How to turn your love of reading into a startup?

Hello, Khabravians!

My name is Semyon, I am a mobile app developer with ten years of experience. I also love reading science fiction and fantasy. I read a good book in 2 days. And then I spend half a year looking for the next one. For the last five years I have hardly read anything because I can’t find it.

I feel like there is a big crisis in the book industry right now. I will write another article about it later. But at that moment I asked myself a practical question: I want to be able to find a book that interests me without spending months scrolling through library websites. This is where the idea of ​​a mobile app — a recommendation system — was born. Now we have a team of seven people, we are preparing to release an MVP and have launched crowdfunding.

In this article I want to tell you about us and the path we have taken.

***

Experience

Ten years ago I got my first job with the idea that I would work for a year, save up some money, quit and start my own projects.

Of course, I didn’t save any money in any year…

A few words about the projects I took on, what I ended up running into, and what conclusions I made.

Summary. A prototype was assembled on Arduino, with the best of the existing sensors. Does not work in sunny weather. As it turned out, laser sensors work in the same range as the sun. Very serious technical work is required to isolate the signal.

Conclusion. If the idea is obvious, but there is no product on the market, then there is probably a serious obstacle to implementation. Before taking the idea into work, you need to carefully analyze why it has not been implemented before.

Summary. The first prototype is ready, a community of interested people has been gathered. The main developer left the project due to personal circumstances. No one was able to continue working on the code, as the technologies chosen were too specific. The board also required a new revision, for which there were no resources.

Conclusion. It is worth taking on hardware development only if you have large resources and deadlines. It is also necessary to choose the most popular technologies.

Summary. Closed 2 months after launch

Conclusion. Just because people say they really need something doesn't mean they're willing to pay money. More research is needed. And remember, details matter. In this case, we chose the wrong location.

There was a lot more, we won't get distracted by everything. The main thing is that with each mistake a map of the labyrinth is built in the head, which we have to go through. And chances to dodge the next trap appear.

Key findings I made these.

• You need to choose an area that you really understand.

• The project must solve a problem. And it must be your own personal problem, not taken from someone else.

• Before you get down to work, you need to verify the idea. That is, make sure that the product will also solve other people's problem. And this problem is so relevant that people are willing to pay for the opportunity to solve it.

• MVP should be implemented with minimal resources and in minimal time

• If someone volunteers to work for free, make sure that they actually have the time, ability, and willingness to do the work. Doing the work is not the same as talking about it.

• In technology, you need to look for simple, reliable and proven ways. Let the product characteristics be worse, but it will not die when an unsolvable bug is found in a new technology. Or when the developer leaves.

And the most important conclusion. Banal, but very important. You need to love your product, project, team. Love and believe. And at the same time, don't forget about healthy skepticism. This is much more difficult than it seems.

What is the problem with existing recommender systems?

Many bookstores and libraries have their own recommendation system. Such systems can be personalized and non-personalized. A non-personalized system is a list of the top 100 most popular books. A personalized system is suggestions based on previous purchases.

Non-personalized systems are useful and do their job. But those who read a lot have long since read everything from there. Besides, if a book is popular, it doesn't mean that you will like it.

Personalized purchase-based recommendation systems also have a number of drawbacks. Avid readers always have a lot of books that they read as children, or bought on other platforms. Let's be honest, most of us download books from pirated resources for free. Such books cannot be taken into account in purchase-based recommendations.

Besides, to rate a book, you need to read it first. A purchase does not mean that the user likes the book.

The goal of stores is sales. The recommendation system is just an additional marketing feature for them.

There are also applications similar to ours. For example, the Livelib application in Google Play is described as follows: “Reading diary, quotes from books and reviews. Stories about books and authors…”. This application really has a lot of interesting functionality. It is a super-application for readers. And, like other super-applications, it is difficult to understand.

We see the simplicity of the interface as one of the key aspects of our application. And tailored specifically to the recommendation system. The interface should be so clear that a new user could dive into the process of searching for books fifteen seconds after opening the application.

The idea of ​​the “What to read?” app

• The interface is tailored specifically for the recommendation system.
  The concept is based on Tinder. Swipeable book cards. But unlike Tinder, there are four actions. “Skip”, “Want to read”, “Already read, liked”, “Already read, didn’t like”.

• The widest possible base due to integration with various stores and libraries. Including resources where authors themselves post their texts. Possibility of customizing the used bases.

• Download links in various stores and libraries. Which will allow in some cases to find a book for free, and in some cases to find the best price.

It's very popular now to use AI everywhere and talk about it.

We say that one of the key benefits is that Not We use AI. All recommendations are based on likes/dislikes.

The principle of operation of the recommender system is as follows:

• Vasya likes the books “Harry Potter”, “Tanya Grotter” and “Crime and Punishment” by Dostoevsky. That's how strange our Vasya is.

• We find users who like “Harry Potter”, users who like “Tanya Grotter”…

• We take the lists of books that these users like. We sum the lists with the repetition count.

• We sort the resulting list by repetitions, get a recommendation feed. We add some degree of randomness.

I will talk about the implementation details in the next article, but for now we will not go into technical details.

Of course, our plans for the app's functionality are not limited to recommendations. However, this is the core around which everything else can be built.

Book community

In the future, we plan to not limit ourselves to the app, but to create a real book community.

For actions in the application, for example, for writing quality reviews, we will award points. These points will give bonuses from partners. Discounts in bookstores, in literary cafes, at various events.

We are preparing integration with city libraries. This will allow, by going to the book card in the application, to find out in which city library it can be borrowed for free.

We are planning to implement book sharing. Peter has a volume of Terry Pratchett's “The Colour of Magic” read three times. Peter is ready to give it to some lucky person. He indicates this information in the application. For all users, the card of the book “The Colour of Magic” will indicate Peter's contact in the application.

We still have a huge number of ideas. The only thing left is to implement them all.

How we went and where we are now

I started working on this topic three years ago. survey.

I sketched out a mockup of the interface on my knee, asked one key question – “how interesting is this to you?” and several additional ones. I threw it into social networks and suddenly 350 people answered the survey. Many even left text comments with their suggestions for functionality.

The survey is still up-to-date, I'd be glad if you take it. It's quite short – six questions.

Then I tried to find investments, communicate with people, organize something, and got stuck for two years. And everyone knows what the last few years have been like, there was no time for startups.

Then I realized that I could do everything myself faster than find investments. And in a month I wrote a working prototype of the application.

Then I ended up in Georgia, talked about the app in several chats, and it turned out that there were a lot of good people who were immediately excited about the project and ready to participate.

This is how Artem appeared – a wonderful backend developer. And I am very grateful for him to those wonderful HRs who do not allow the coolest specialists to a technical interview due to a lack of commercial experience in their resume. Thank you, HRs!

Overall, it's awkward, of course. But I realize: if Artem had gotten a normal techie instead of an HR guy, he would already be working at a job for good money.

Also Anya joined the team, to whom I gave all non-technical tasks. Anya is great, other people started coming to us from Anya.

I won’t go into detail about the whole team here; that’s a topic for a separate article.

As a result, I am now finishing the application, implementing a new professional design from Dima.

Artem is finishing the backend.

With Anya's help we launched crowdfunding.

We really hope to release it to closed access on Google Play in about a month and begin internal testing.

How to be a g̶o̶b̶r̶o̶m̶beaver

There is one more topic that I would like to talk about in this article.

When it comes to business, everyone immediately remembers competition, markets, struggle, and so on. You need to get into someone else's market, keep quiet, gather strength, and then crush everyone with your hitherto hidden ego. Or take it all at once by storm.

Wrestling and pushing butts is something we really don't want to do.

In game theory, there are concepts: “win to lose” and “win to win”. There are others, but this is a topic for another article. “Win ​​to lose” means that there can only be one winner in the game. The rest lose. And here the struggle is inevitable. But in a “win to win” game, everyone can win. For such a game, the most profitable strategy for all participants is cooperation and mutual assistance.

So, how to be a beaver in a “dirty” market?

We need to find a concept that will make the game a “win to win” game, not a competition.

This is what I particularly like about the idea of ​​a recommender system.

Our project does not compete with any market participant.

The book cards will contain information about all legal sources where the book can be bought or received for free. This is beneficial for all stores because it is enough to temporarily reduce the price of some goods and users will go to them. Our application is a great tool for honest marketing.

For readers, we are useful, of course, for the recommendation system itself. But also for the opportunity to find where to get a book cheaper, and sometimes for free.

We are also useful for all organizations that associate or want to associate themselves with literature. We plan to award users with a rating, for example, for writing high-quality book reviews. And we really hope to find partners who are ready to give bonuses to our users. For partners, this is marketing, and for users, these are the same bonuses. Plus, the application will have many high-quality reviews, which will be nice for both readers and authors.

So we really hope that we can remain cute fluffy cats that everyone loves and appreciates. And we will leave all the fighting, pushing and competition to others.

Conclusion

This is the first article from our blog. I tried to tell about the idea of ​​the “What to read” app, how it was born, how the idea turned into a real app, where we are now and where we are going.

If you were interested, go to the page crowdfundingyou can read about everything in more detail there.

We tried to come up with good rewards for sponsors. You can get an eternal subscription to a special additional functionality of the application – for example, the ability to find people with similar book interests. Or we can send you a personalized thank you card from the team.

Thank you for your attention!

You can also subscribe to us:
https://t.me/whattoreadapp
https://www.instagram.com/whattoread.app