PAYMENT PROCESSING CASE STUDYFULL TILT POKER CASHIER REDESIGN
By nature, online poker software contains some of the most complex payment processing user flows, due to the past volatile nature of regulation within the online gaming industry. During the earlier years, convoluted flows involved prepaid phone call cards, telephone verification & changing regulation on age verification & self-exclusion.
Here is my story of how I used UX techniques & tools to uncover new insights, build trust in the UX process & build consensus across divided teams to create an efficient & effective UX architecture resulting in some of the highest seen success rates for merchants with over 64% reduction in deposit failures.
HIGH VOLUME TRANSACTIONS
Full Tilt Poker was a downloadable desktop poker app, built in C++ using the Qt framework. With over 23 million registered users, peaks of 150,000 concurrent users (real & “play money” users), and over 15,000 real-money cash games at any one time, successful, friction free deposit UX was critical to success of the business.
“The goal of a designer is to listen, observe, understand, sympathize, empathize, synthesize, and glean insights that enable him or her to ‘make the invisible visible.’
UX DESIGN LEAD
Before my engagement, the company did not have any UX process in place. Throughout this project I was the sole UX designer responsible for the UX work on this project & introducing the process to the organisation.
I lead the UX work, producing all major deliverables & presenting solutions to the internal client.
Working closely with the payments teams, I analysed trends & data on user behaviour patterns to create requirements & specifications for graphic design & development teams.
SIZE OF THE AUDIENCE
- 23 million registered users
- Peaks of 150,000 concurrent users
- Over 15,000 “real-money” cash games at any one time
INTERNATIONALLY SCALING PAYMENT PROCESSING
Up to this point Full Tilt Poker had been a largely US focused offering. However more recently it began scaling at speed across Europe, the Middle East, Australia, Russia & Asia. A large choice of payment processors needed to be communicated to rapidly expanding geographically diverse customer base, across multiple languages, with different regulatory requirements.
At short notice, processors would change, and user flows for lesser known processors would become part of the UX. Alongside this, enhanced validation such as VBV was introduced during this time, which at its earliest stage caused significant disruption to users.
HIGH RATES OF FAILED DEPOSIT ATTEMPTS & ABANDONMENT
The challenge was to improve the UX of the deposit process for customers overall, but specifically included reducing deposit failure attempts which were frustrating for user and badly affected deposit rates and customer retention.
This meant streamlining the flow of the deposit experience, to reduce friction as much as possible, to evoke trust & credibility in the interface, & guide the user through remediation of errors along the way, through to successful use of deposit on game.
It also required streamlining the wide range of deposit options available, promoting the payment processors with highest potential for success by (legislative) region and the most popular methods, while also making lesser used, but equally important, options easily available.
The UX for this flow would affect payments interfaces for over 23 million registered users. Any changes to the UX/UI were risk critical to the business. Improvements had to be carefully rolled out to minimise impact on existing high value users.
In addition, the flows included increasingly complex regulatory elements such as ID verification for new users, fraud prevention algorithms, self-exclusion functionality & processor variables such as minimum & maximum deposit limits over varying time periods.
LEVERAGING UX TECHNIQUES & TOOLS
The project was high impact & high profile, so the process needed to be thorough.
It involved comprehensive quantitative & qualitative research, several rounds of design with iterative cycles to optimise.
FIRST UNDERSTAND WHAT WE DON’T KNOW
Speaking to stakeholders & team members, I realised that there was a lack of understanding of the range of user cohorts & their needs. Previous to this project, the teams involved had not had any UX input & to date had focused their efforts on high value VIP users only.
These customers were mainly existing users, who would use the software for intense periods of time and were highly motivated to overcome any barriers along the user journey. This resulted in a lack of understanding of the user journey of new & casual users & impact of friction along the onboarding path.
The company lacked a holistic view of their audience & their current user journey which resulted in a disparate UX for users, with inconsistent branding, eroding trust & credibility of the product.
USER JOURNEY MAPPING
STARTING ON THE SAME PAGE
A shift of mindset needed to happen in the organisation.
In order to illustrate the problem to stakeholders I needed to make it easy for them to see the problem.
To illustrate it, I created a screen flow diagram and user journey mapping the journey of the user from discovery to installation of the software, through account creation, to depositing on the site, & finally, successful play.
This was presented to stakeholders & teams so they could quickly grasp the scope of the issue.
“Get in over your head as often and as joyfully as possible.”
– Alexander Isley
GETTING IMMERSED IN THE SUBJECT
LEVERAGING ORGANISATIONAL INSIGHT
The product offered a of the wide range of processors each with their own & requirements for ID, verification, fees, minimum & maximum deposit limits over varying time periods & speed of processing.
There were several flows for verification such as telephone verification, address validation, & ID documentation.
In order to understand the flows, I worked with domain experts to put together a table of the various so that we could track the requirements from the users perspective & map a full user journey through to create a task analysis.
MAPPING PAYMENT PROCESSORS
CAPTURING THE DETAIL
Each individual flow was then mapped.
BUILDING TRUST THROUGH NUMBERS
The company had implemented an analytics system for user payment processing activity. At that time, they could see high levels of deposit failure & abandonment, however did not use the analytics to build a picture from the user perspective.
We worked to to identify most common deposit failure points & user journey flows for deposit failures were examined right through to contact with customer support &resolution.
Depositor cohorts were defined based on behavioural data. These were used to inform Personas later in the process.
Some of the cohorts identified included; initial depositors, recurrent depositors, frequent micro depositors, infrequent lump sum depositors, etc.
LISTENING TO TEAMS & ALLAYING FEARS
One of the biggest impacts of any changes to the UX would be on the Customer Services department.
The department was not usually a part of the design process. This was partly because they were located remotely & in a different time zone. I had contacted them to get some information on deposit failure flows. They became concerned when they learned of the scale of the project & their lack of knowledge up to this point about the forthcoming changes.
I realised that not only had we completely failed to engagement with this important team but also overlooked a valuable source of user data. I arranged a video call with the CS reps & listened to their concerns which were allayed as soon as they felt they had a voice in the process.
From this point we worked closely with the CS department to draw from as a source of data & feedback. In return the department was happy to have a preview of upcoming changes to the payment processing interfaces so that they could prepare further in advance.
INTRODUCING QUALITATIVE TO QUANTITATIVE DATA
BRINGING LIFE TO THE DATA
Now that we had a data based understanding of where we were, it was time to bring this picture to life.
It was clear to me that to shift the mind set in the organisation, stakeholders would need demonstrative evidence of the issues in order to secure buy-in to the necessary changes.
On presenting the results of a heuristic analysis, I pitched the idea of guerrilla (informal) usability testing to test the assumptions & won buy-in from stakeholders.
As a numbers focused organisation, this was the first time the organisation had experienced usability testing. It was important to build trust in the process through transparency.
I worked with the marketing department to prepare demographic data to inform the requirements for participant recruitment.
10 male poker players between the ages of 21-55 were recruited & were interviewed & tested in their regular play environments (generally at their homes). Doing this enabled me to also achieve a “guerrilla style” contextual inquiry & achieve less biased results.
My work included:
- Participant recruitment
- Test preparation – task definition, test set up
- Test facilitation – using TOL (thinking out loud) protocol, observing, prompting where needed etc.
- Follow up -satisfaction & emotional recall testing.
- Reporting to stakeholders
LEARNING FROM OUR CUSTOMERS
Usability testing delivered some really interesting insights into the user journey of the customer, as well as friction along the flow. The biggest revelation of the testing was the impact of that friction along the path & the level of frustration it was causing players was severe. On reviewing the results, there was a shift in perception of the business of the urgency to resolve the issues.
…you see, ‘Cashier’ would be an unfamiliar term to me for where I change my money to chips. I know it’s a name that’s used in casinos but I’m not in a casino. I would have just said ‘Add money to account‘ or ‘Top up your account…
Usability participant Dave’s on struggling to find the payment processing interface.
Did not immediately identify the route to deposit
Found the deposit options grid to be confusing
Felt the deposit process was too long
Said deposit failure message looked like it might be phishing
USABILITY TESTING FINDINGS
A DIFFERENT UX FOR PAYMENTS
The usability testing underpinned to the organisation that while we were a gaming & poker company synonymous with risk, thrill & excitement, we needed to change how we projected our UX when it came to the payment processing piece.
We learned that users’ needs shifted when dealing with payments & they wanted to feel a different sense when interacting with those interfaces.
BUILDING TEAM CONSENSUS
PERSONAS AS A PROCESS
To encourage buy-in from business & tech teams to the concept of personas, I workshopped with them to develop a series of personas.
Enabling them to see how their data & insights were being used, helped not only to build trust & credibility, but also consensus between the teams & broaden their own insights.
The personas were specific to user behaviour in the Cashier & for this purpose, we distilled them into 4 archetypes built on payments behaviour. To simplify, we created 4 “flash cards” with essential data which was relevant to & could be used by non-creative teams such as QA & development. The simplified flash cards meant that the personas were more graspable by all teams, memorable & therefore used more widely.
These variables categorised into experience, frequency of use, & variety of payment methods amongst others motivations.
“How well we communicate is determined not by how well we say things, but how well we are understood.”
DESIGNING FROM OUR USERS’ PERSPECTIVES
Now that we had defined our core issues with the current flow & our target objectives for the new flow, it was time to start defining the vision for the new UX.
To understand our users’ perspective, we also had to look at the conventions & norms which they had been exposed to in other payment processing contexts. We needed to be aware of the patterns & conventions they had become used to & come to expect. This meant looking further afield than just the gaming context.
As well as gaming competitor analysis, wider domain analysis on market leaders such as iTunes & Amazon payments interfaces was equally important as they influence consumer expectations on a global scale.
BUILDING TEAM CONSENSUS
BUILDING TRUST & TRANSPARENCY THROUGH THE PROCESS
I workshopped with the development team to learn about the logic behind the flow. Using rapid sketching and design mock-ups. This created early team consensus & alignment & sparked imaginative ideas & solutions from the team.
This workshop also helped build a sense of ownership for the project across disciplines within the organisation.
DESIGN AS BUSINESS STRATEGY
A solution was needed urgently as the current processor interface was already struggling to accommodate the addition of new processors.
The brief set by the business was to improve the existing deposit screen to accommodate the rapidly growing number of processors & address the relatively high number of deposit failure attempts.
The solution had to reflect not only these needs of the business, but also the technical resources, & most importantly, the needs of our Customers.
Meeting the needs of the business
The solution had to be agile & flexible enough to not only accommodate the growing number of processors, but also enable the backend business logic around them. The logic centred around the display of different amounts of processors per region & the promotion of certain processors over others based on business variables.
Lean & agile tech at speed
Changes to the UI in the payments area of the game were more costly on the backend than changes in the rest of the software. For a rapidly scaling company, technical resources needed to be used efficiently & every change to the UI needed to equate to a win for the business. The solution also had to be efficient & achievable in terms of technical effort & development timescales. The payment processing changes needed to be deployed as quickly as possible as the interface was already struggling to cope with the current payment processor offering.
Meeting the needs of Customers
The goal was to make it as quick & easy as possible for a user to make a successful deposit so that he could get back to enjoy his game. This meant reducing friction along the path & providing graceful error & failure recovery.
DESIGN FOR DESIGN’S SAKE
AVOIDING AGE-OLD DESIGN TRAPS – FOCUSING ON FEATURES USER NEEDS
Overall, it was important to put the goals into context. From a business perspective the UX/UI was critical to business success. Because of this focus, there was a danger of over-designing & over-engineering the solution.
From a user’s perspective the payment processing UX was a means to an end, & we needed to be careful that we did not fall into the trap of running with a redesign for design’s sake.
Changes to the UI in the payments area of the game were more costly on the backend than changes in the rest of the software. For a rapidly scaling company, technical resources needed to be used efficiently & every change to the UI needed to equate to a win for the business.
We needed to ensure we were designing for impact.
LEAN & AGILE SOLUTION – IMPACT AT SPEED
It was during this time that we discovered that significant technical debt had been built up due to resources being focused on the main game area. This severely restricted any major changes to the UI. This included restrictions to the size of the window, the treatment of popups & modal windows within that window & the decision was made to refactor the code to allow faster & more flexible changes in future. A design solution was needed which was technically efficient but addressed the same issues.
Instead of focusing the redesign on the deposit screen itself, the biggest bang for buck was in redesigning the user flows around the screen.
TASK ANALYSIS & USER JOURNEY CREATION
BUILDING TEAM CONSENSUS
I started crafting ideal user flows for each persona type, reducing barriers & effort to accomplish their goals.
The flows contained geotargeted business logic whereby the payment processor with the greatest success rates per region & the most popular (widely used) per region, as well as those preferred by the business, were promoted in the visual hierarchy over the range of alternative processors.
With informed data it was time to move into the design process.
“Any intelligent fool can make things bigger and more complex.
It takes a touch of genius—and a lot of courage—to move in the opposite direction.”
– E. F. Schumacher
DEFINING THE UI
With the background research completed, it was time to start defining the UI.
The earlier developed persona types were used to inform the flows & optimise the interfaces for each persona type.
As the software had a relatively specific graphical style defined already, it was easy to produce high fidelity wireframes from the start with little additional effort. This helped in the crossover to the UX process for the organisation as it made it easier for them to visualise & review the designs.
Wireframes were created, & several permutations were explored, with iterative redesign cycles taking place upon feedback from stakeholders.
After this functional prototypes were developed & reviewed.
DOCUMENTING THE UX
A LEAN & AGILE APPROACH
With many permutations of processors & related messaging, wireframes were designed as templates. Areas were defined where processors & messaging were swapped out as necessary.
AVOIDING LENGTHY DOCUMENTATION
To avoid lengthy documentation, the wireframes were annotated with callouts & notes on function.
UX AS A CROSS ORGANISATIONAL TOOL
Wireframes were integrated into technical specifications used by the development teams. And in parallel the wireframes were stylised by the brand & graphics team. In addition, the wireframes were used by the QA & Customer Support teams as an early indicator of upcoming work.
Note: For the purposes of presentation here, annotations have been stripped out (in such a time sensitive environment, time was not wasted on making annotations look pretty!).
INITIAL REAL MONEY DEPOSITOR
The goal for initial depositor types was to reduce the initial deposit abandonment rate. This meant making it as easy & convenient as possible for users to make a deposit.
As well as improved messaging & a link provided directly after signing up to an account, the 1st cashier screen was pared right back to display 1 to 2 primary payment processors only, with the number of processors displayed determined by region & success of the options. In regions where 2 processors sat at the top most used list & were equally used successfully, for example VISA & Mastercard, the 2 option screen was displayed.
This reduced the burden of choice from the user & promoted payment processors which required less validation actions from users & instant processing.
In regions where only processors with more complex flows were available, improved messaging was communicated on this screen, to guide the user through the process.
The goal for Regular Reloader types – those who regularly deposited using the same payment method – was to provide quick access to the same payment method, displaying relevant information such as deposit limits.
This reduced the burden of choice from the user & enabled him to reload his account as quickly as possible.
DEPOSIT ASSISTANCE CENTRE
IMPROVED FAILURE & ERROR RECOVERY
In order to assist users when payments were declined, a new Deposit Assistance Centre was designed. The goals for this flow was widely informed by the insight provided by the Customer Services team.
Rather than continue as the previous screen & provide users with a single generic error message. the new flow utilised a clever back-end which enabled customised messages to be displayed based on the flow steps previous to this screen. It provided the ability to make error recovery recommendations to users based on their payment method & historical Customer Service insights on failures for those processors. They were also provided recommendations on alternative processors with high potential for success in their region. For example, for credit card payments, the most common failure was error in card number entry – so the new screen messaged users on this potential for X number of attempts, & any failure beyond this, the user was prompted to contact Customer Payments Support directly.
EXPEDITING ACCESS TO PAYMENTS
The biggest change to the payment processing flow was the addition of a Single Click Deposit functionality from appropriate locations in throughout the software. The click-path to access the payment interface was reduced from 5 steps to 1.
This functionality enabled the user to access a previously used payment method(s) directly from the a table without having to navigate to the cashier.
Full Tilt Poker was knows for its secrecy around revenue information. This project was completed without ever knowing the actual numbers for amounts being processed. However a well known point of reference was that at one point Full Tilt Poker was processing up to $1 million per day.
Despite this, the project was recognised as a win for the business, with some of the highest seen success rates for merchants. There was over 64% reduction in deposit failures. The project enabled a lean & agile solution to be brought to users, enabling the tech team to refactor technical debt in the process.
The cashier was developed further in the following years, however it always maintained the designs of these core screens & concepts, making the solution one with longevity.