Crypto Autotrading

CoinLion had a vision of making crypto easy for everyday investors, but their complex interface was more suited to technical traders and engineers. Less experienced users struggled to understand the status of their investments, and sometimes made costly errors.

CoinLion’s signature “autotrading” feature – allowing users to subscribe to and automatically mimic the trading strategies of more experienced traders – promised to take the hard decisions out of crypto trading. However, this hidden feature lacked a front-end interface and was only available by special request.

While defining the information architecture of strategies and subscriptions, some of my first work was to diagram and propose clarified terminology (below).

Our design needed to address two primary audiences with very different workflows. 95% of users were novices, valued simplicity over detail, and simply wanted to subscribe to the trading strategies of others. But 5% were experienced power users who needed the ability to create strategies – complex trading algorithms that buy and sell in response to market signals. Their goal was not only to make a profit, but also to attract subscribers.

Crypto was new to me, so I took a deep dive into terminology and competitor products. Meanwhile, I also started diagramming some discrepancies between my mental model of how I expected an investment product to work, vs how CoinLion’s backend was set up.

The difference between reinvested growth and rebalanced growth turned out to be key. In order to compare the relative success of different strategies, a user should be able to choose to reinvest profits in the same strategy that generated them. However, CoinLion’s back-end was only set up to rebalance into a single pool of funds per account.

Without time for formal research, I asked colleagues and friends which behavior they would expect from an investment product. Overwhelmingly, they chose reinvestment. It turned out, even CoinLion’s executives agreed.

The reinvestment option, like many other feature concepts, didn’t make the cut for our MVP launch. An ambitious timeline meant we had to be strategic about choosing the most essential package. We worked with the client to compile a complete wishlist of new features and functionality fixes. Using a balance of business goals, user needs, and implementation difficulty, we created a roadmap of priorities for an initial launch, followed by subsequent releases.

We used several rapid rounds of low fidelity wireframes to seek feedback from stakeholders on a vision for how autotrading might look and feel. As the pictures above illustrate, our homescreen evolved to visualize autotrading activity in a unique way.

Where traditional financial products use a line graph to focus on performance over time, we settled on a bar chart that would track the amount of profit or loss from each autotrading event.

The image below shows the final design.

Although the flashiest screens in the app aimed to serve the everyday investor user, we hadn’t forgotten the importance of the expert traders. Designing the custom strategy creation screen was one of the most fascinating challenges.

Repeated interviews with a seasoned trader helped me to design a system for easily adding and subtracting different types of signals from “buy” and “sell” categories, as well as “pause buy” and “pause sell” for safety.

Both the autotrading marketplace and the trade history screen focus on showing performance data that will help investors make informed decisions about who to subscribe to, and how their current subscriptions or strategies are performing.

If I could revisit the project, I’d love to dig deeper into social features that would increase the level of accountability and reward for each publisher, as well as give subscribers the ability to review their experience.