sitting by your side whose shoulder you could tap when you got stuck. Instead of being a story writer, Shortly became a personal writing assistant.
Shortly had reached a point that few successful companies ever reach. He had product-market fit. A fast growth speed for a young startup is 10% weekly growth. Shortly was averaging 20% growth every week - his revenue was doubling every month. Eight months after charging money for Shortly, it reached 3000 businesses paying monthly subscriptions. This degree of progression was almost mythical.
The product was so good it was picked up by media outlets like Vanity Fair, Vice and popular YouTubers. People couldn’t get enough of it and influencers would post about it on TikTok. Qasim capitalised on this by doubling down on influencer marketing.
The unexpected message on Twitter
Around Spring 2021, Qasim received a message from Dave Rogenmoser, the CEO of one of his direct competitors, Jasper (then conversion.ai).
Since GPT-3 came out many people had the same idea as Qasim. So, there was a lot of activity in the market for AI writing assistants. However, it’s not about the idea, it’s all about execution. There were very few players who had built great products, and Shortly was one of the best. The big players were scrambling to win in this space. Copy.ai, one of Shortly’s direct competitors, had raised investment from Sequoia, one of the major VCs.
Qasim didn’t know this at the time, but Jasper had also raised a large round of investment, and they were planning on consolidating this niche market. They had recently acquired another direct competitor, HeadLime.
Dave’s message was simple, he wanted to ‘team up’ i.e. acquire Shortly.
The thought of an acquisition had not occurred to Qasim. He had built a rocket that he was just about keeping under control and was more focussed on how he would maintain it and balance it with medical school.
Qasim reached out to his deal doula: Ismail Jeilani. Ismail is my Co-founder and the CEO of LiveLink. He’s a natural salesman and is responsible for fundraising in our team. Ismail was part of Qasim’s in-team. Throughout the rest of the deal, they would plan Shortly’s next steps together.
Deal doula · deel doo-la
An advisor who is there to provide guidance and support to a founder during their acquisition/exit process.
After his first discussion with Ismail, he realised he had three options:
Sell the company. Qasim had an ideal number in mind, and if they were willing to go with that, he would be willing to sell. This may have been his first company but it wasn’t his last. Selling would free him to focus on other projects.
Raise capital and accelerate growth. If he wanted to pitch VCs, he would have to pitch a billion-dollar vision and probably commit 10 years or so to taking this as far as possible. There was a lot of upside that could be captured.
Run the startup as a side project and take profits. Shortly already had 3000 paying businesses, and his costs were almost exclusively GPT-3 licensing and cloud hosting. His margins were good and he knew they would scale predictably.
All of these options were lucrative and attractive but in his heart, he knew which route he wanted to take.
If he raised investment, he would have to expand his one-man band, which meant building a company and team around the product he had created. That entailed management, operations, engineering, sales and marketing personnel. Speaking to other founders, he knew this would be tough.
Making sure his product had technical defensibility would also become more important over time. Shortly’s product experience was reliant entirely on GPT-3. He knew its pros and cons, and he didn’t know if he was willing to build the tech that would cover for its shortcomings.
Shortly, despite its success, was another one of his curiosity projects. He loved AI and seeing how far he could push the technology. Did he really want to spend the next 10 years helping businesses write better copy?
Selling the company was the most attractive option but for any good negotiation, you needed leverage. The ability to walk away from a deal is the ultimate leverage.
Whether he wanted to or not, for the sake of the deal, Qasim would have to begin fundraising. If Qasim could get a strong valuation from external investors that would improve his asking price for the acquisition.
Shortly’s revenue numbers clocked the company in at the Series A stage. Startups in this range usually raise at valuations of $15m+ as they have a lot of growth potential. Ismail began introducing him to potential investors.
The deal gets serious
Qasim responded to Dave saying he would be interested in discussing ‘teaming up’ and that he was actively fundraising for Shortly given its success.
Since Jasper had recently made an acquisition, they knew the whole process and already had all of the documentation from their purchase of Headlime. They were also using GPT-3, so they knew the costs of running a business based on it. There was every reason for this deal to go quickly.
Jasper began their diligence process by asking about revenue, metrics, costs and team members. They were shocked to find that this juggernaut of a company was a self-taught hobbyist in medical school.
It was clear to Qasim at this stage how the two companies compared. Shortly’s product excelled at long-form writing. He found his competitor’s users would leave and turn to Shortly for this feature.
Jasper sent a napkin offer. It was nothing set in stone but it was surprisingly close to Qasim’s ideal number, so he knew there was hope. Qasim quipped back that Shortly’s revenue was fairly high and that he knew his company was worth more.
Qasim's ideal number was less than the company’s true value, but he would be very happy with it. Despite this, he pushed back for a much higher number. Again, he made it clear that he had alternatives and was still fundraising. In Elon Muskian fashion, he gave his final number, which was a huge discount on the valuation he would give to VCs.
Jasper’s team immediately agreed and began their deeper diligence process. They requested all sorts of metrics as well as evidence of it. That evening Qasim was given a more official offer.
Jasper’s team sent a letter of intent (LOI) and a non-disclosure agreement (NDA). Now things were getting serious. It was time to bring in the suits - lawyers that is. Qasim reached out to other founders who had done acquisitions and found a reliable lawyer for himself. Someone who would fight tooth and nail for his deal.
Both sides were smitten and wanted the deal to go through fast. After all, Jasper had just finished the process of acquiring Headlime, they had everything in place already. They estimated they would get this all done and pay Qasim in two weeks. Spoiler alert: your estimates are usually way off.
There was a lot of back and forth on the deal but these were the relevant changes Qasim pushed for:
The deal should be done in British Pounds (GBP) and not USD. As a result, Qasim wouldn’t have to deal with transaction fees.
The deal would have to be done under English law. It would be easier for Qasim to resolve potential future issues, as he was UK based.
Jasper agreed, updated the LOI, and signed the document. Now the acquisition process could truly begin. Jasper wanted to see his codebase, roadmap, known issues and proof of literally everything the company had done. Shortly was legit, so this process was smooth sailing.
In the meantime, the lawyers had gone to work drafting the share purchase agreement. The constant back and forth was taking its sweet time. With each week that passed, Qasim’s revenue was growing fast, and the valuation he had agreed to was becoming outdated.
Although Qasim was content with the deal price, his advisors pushed him to ask for more, as the slowness of the deal was shortchanging him. He had to defer his medical school exams to the summer so that he could give the deal his full attention. Jasper was understanding and agreed to the higher valuation.
The agreement was drafted after a month but now there was more back and forth. These sorts of acquisition agreements can very quickly become a big land grab where lawyers try to get upside for their clients from every angle. Both sides told their lawyers to be lenient.
The acquisition began in early April and ended in June when the payment was made. As we in the startup scene know, the deal isn’t consummated until cash is in the bank.
What happened next?
Jasper took over Shortly, and Qasim helped them transition over the next two weeks.
Qasim had his final year of medical school remaining, so he decided to finish that off. So that’s Doctor Qasim Munye to you.
He went back to following his curiosities, namely AI learning and research. His passion for AI and his exit opened doors for him. He met with Greg Brockman, the founder of Open AI, who then introduced him to Sam Altman, President of Y-Combinator. Both gave him great advice on navigating post-acquisition life.
Qasim keeps up with the state-of-the-art in AI. He loves working with large AI models and musing over how we can use them to create great, reliable, and scalable products.
Insights from Shortly’s successes
In any deal, create leverage
If the other side thinks you are their only option, then they dictate the terms.
Qasim knew he needed a better position to start discussions from. He strengthened his side by beginning fundraising, which is not uncommon for founders thinking about the next stage of their startup. Shortly wasn’t as big as Jasper, but if Shortly was fundraising that means they had big plans, and missing out on the acquisition now could lead to the creation of a major competitor for Jasper. This reframed the conversation and made it seem like Jasper could get two birds with one stone.
Beginning fundraising also put a more definite number on the company’s value. If Shortly was fundraising at valuation X, then that was a good number to start conversations with acquirers. Now, if Qasim turned around and says he would sell for a lower number, that adds to the case that Jasper is getting a good deal.
Opportunism beats luck
When you see someone beat the odds, it’s easy to say it was luck. When you see them do it consistently, you know there’s more going on behind the scenes.
Listening to Qasim’s story one thing stands out: he seemed to always be at the right place at the right time. You could say it was luck, but many people never did what he did. What distinguishes his achievements from others was that whenever an opportunity presented itself Qasim didn’t falter, or add the idea to his list of ideas. He acted on it. Knowing how to code, he didn’t have to rally a team together to get a project off the ground. He could go from idea to building products as fast as he typed.
It’s not about resources, it’s about resourcefulness
I was also in the Kings20 startup accelerator with Qasim, and I remember that he was never afraid to reach out and get advice. Whether he was asking a more experienced developer for best practices or a designer for feedback, he was always learning from those around him.
It makes sense that he got what he wanted out of his acquisition - he wasn’t afraid to reach out to others in his network.
Others may limit you, there’s no reason to limit yourself
Constraints help to focus our thinking and direct us. But they can also pigeonhole us and stifle creative exploration. Qasim didn’t let labels like ‘medical student’ stop him from becoming a developer. He didn’t let being just a ‘developer’ stop him from building a successful company.
Your background and context don’t define you as much as you think it does. Thinking from first principles, it’s easier to identify what is truly stopping you and when you are letting labels and titles get in the way.
—
As he continues his journey, one thought keeps coming back to him when others ask for advice. If he never sold those vouchers at 14, he would not be here.
“The journey of a thousand miles begins with a single step.”
Lao Tzu
You can find Qasim on Twitter.
Thanks for reading this issue of The Exit Game. If you’re finding this newsletter valuable, consider sharing it with friends, or subscribing if you haven’t already.
Your friend,
Mujavid 👋
Really inspiring journey and fantastic first edition. Looking forward to more stories like this