The Problem:
Synthesizing the results of technical and natural language AI analysis, gathered from different datasets and formats (social media posts and technical indicators), without overwhelming the user with irrelevant or confusing technical data.
‍
The Objective:
To create a user experience that should be easy to use, fast, actionable, educational and engaging in content.
Ideally, the app should provide enough actionable information and show the state and prediction of the general crypto market sentiment within a few seconds of opening it, and it should also notify the user in real-time about any shift in the market state.

The branding should also reflect the main function, which is to provide signals in order to anticipate future market moves.

A couple of initial brainstorming sessions with the developer took place in order to assess the feasibility of the objective, after which I've initiated my research.

While researching competitors, several different services were evaluated, and the following features were considered for analysis:

- Signals format
- Data points
- Data sources
- Ease of use
- User interface
- Branding

While some crypto signal and trading bots on the market were identified during my research, the competition falls into two distinct categories:

- The first type uses machine learning to power automatic trading bots, which either connect to centralized exchanges via API and utilize the user's entire portfolio (such as Stoic), or are integrated into the exchanges themselves (such as Crypto.com).
This method forces the user to trade using a "set and forget" approach, requiring a high level of trust in the technology, involving obvious risks. Typically, the user experience in these services is not transparent regarding what or how the trade decisions are performed in the background, with no comprehensive information being shown due to the use of simple presets that control the behaviour of such bots - therefore these are also not educational.

- The second type does broadcast comprehensive signals, but these are compiled and edited by real-life traders (sometimes making use of some kind of AI tools to aid technical analysis).
This method poses the risk of market manipulation or human error, and the signals are mostly quite technical in their approach.
It was also observed that quite a few of the competitors use 3rd party messaging apps such as Telegram to broadcast the signals instead of a native app.

It soon became clear that the signals screen UX should probably be unique and with its own patterns since there were no exact counterparts to what we intended to develop, while the Portfolio tracker could partly rely on existing patterns identified on an parallel research, which analysed Coinmarketcap and Coingecko's proven functionality.

This research provided me with many useful insights on how to conduct the interviews and focus group.

The target group was defined through analysis of reports of the average crypto investor population, as well as product specific requirements - such as awareness of trading bots and signal apps.
‍- 20-60 years old
- Any gender
- Retail investor for less than 3 years
- Some knowledge of the crypto industry
- Awareness of crypto trading bots and signal apps
‍
‍
During the session, which involved 6 subjects, the following points were discussed, mostly based on how they conduct their own cryptocurrency market research:
- Time of day that research is conducted
- Time spent on research
- Frequency of research
- Data sources consulted
- Bluesky ideation / Wishlist of features


Each subject was also interviewed privately through a survey about some additional, slightly more sensitive information, such as:
‍- Likeliness of following signals of known Twitter and Youtube influencers.
- Likeliness of investing based on technical indicators such as Bitcoin MACD or Heiken Ashi.
- Likeliness of behavioural changes reflecting macroeconomic factors.
‍
The information gathered gave us the chance to discover some of the pain points of the subjects regarding their current preferred tools and methods of research, and also to learn about their needs, such as showing the future health of the portfolio based on the current investments (which would be the biggest challenge for development).

Some more insights were also acquired regarding aesthetics, branding and tone of voice to be applied throughout the product.

The main takes from the research were singled out as:
- Fast to consult
- Direct and easy to understand
- Reliable and accurate

Two different qualitative personas were developed, with distinct behavioural archetypes and levels of knowledge. This data was collected and averaged from real personality traits and life stories of the interview subjects.

The traits of each persona were leaned out to the minimum relevant for the specific product.

A common trait of all subjects was the lack of time or wish to spend as little time on research as possible - making the product quick to consult was always at the top of our priority list.

The time of day in which research is usually done (mostly at night, before or after working hours) also had an impact on the MVP's visual design, explained further.

Applying the insights from the research from the previous phase, a user flow was laid out.

I've tried to cut unnecessary steps to a minimum of screens in order to optimize the amount of time spent by users in the product - one of the main topics of discussion during the focus group session was the need to design the product as minimalistic as possible, to be consulted on-the-go. Cutting out unnecessary steps was key.

For instance, the first version of the flow included one step for each signal type, but I was already predicting at this phase that there was the need to cut this down to one main screen, which would somehow include all the signals, also because having extra steps would make the UX cumbersome as more signals would be added in later releases - more on that in the following step.

In order to iterate faster at this early stage of design, my preferred method is to start with pencil and paper, as it allows my ideas to flow quicker than using a mouse or a tablet.

One of the biggest challenges was to design a clear experience, so that each data point would not be confused with the other, and without having too much information in one single screen.

It had been decided that there would be three distinct signal types in the prototype and MVP - but with more to be added on upcoming versions - therefore the UI should have room to accommodate more signals in the future. Having all the data in one single screen at the same time would be unsustainable in the future due to screen real estate, and a solution would have to be created to tackle this issue.

The solution I found was a carousel-type of component to show the signals, which would be the hero of the product. It was perfect to grow, and it made possible to add many more signals without cluttering the screen. As an added advantage, this solution also left room to focus about a more striking visual design and interaction at later stages.

The crypto portfolio design was straightforward, as it would rely on existing patterns used in the crypto and fintech industry, adapted to our needs.

The Data Sources screen would also be quite simple, as the default sources are added by the team on the backend, with a simple UI on user side just to turn them on or off as needed (so there was no need of an extra screen to add sources). There is a minimum of one active source required for the signals to work.

Working on the insights gathered during the competitive research and user interviews, a few keywords were gathered to define what kind of brand identity we wanted to create for the product.
‍
The decision to tackle the brand design before proceding to the prototyping phase was intentional, to dig deeper into the "personality" of the product, and perhaps gain new insights regarding the style of the user interface - especially the tone of voice used on the copy, which should be already available to test on the prototype phase.

It became clear that the functionality of the product - heavily based on Artificial Intelligence - conveyed an aura of "technological mysticism" to all subjects of the group.

Some of the keywords gathered were "oracle" and "prophecy" - which by themselves could not be used directly as a name for the product, since these are already either in use or are too generic. These terms are also frequently used in the lingo of investors, cryptocurrency media outlets and influencers. We wanted a name which would be deeper in meaning and more unique, so another research was conducted to find terms which could be cross-referenced with these ideas and feelings.

Upon researching some online resources such as Wikipedia and Reddit, I came across the term "Sibyl", from the Greek "Σῐ́βυλλᾰ". Sibyls were the prophetesses of ancient Greece, which predicted the future based on the will of the gods - the meaning was perfect for our product, the aesthetics of the name were very pleasing, and it was also short and easy to remember - the team decided to move forward with it. By changing the spelling to "Sybyl", we made it unique so it would not have overlap on Google and App Store searches.

On a side note, although a company of the same name was found to exist, it has no connection to the cryptocurrency industry. We have decided to add the claim "Crypto Signals" to the name, in order to reinforce and clarify the functionality of the product.

A moodboard was made to give some inspiration for the logo, based on a search for the terms "oracle" and "sibyl".
‍
Since some of the keywords gathered during the focus group session were also around the "reliability" and "solidity" of the results, I researched to find an element which could be translatable to a simple and modern logo, held some connection to the name, but also transmitted a sense of strength or robustness.

Back to the drawing board and inspired by the moodboard, again my preferred method at this stage was pencil and paper.

One of the first ideas before sketching was focused on the profile of a goddess, and some ideas surrounding a temple (of Dephi), but soon these ideas was abandoned in favour of a stylized Ionic column, reminiscent of the temple of Delphi.

This would allow for a quite simple logo which could be recognizable even in small resolutions, and conveyed the idea of strength and solidity much more effectively than the first ideas.

After some exercises and feedback / voting from the group, the final sketch was approved.

Following closely the sketched lo-fi mockups, I returned to Figma to start designing and prototyping a series of low fidelity mockups for iOS.

Three distinct Signals screens were designed at first, with different orders and methods of data visualization.

Although we were inclined to choose one of them due to the possibility of adding new data sources, I've decided to show all three to the group and test their usability.

I've proceeded to design three distinct clickable prototypes with the different signals screens, and these were sent to the group for testing.

The Portfolio and Signals use well-proven existing patterns, found on similar products: therefore the focus was to discover which Signals screen would have the best experience.
‍
You can try the selected design yourself in this simplifed Figma prototype!

The first solution was rejected by nearly all participants, due to the need for having a quick way of consulting the status of the market. The preferred method by all participants was the data visualization mode. The density of information on the screen was also an issue, which was described as "confusing" and "cumbersome" by some.
‍
Both of the proposals with some form of data visualization were well received, but due to the need of adding more data visualization components in the future, we have selected the carousel version which allows for multiple options while keeping the interface cleaner and simpler to use.

As stated in the Personas section, we learned that most users (if not all) prefer nighttime or early morning just after waking up to perform their research about the markets.

This led us to decide for a dark, high contrast colour scheme for the MVP (day mode is planned to be tested after the MVP is shipped).

After many trials with different typefaces, I've decided to go for a clear geometric Sans-Serif font: Plus Jakarta Sans.
This font also has a wide range of weights, making it very flexible. Furthermore, it features beautiful ligatures, which gives character to the copy, makes it look more sophisticated and at the same time is perfectly readable in nearly any size.
‍
The colour scheme was selected after a thorough research on the crypto market, but also the underlying psychology of the purple colour, which historically been associated with wealth and royalty, but also with mystery - all these traits can be associated with the meaning of the name of the product and its function. The components use high-contrast colours in order to aid for readability, but also to be informative.

The main data visualization components are centrepieces of the whole functionality of the product, therefore these were designed to be clear and striking in appearance. Also, animated microinteractions were added to the carousel's UI with colour coded feedback for low to high values (a green-yellow-red tricolour gradient).

Due to the impracticality of animating sweeped circles in Figma, the animated mockup (seen below) was made in After Effects.

Returning to the logo design, I have based myself on the single Ionian Column which was selected from the sketches. After scanning it, I vectorized it using Illustrator.
‍
The design was simplified and modernized even further in line icon style, which works quite well also on smaller resolutions - especially in its main application, the iOS icon and Splash screen.

Zen Dots Regular, which is a typeface with a modern style, monospaced and with a serious/technological feel was selected for the logo. It was then paired with the uppercase main typeface used in the UI, which was used in the claim.