Taking small steps and getting big results at a start up
Edit: from 2017–2018 I worked for Pelican as their Product Designer. You’ll learn more about the app below, but if you’re thinking about starting to trade for real, think twice! Between 74–89% of retail investors lose money when trading CFDs. You should consider whether you understand how CFDs work, and whether you can afford to take the high risk of losing your money.
Pelican holds a unique regulatory license, allowing them to provide trade execution and peer-to-peer messaging on the same platform, making trading accessible for everybody. Whether someone’s serious about making big money, or just wants to see what all the fuss is about, Pelican helps find the very best traders to follow and groups to join while keeping track of your trading journey, all in an open and transparent format. No other trading app does all this.
I worked with a small team of developers, stakeholders, investors and founders to make improvements to the app. In 18 months this is what I learned.
18 months ago the app was three years old and struggling to retain users while nurturing growth. Features were in abundance, but there was no thought to functionality or accessibility. It also became increasingly unclear who the product was for and crucially, what they wanted. The app had been built mirroring existing desktop solutions with little to no thought to the mobile experience. It was time to ask serious questions and use three years of untapped data to optimise and move forward.
A trader at any level, is still a trader
One of our biggest challenges was catering to the needs of both the beginner and expert trader in a single app. Seasoned traders (the experts) were reluctant to escape the comfort of their desktop application, powered with seemingly infinite statistics and just as many screens to match. They wanted as much information as they could get, no matter the cost of legibility. Where as our beginner users wanted a more educational approach, eager for answers to questions like…
“Who’s the best trader?”
“How do I know if they’re any good?”
“How do I copy them?”
Pelican had the answers and all the statistics, we just needed to develop a language to make them accessible.
Identifying the experiences
For every page, feature, detail, we asked ourselves…
“What is it supposed to do?”
“Does it really need to be there?”
“Could it do it better?”
When looking at some of the main experiences we saw that…
First impressions count
Users arriving on the app for the first time were bombarded with too much technical jargon to know how to continue. Data showed vast numbers of users going down the wrong path, dropping off and not returning.
Users were asked to set up either a live or demo account. Live accounts enabled you to trade with real money and required a lengthy KYC check as a prerequisite. Essential to trade with real money, but not to use the app and explore the same features with demo money. Of the users who choose to setup a live account, only 12% successfully finished. Live accounts are an important revenue stream for Pelican, but were massively ineffective when presented at the start of the journey. We were losing users before we even got them.
For the new users that made it past our sign-up process, the first screen they saw was a static list of text. Simply a list of markets you could trade. While trading is the core of the app, a list of text isn’t going to get users coming back for more.
Too much of a good thing can be bad
The Activity Feed showed trade signals from users you follow, the My Messages area allowed you to access the groups you joined, and the Discover page is where you found the best traders and groups to copy. All these were crammed into one place on the bottom navigation bar.
The feed required a fair amount of engagement and understanding. Without initially being populated with followers, it required users to not only understand the concept of having to follow other users to see updates, but it also required them to understand each stage of the trade as it appeared. With multiple trade updates from multiple users all going on the Feed, it soon became overly complex and cluttered.
Messages were not too different, lots of content without structure. Arguably the biggest USP of the app was Discover. It populated every page and was the only source of new and exciting groups and traders. Data showed that despite the Discover section placement at the top level of the navigation structure, a minimal percentage of users even viewed this page and even less were interacting with the content. The Discover page was a static list determined by the business, only editable by a developer and rarely updated. Profiles and groups also lacked calls to action, a user had no way of knowing how good that trader was or what they traded. There was no reason to follow anyone.
At the heart of everything was Pelican’s trade experience. Data showed that after several attempts users understood the basic steps to open a trade, but constantly got caught out making mistakes due to clunky interactions. Reliability for more experienced traders was an issue, and basic education for beginner traders was missing.
How it works is just as important as how it looks
Anyone investing in anything needs to know exactly what’s happening with their money, at all times. Data revealed sporadic drops in our market price stream on the Positions page. This adversely affected the time it took users to see their live balance and take action on trades while the market was at a certain price. This continued to be one of the biggest sources of irritation for our users. Perhaps the most basic page, simply a snapshot of all your trades, was one that our users spent the most time looking at.
Empowering users is key
As we previously saw on the Discover section, Profiles and groups were under-used and ineffective. The information we had available on users and groups provided no hint for further interactions. Pelican’s unique selling point was the ability to find the best traders to follow and groups to join. With no performance figures or transparency this was impossible.
Taking our understanding of the existing product, we decided to strip it down to its very core, orienting the entire product around some of its most used features…
Get our users into Pelican, help them find what they’re looking for, and have them interacting with content as quickly as possible.
Move the activity feed to the home page so our users can stay up-to-date with their friends and followers while providing clear information which they can understand at a glance and at their leisure.
Allow users to discover content on their own time. Allow them to find someone at the top of their game and see exactly why a group is worth joining. All of which is constantly updating.
Clear and precise actions to help our users open, edit, and close trades with perfect execution.
Allow users to stay on top of their trades with market prices that never miss a movement.
Groups and accounts
Clear and concise information that help our users quickly tell why they should follow someone or join a group while keeping track of their own performance.
These experiences formed our base for rebuilding the information architecture and making sure the most important features were always obvious and emphasised.
As work began on redefining the product, we now had the chance to confront and properly challenge not only Pelican’s foundation, but also that of mobile trading itself.
A closer understanding
Designing how these experiences looked was fun, however it was also our initial downfall. With the internal excitement around the project, scope for features kept increasing and we found ourselves in long unwinding debates around potential ideas and enthusiastic suggestions for very lofty ideas. A lot of time was spent with stakeholders refining pie-in-the-sky features which soon become completely out of scope and only ended in disappointed investors and developers on the brink of exhaustion.
With such a small team the point was to answer problems with small manageable solutions we could release, test and repeat.
To cut down the time wasted on over-design, we asked ourselves…
“What can we design, with what we have, to improve each experience by 1%?”
If aiming for 100% wasn’t working, then improving 1% here, and 1% there would soon ad up.
Adopting this approach became our mantra. All of a sudden we were prototyping, testing and delivering quickly. The business, users, stakeholders and developers were all seeing the benefit.
Information you could read
As we know, seasoned traders expect a certain level of information when they look at a trade. It became clear that while beginners wouldn’t know or care about the majority of this information, we could vastly improve the way they consumed it to make it more understandable. Skipping all the jargon, on the most basic level a user simply wants to know if the trade was in profit or loss and therefore worth copying.
This resulted in a careful consideration of how we represented each trade. We decided to allow more room for each trade card, which increased legibility and represented key information with colour, not just figures. Simply put, green trade meant profit and red meant loss. Our users could quickly hunt down the best trades and spend more time seeing why they were the best. It was important to not take away information, but use it more efficiently.
For the rest of the application this attitude of clarifying and simplifying became another underlying principle. Rather than over-communicating we would focus on always trying to radically simplify, making information you could read.
Explore and discover
We dragged the Discover page into its own space on the bottom navigation, giving users vital quick access to the very best traders and groups on the app. The goal was to create a place that continually changed and updated. Our data showed that the one thing all our users understood and wanted were leaderboards. Users wanted to be at the top, to see who was at the top, while doing anything to get there. This was exactly what we wanted.
Lots of easy, bite-size leaderboards gave everyone a quick way to find what they wanted in a variety of ways. Users could discover who was doing the best today, this month as well as over all time. We surfaced the best mentors and educators so beginners had traders they could contact openly and learn from. We even exposed who was on a Live Account, trading real money and risking it all. Discover was soon becoming the reason you opened the app.
But it was not only important to expose the traders and groups doing well, but also to let our users know why they were doing well. The more we could force our users to ask “why”, the more they would use the app to find out.
Who and why
As I mentioned, to populate areas of Pelican you needed to follow users and join groups, but first you had to ask the “why” part. To answer the why, we leveraged our trade data to provide stats per group and profile. Basic but essential.
The result was exposing stats any beginner or pro could understand in a defined area recognised across the app. How much is this person or group making now and how much have they lost in the past. Users could finally dig into a trader or group’s profitability, points, and trade frequency. A concept so obvious it was a no-brainer to implement. We even made awards and badges to add visual elements to answering those questions, but more on that later.
While the Discover page provided competitive content, data showed users really only wanted to see the content relevant to them. The best trader who didn’t trade crypto was useless to a user wanting to buy Bitcoin.
We found a user was 38% more likely to open a trade if they were in a group or following someone, and once you opened a trade the drop off rate was 55% less over seven weeks. So when a user created an account we decided to ask them what they were interested in, before automatically following users and joining groups which matched their interests. They were instantly half way down the funnel able to interact with content without having to search for it.
We also decided to completely remove the sign up for a Live account option on the login screen. Users were more likely to take that step after they experienced the app. The best thing we could do was get users in, as quick as possible, interacting with content. Stripping down and streamlining those key pages proved key.
While an optimal frontend experience is fundamental, users made it clear nothing mattered more than speed and performance. Over time we made sure to continually enhance expensive calls to APIs and reduce data drop-offs in order to improve stream updates and user experience.
With every release we made sure to add equal measures of new features, bug fixes, enhancements and speed updates. Ultimately this led to more activity, which was great for the business, and provided more data we could capture, only helping inform frontend design decisions. Slowly our process became self sufficient.
Small but powerful
Sometimes the smallest experiences can be the most effective. A great example of this is Badges. We implemented badges as a reward system to encourage activity.
Trading is all about numbers and graphs, so when the opportunity arose to add something visual we all got very excited. While the implantation of badges could have escalated, it was important to maintain our 1% philosophy. We built it in a way to ship quickly while having maximum effect. We hid the badges until a user had made profits on trades, which not only helped users to quickly see if a trader is good, but increased trade activity through small incentives, i.e. gamification.
We also rewarded users for other activities like commenting, copying and liking trades, causing increases in activity in other areas of the app. Simple, lightweight, effective and fun.
Often designers can be tempted to do a re-skin of the product they’re working on. After tinkering away out-of-hours designing a dream version of the app, we decided that the benefits of a refined user interface were absolutely worth it. Together we began to rethink our brand image as well as who the product was for. In the past, Pelican had suffered from little to no brand identity. Nothing made it stand out or appeal to its target audience. This made the business reluctant to spend money on advertising and brought about a lack of confidence in the product.
The re-skin not only created excitement within the team, but it also improved legibility through type hierarchy and sequenced styles. We added colour to give direction and help with decision making. We also built reusable components to help prioritise developer time and add new features to multiple pages quickly. We launched a brand new website and accompanying marketing videos. The re-skin arguably had the most effect on the business. It not only gave us confidence, but it gave our users confidence that they were part of something cool and expanding. It was something they saw change over time and something that they could be proud of.
While it felt frustrating at the time, the re-skin took several months to complete. However staggering its release on a per page, per component or per feature basis was essential. Not only was it more manageable for our developers, but it let us gather crucial user feedback which went on to inform further design decisions. This was fundamental to effectively implement such wide reaching changes and ultimately create the best, most effective user interface for our users.
Just the start
When we started this project there were no big meetings where we outlined a grand vision on post-it-notes. The process was developed over time. It was tough, dirty and we made mistakes. Arguably some would say that if we paused for creative sprints and bean-bag blue-sky-thinking sessions, we could have been more effective. But not every business has that luxury. There’s a trust that needs to be formed with stakeholders and investors. Big changes take time. To begin with, selling those changes were equally as hard as implementing them. Pelican did solve problems, it just needed help to be more efficient in practical ways that kept the business going and kept their existing users happy.
After eighteen months the data clearly shows improvements in all the changes we made, but naturally any change only exposes new problems. Pelican is far from finished. The team is actively iterating, responding to user feedback, and designing new experiences.
Over eighteen months our team unearthed Pelican’s unique selling points and gave each of them a space to evolve. We developed a process that was effective for us. We got better at releasing faster, on time, meeting expectations while improving all areas of the app equally. We learnt to ask the right questions, change what wasn’t right and stop when we lost sight of our shared goal. Every improvement opened a new door and is only continuing to do so.