foreword
I think a lot of people have read .
As they wrote in the comments to this article, the statistics are depressing. Therefore, I would like to tell about myself in order to correct the statistics and give some practical advice on how not to be blown away after the hackathon. If at least one team, after reading the article, does not stop developing their cool idea after the hackathon, takes my advice and creates a company, this article can be considered successful 🙂
Warning! This article will not be technical details of the application implementation. At the beginning, I will tell our story (TL;DR), and useful tips that we have deduced for ourselves during development are given at the end.
“Success” story
My name is Danya, I founded emovi, a service for selecting a movie by emoji, which has grown virally by 600% over the past few days. Now the application has 50 thousand downloads and is in the top 2 of the App Store and Google Play. In the team, I am engaged in product management and design, and earlier Android development. I study at MIPT.
Disclaimer: We understand that this is just the beginning and not a "success story". We have a chance to both continue to grow rapidly and lose everything. But, taking this opportunity, we decided to tell our real story, hoping to inspire those who want to launch their own startup someday, but have not yet come to this.
The path of our team started at the Finnish hackathon Junction, where there was a track dedicated to film services. A team from Phystech won that hackathon, they managed to do more, but did not continue to develop . At that time, we formed a concept - search for films by the emotions they evoke, using emoticons. We believe that the abundance of information about the film: long reviews, ratings, lists of actors, directors - only increase the search time, and choosing a few emoji is quite simple. If the ML algorithm that determines emotions in movies works well, and we remove the movies that the user has already watched, then it will be possible to find a movie for the evening in 10 seconds. But the reality was then completely different, and with such a project we .
After the defeat at Junction, the team needed to close the session, then we wanted to continue developing the project. It was decided to move towards a mobile application, due to the lower level of competition compared to websites. As soon as we started getting together to work, it turned out that not every team member is ready to devote their free time from study (and for some from work) to develop a project that:
- complicated
- labor intensive
- requires full commitment
- not the fact that someone needs
- will not bring profit in the near future
Therefore, very soon there were two of us left: me and my friend from the Higher School of Economics, who helped with the backend. Coincidentally, I lost interest in scientific activity just at this moment of my life. Therefore, despite the good performance, I decided to go to the academy. I hoped to have time to launch a new project in a year and find myself in a new activity. It's also worth noting that the problem of choosing a movie for a long time on Kinopoisk has always been my pain, and I wanted to alleviate it by offering people a new way to choose.
The challenge was in building an algorithm for determining the emotions of the film and collecting a dataset, also because we did not have a professional specialist in data science. And also, as a developer, to make a convenient and new UX, but at the same time a beautiful UI. After I redid the design 10 times, I ended up with something quite comfortable, and even good looking, thanks to some innate sense of beauty. We started writing backing, collecting a database of films, the dataset we needed, and developing an Android application. So the spring and summer passed, there was a movie base and API, an MVP of the Android application was made, a dataset appeared, but there was no ML algorithm for predicting emotions.
At that moment, what was to be expected happened: my friend, who was involved in the backend, could no longer work for free, got a part-time job at Yandex, and soon gave up on the project. I was left alone. All I've been doing these six months is a startup and part-time tutoring. But I did not abandon it and continued to move on alone, simultaneously offering to work on the project with various DS with FKN, but no one had the motivation to work for free.
In September, I went to Fiztekh.Start, where they didn’t take me, but where I met my current co-founders. After talking about the project, I convinced the guys to join me. So, before the October hackathon Hack.Moscow, we were engaged in a full-time project. We made an iOS version of the application, and wrote the main algorithm that uses NLP to determine emotions in films. On we came with a finished project (the track allowed this, it was called “My track”) and only dealt with the presentation for 36 hours. As a result, we won, received good feedback from mentors, were invited to in December and were very enthusiastic.
After the hack, work began 24/7 on the product ahead of Launchpad. We came to it already with a finished product, a working beta on Android and an alpha of iOS, and a new co-founder from the HSE Faculty of Computer Science, who replaced me with the backend, since I could no longer do Android, backend, design well and think about it what else users need from the product. At Launchpad, we were greatly upgraded in marketing and product management. In a month, we completed everything we wanted, released and ... nothing happened.
The application itself did not gain anything, although it seemed to us that it should (we just made publications in our social networks, Peekaboo and in a couple of telegram channels).
When the first disappointment from our own misunderstanding passed, we began to analyze what went wrong, but everything was exactly as it should be, because then we did not understand anything about marketing and PR, and the product did not have viral features.
Since there was almost no money, we survived on cheap advertising in public VK, which kept us growing by 1K installs per week. This was enough to test product hypotheses on this audience and at the same time look for investments, having become acquainted with most of the Moscow venture capital industry through various pitches and conferences. We went to the HSE Inc accelerator, where we worked on the product, business development and attraction of investments, additionally went to the course “How to make a product?”, the founder of Prisma and Capture Alexey Moiseenkov, which helped a lot to understand what to do next. But things were not going as well as we would like: there was little growth, and our Data Scientist left to work ... guess where?
— Yes, to Yandex!
- Who?
- Producer.
We almost developed a new section in the product related to video, we were engaged in attracting investments, which helped to work out the understanding of the streaming market, business model and vision. I learned how to convey this to investors with varying degrees of success, but otherwise there was no clear light. There was only faith in ourselves and in our insight that no one had solved the problem of choosing a film on the Russian market in free services. By this point, the money ran out, we started doing zero-cost marketing, which brought very little. It was very hard, but faith and XNUMX% focus saved me. During the accelerator, we actively communicated with various experts, investors, received a lot of feedback - not always positive. We express our gratitude to all the guys from HSE Inc for their support in difficult times. As founders, we understood the specifics of a startup and believed that nothing was lost yet.
And then we made a post on Peekaboo and went viral. Basically, the main task was to find users who really need our application, they turned out to be the guys from the “Serialomania” thread on Peekaboo. They were the first to pick up the wave, liked and shared a lot, took us to “Hot” and then we only had problems
We reached the top of the Play Market and the App Store, received 600 reviews, we fell and rose, and at the same time wrote press releases for publications and gave interviews ... Special thanks to the largest community of hackathons where people helped us solve technical problems for free.
By evening, the hype had subsided, the servers were working normally, and we were about to go to bed after a 20-hour marathon, when the incredible happened. Our application was liked by the admin of the NR Community public and he made a free post about us in his group of 5 million people without our knowledge. The servers were already better able to withstand the load, but we still spent most of the time optimizing.
But, as YCombinator says, if your servers go down, then it's a success (Twitter is cited as an example). Yes, it would be better if we were ready for such a load in advance, but we did not prepare for such success after this post.
At the moment we have an offer from an investor, and we will develop further. Our main task is to refine the product so that it suits the majority of our users.
Now let's move on to the tips. Our team has a lot of faith in survivor error and believes that advice like “Do A, B, and C” is useless. Let the business coach talk about it. Peter Thiel wrote in "Zero to One": "Anna Karenina begins with the words 'All happy families are happy alike, all unhappy in their own way', the exact opposite is true of companies." The path of each company is individual, and no one will tell you how to do your business. But! You may be told exactly what not to do. We made some of these mistakes ourselves.
Tips
- Due to the high competition with large companies, a b2c startup needs a high quality product, which is extremely difficult to implement without experience in creating b2c products, without people willing to devote themselves to this for a year for free, or angel investments that give you time first of all. We are sad to state this, but it is almost impossible to find b2c angel investments in Russia without growth or great experience, so if you have hypotheses about opportunities for b2b, it is better to do b2b in Russia for now, because there your first revenue will happen earlier.
- If you still decide to do b2c without money, the problem you are solving should be your own. Otherwise, you will not have enough strength and desire to finish it to the mind and motivate your team.
- If after your pitches (approx. presentations to an investor) your project is reacted extremely badly, then there are two options: either you really should listen and make a pivot, or the market simply does not understand you, and you have found an insight that many overlooked. It's the kind of things that others overlooked or considered unimportant that help some startups grow fast every year. It is clear that the probability of the second is less than 1%, but always think with your head after you have listened to everyone, and do what you believe in, otherwise you will never find such insight.
- That is why an idea is worth nothing, because if it is worth something, then only 1% will believe in it, and 1% of them will start doing it. About 1000 people come up with the same good idea at the same time every day, but only one starts doing it, and more often than not, it doesn't finish. Therefore, do not be afraid to tell everyone about the idea.
- All that you consider it necessary to do is your hypotheses, according to which KPIs are needed to confirm them. Your time must be planned, you must know what you are doing on which day, what hypothesis you are testing this week, how you will know that you have tested it, and what is the deadline, otherwise you will be bogged down in constant “doing”. Your answer to the question “What did you do all week” should not be “I did X”, but “I did Y”, where done most often means testing some hypothesis.
- In b2c, your hypothesis can be tested either by competitors' products and the market (for example, there is already a service that solves the problem, but you can do it several times better), or metrics in product analytics, such as Amplitude, Firebase, Facebook Analytics.
- If you are doing b2c, listen less to the fans of the CustDev methodology popular in Russia, who apply it where it is necessary and not necessary. Qualitative research and conversations with users are needed to identify insights, but they cannot statistically test the hypothesis in any way, since they are not quantitative research methods.
- Invest only after the MVP and testing of the main hypotheses, unless, of course, you have a startup experience in the past. If you have a b2c startup, then without revenue it will be very difficult for you to find an investor in Russia, so think about either how to start growing in terms of users, or how to start earning.
- A startup is, first of all, the speed of growth and decision-making. In the current venture realities in Russia, a quick move for a b2c project is not always possible, but do everything to move faster. This is why the founding team usually consists of 2-3 people working full-time, and a team of 10 part-time friends at the very beginning is a mistake that will kill you. A lot of people also feel bad about the fact that a new problem arises: there should be a separate project manager who does just that and gets a share simply because you could not find enough motivated co-founders.
- Do not combine work and startup. It's just not possible and will kill you sooner or later. you as a company. Personally, everything can be “perfect” for you, you will be hired by Yandex and you will receive a large salary, but you are unlikely to be able to build something big there, because your startup will move too slowly.
- Don't get carried away with everything. One hundred percent focus is very important to you, without which you will be pivoting (changing course) 3 times a week. You must have a strategy and an understanding of what to do, where you are going. If not, start with an analysis of competitors and their position in the market. Answer the question "Why didn't X do what I want to do?" before coding anything. Sometimes the answer may be “they considered it a non-priority and they made a mistake”, but the answer should be.
- Do not work without a quality metric (this is more about ML). When it is not clear what needs to be improved and how, it is not clear what is good and bad now, it is impossible to move on.
That's all. If you do not make at least these 11 mistakes, your startup will definitely move faster, and growth rate is the main metric of any startup.
Materials
As material for study, I would like to recommend the excellent course of Alexei Moiseenkov, the founder of Prisma, from whom we learned a lot.
He will tell you what an IT company consists of, how to distribute roles, look for founders, and make a product. Such a direct manual “How to build a startup from scratch”. But watching a course without practice is useless. We watched it in the video version and went through it in person, practicing in parallel.
Every startup should know YCombinator, the best accelerator in the world, which has released such teams of founders as Airbnb, Twitch, Reddit, Dropbox. Their course on how to start your own startup at Stanford University is also available on YouTube.
I also highly recommend the book by Peter Thiel, the founder of PayPal and the first investor in Facebook.
What are we doing anyway?
We are making a mobile app that searches for movies by emoji with personalized recommendations based on movie ratings. Also in our application, you can find in which online cinema you can watch a particular movie, and user ratings are taken into account in emotional search. Trust us, the emotions are not arranged manually, we have been working on this for a very long time 🙂
You can learn more about us at .
And who wants to download, you are welcome.
A little insight and conclusion
At the end of the article, I would like to strongly recommend not to abandon your projects after hackathons. If you can make a product that people will want, you will never be late to go to work, because you will work many times better and more efficiently than people who have never made a startup. In the end, it all depends on your ambitions and goals in life.
And I would like to end with a phrase that Steve Jobs said to John Scully (at that time CEO of Coca-Cola), when he called him to work at Apple:
“Do you want to sell sugar water for the rest of your life, or do you want to change the world?”
Over the coming months, we will be expanding our team, so if you are interested in working with us, send your CVs and motivations to [email protected].
Source: habr.com