The path to the button "loot" on the road from a rake

    In this article I want to tell how we created a cryptocurrency startup, and what came of it.

    A picture to attract attention

    0. Product Idea


    The idea is to make a loot button.
    It is clear that there is no such button that, when clicked, sends “the very precious ones”, although there are services that promise this. And yet, we tried to develop the idea as much as possible.
    We understood that the “loot” button is something complicated inside, but at the same time simple outside, connected with money or assets, packed in an intuitive interface.
    What would such a service look like? For example, the “Follow” button, when clicked, the system starts repeating the trading actions of one of the professional traders. The user only has to be enchanted to monitor the balance change.
    Feasible? Yes. Great, sawing!

    What they did right: The idea is interesting and inspiring.
    What did wrong: Started to work without a detailed and detailed project.

    1. Team


    A classic of a Russian fintech startup: there are a lot of developers, marketers are empty.
    Initially, the staff was 6 people: 2 programmers, 2 managers, 1 trader, 1 marketer.
    Later the team included 4 programmers, 2 managers, 2 marketers, 1 layout designer, 1 trader, 1 designer.
    Everyone is incredibly talented, and as a goddess star. With such an introduction, the communication structure turned out to be “flat”.
    The udalenka took root, but not very much. In a startup, a lot is decided in the course of oral informal communication. With great speed and constant changes, it is important to be in context.

    What they did right: They recruited diverse specialists, worked well together. We didn’t inflate the team (two-pizza team), which helped to be fast and flexible.
    What did wrong: They took a content manager who was not interested in the topic of crypto, had to leave.

    2. Implementation


    For implementation, we chose the Binance crypto exchange according to the following criteria:

    • Open API
    • The largest in terms of trading volume (more than 50% of the market)
    • Many currency pairs
    • Her story lacks major hacks and theft of accounts
      (UPD: while writing an article some incident nevertheless occurred, but everything went well for the exchange)

    We were not happy with creating a web version of the system, because the priority was speed of access to balance, mobility, security, and access to the wallet without unnecessary authentications. And such an option was found, we created a Telegram bot that connects to the exchange.
    Telegram was chosen because of the wide possibilities for creating bot interfaces, and also because this messenger is popular among the crypto community.
    The main code was written in JS, and, after some time, its flexibility and dynamic typing began to slow down progress. It was decided to switch to TypeScript.

    What they did right: They chose the exchange and the messenger correctly - both sites continue to grow.
    What they did wrong: Lost time in a dynamically typed language.

    3. Pilot version


    In the beta version, the wallet balance was displayed, there were notifications about trading operations, there was a profit lock button.
    Those. the task was absent, so some functions were invented during development, for example, the “FIX” button. This button instantly converted all coins to a stable currency if the market began to fall. So the accumulated on growth remained. It's nice to know that you have a stop crane.

    Balance screen from the first version:

    Old Benice balance interface

    Now the balance interface looks different:

    New Benice balance interface

    What they did right: They started as soon as possible.
    What did wrong: Then I had to redo a lot.

    4. Promotion Methods


    We chose an English-speaking audience, because, in theory, they are more loyal to cryptocurrency, and the size of such an audience is considerable. All content was created in English, without professional translators, on their own.

    4.1. Social networks
    I wanted to be everywhere on the network, which means nowhere, so I chose the sites where the crypto audience sits the most. Telegram (news channel + chat), Medium, Facebook, Twitter, YouTube, Tradingview got under the spotlight.
    So far, no resource brings the desired effect, but the content continues to be published persistently.

    4.2. FB targeting
    Due to the subject matter of the “Cryptocurrency” Page, several licenses had to be presented according to FB requirements. More specifically, it is a “Coin Listing Document”, “Public ICO Document” and something else. We didn’t have these papers, because we are “software”, which works with the crypto exchange, so on the next attempt to explain this in the ban report, a thick bolt was put on moderation. In addition, 3 advertising offices flew into the eternal ban; it was impossible to provide access to managing a banned office.

    4.3. Telegram advertising posts
    A separate story about the cost of advertising posts in crypto channels, it is unreasonably expensive. And another story is that 60% of the channel’s audience is either catch-up bots or people who don’t constantly monitor content, so posts collect 30-40% of the total views, and then the post is permanently deleted.
    Where it was possible to place native advertising forever, it showed a satisfactory result.
    Example: t.me/tgTONru/316

    4.4. Programmatic advertising A
    monthly package for contextual advertising in the Hybrid service was bought, but there was no launch yet, because the site is not ready for “super targeted” traffic, and causes more doubts about the product.
    Contextual advertising on Google did not run in order to save budget, and for the same reason, questionable content on the site.

    4.5. Collaborations with opinion leaders
    Not a single opinion leader did a review on a product, did not recommend it to its audience.
    During the negotiations, so far only refusals have been received, or the advertisement is worth 6-digit space amounts.

    What they did right: The bet on native advertising worked.
    What they did wrong: We got a weak result from social networks. We did not use such sites as Linkedin, Reddit, large crypto-forums. Email contacts were not collected anywhere, so this type of marketing was designed.

    5. Monetization


    Monetization takes place by paying a monthly subscription to the premium features of the bot, and is paid using BTC. Before paying, you must first enter the API keys through which the account is connected to the exchange, then find the wallet menu inside the bot, and transfer BTC equivalent to $ 30 to the specified crypto account. After that, an internal settlement coin will be credited, for which access is paid.

    What they did right: There is a clear monetization model.
    What they did wrong: To pay, the user needs to click somewhere 20 different buttons, and this is only in the best case. In addition, the monetization model had to be changed later (more on that in the next article).

    6. Metrics


    Bot admins can see how many users logged into the bot, but this was not enough for the marketer, and I wanted to know which buttons users click on. And with the condition that you need to enter API keys to connect, you need to know how many people got into the main menu, and not just pressed the START button and left.

    To do this, they tied Google Analytics to the bot and set up events. How this is done is written in this article ( https://habr.com/en/post/442610/ ).

    What they did right: They were not afraid and made friends with Telegram and GA.
    What they did wrong: Everything is right.

    7. Website


    I would like to show the first version of the site that creates a wow effect, but it turned out to be very useless in terms of sales.
    ( https://benice-tools.github.io/benice-space-landing/index.html )

    After the realization of the beautiful dummy came, the site became more serious and “corporate”.
    ( https://benice.tools/ )

    What was done correctly: The structure of the site immediately changed depending on the general vision of the project.
    What they did wrong: They did not investigate the audience and tried to please everyone at the same time.

    8. What next?


    There is a category of telegrams in Telegrams that publish trading signals (recommendations for profitable transactions). These publics have VIP channels where there are better signals. Public administrators showed interest in integrating our bot with their channels, and since we put in the bot a system of interest payments, for them this is an additional way to monetize and simplify their routine.
    portfolio of public trader

    Previously, in order to apply the signal, you had to put trade orders on the stock exchange with your hands. With our service, you can forget about this complicated routine and fear of the exchange interface.
    Now, trading signals will be applied automatically, even when users are offline.

    9. Summary


    portfolio statistics

    So, despite all the obstacles, the project is up and running. On the one hand, this is a chance to thank the crypto community. On the other hand, the service daily makes life easier for the creators themselves. The process of birth of the service was fascinating, and we do not intend to stop.

    Also popular now: