10th August, 2026
Exploring Text-Based Transaction Interfaces
Keep In Mind
For the purpose of this exploration I will be using the Opay app and trying to improve it's specific user-flow since it's the one I'm most familiar with.
Transferring funds to accounts can sometimes be annoying, you start the process and you are met with all of these infuriating "confirm" screens; "confirm bank account", "confirm recipient", "confirm transfer amount", "confirm your identity to send".
It's genuinely annoying, sometimes you're in a hurry and just want to quickly send money to a merchant you've just bought from without having to deal with all that friction.
I had one of those days a while back, I opened the bank app to send money to a merchant and as usual I was met with a cascade of confirm screens, and like most people I blindly spammed the "next" button; I just wanted to get this over with as quickly as possible and leave the seller's store.
When I was done with that particularly unpleasant encounter, I decided to look into ways to make the experience better.
I had two goals;
a. Shortening the flow by redesigning all the screens involved, optimising for ease & speed.
b. Find a new way to present the amount that's about to be sent without using numbers.
Overview
What a Fixation is
A fixation is the short moment when the eyes stop moving so that the brain can process the information it sees. More or longer fixations usually means that the information the brain is decoding is more complex.
For the first goal I decided to cut out entirely the first “Confirm Recipient” screen.
My rationale is simple, confirm screens are by design meant to slow users down, i.e force you to show restraint and actually look through the details, but doing so requires willpower a resource that we have a limited amount of, so the longer the flow is the more willpower it would take to get through, and sometimes that reservoir is already low or even empty before the app is opened (like say a user who has had a particularly hectic day at the market and just wants to send funds to a merchant and go home) it's why sometimes we enter that autopilot state where we just blindly tap the next button without even looking at the details carefully.
To help fix this I decided to shorten the user-flow so the steps become less demanding and users end up completing the process faster while also using up less of that reservoir in the process.
First Objective
OPAY'S CURRENT FLOW
As for the second goal I wanted to do a bit of research before trying anything new and it turns out that using numbers to represent the amount that’s about to be sent might not be the best option because according to the research, reading numbers is way harder for the human brain to do compared to letters1
Second Objective
According to the research we make around 2.5 times more fixations when reading short numbers compared to short words and that difference becomes even larger the longer the numbers get, as much as 7x in some cases.
So! In conclusion;
A shorter flow - reduces the likelihood of a user entering an autopilot state.
Written amounts - make the critical information easier to consciously verify.
A swipe-to-send gesture - requires a more deliberate physical commitment than a tap.
Biometric/PIN - remains the actual authorization mechanism.
In the end this was an exploration and wasn't widely tested, so I can't say that these ideas would scale across every user or financial product. But based on the existing research, and the prototypes I designed, I believe this approach has the potential to reduce cognitive load while making transaction flows more engaging.
Summary
Saccades was also measured and we need 3x more saccades when reading short numbers and as much as 9x more when reading longer numbers compared to text.
What are Saccades
Saccades are rapid jumps our eyes make between points. We do this because unlike words that the brain processes as a whole, numbers are broken up into chunks which the eyes then move through before reaching the end of the number.
Lastly fixation duration and saccades amplitude was measured, long numbers have about a 20% increase in duration when compared to similar words. There is also a .83 character drop in saccades amplitude when reading long numbers too.
All of this suggests that users are likely to benefit from it if the confirmation screens used text as the main medium and not numbers, but the implementation approach was important though.
Lastly, fixation duration and saccades amplitude was measured, long numbers have about a 20% increase in duration when compared to similar words. There is also a .83 character drop in saccades amplitude when reading long numbers too.
All of this suggests that users are likely to benefit from it if the confirmation screens used text as the main medium and not numbers, but the implementation approach was important though.
Think of it like this. We need to break numbers up into chunks before we can understand them, the longer the number the more the chunks. Any time our eyes move between chunks that’s saccades, when our eyes stop on a chunk and starts processing the numbers within it that’s fixation.
SHOTS OF THE NEW FLOW
At first I used a text only approach and when I ran informal tests with five people, they said it felt too different and that despite numbers being harder it was more familiar, so I iterated and came up with a hybrid instead, now the UI has both text and numbers, although the numbers were made smaller so as not to be the main focus.
I also replaced the “Transfer” button with a “Swipe-To-Send” interaction, this is because unlike a tap which can become reflexive, a swipe needs a continuous gesture and a higher level of intentional engagement and commitment. The swipe doesn’t replace authentication though, users would still need their PIN or biometric after.
Citations
OPAY'S CURRENT FLOW
I also made certain key information about the users current position in the flow bolder & persistent the whole time so that the user always knows. Some key UI elements where also repositioned.
At a Glance
TL;DR
Started from a real frustration: too many confirm screens on money transfers lead to blind, autopilot tapping
Cut the flow and tested representing amounts as words instead of digits, backed by existing eye-tracking research on numbers being harder to process than text
Ran informal testing with 5 people, a text-only version felt too unfamiliar, iterated to a hybrid of text and smaller numbers, added swipe-to-send for more deliberate commitment
Role: Solo — independent design exploration
Type: Concept exploration
Focus: Reducing cognitive load and friction in Opay's transaction confirmation flow
Exploring Text-Based Transaction Interfaces
10th August, 2026
At a Glance
Role: Solo — independent design exploration
Type: Concept exploration
Focus: Reducing cognitive load and friction in Opay's transaction confirmation flow
TL;DR
Started from a real frustration: too many confirm screens on money transfers lead to blind, autopilot tapping
Cut the flow and tested representing amounts as words instead of digits, backed by existing eye-tracking research on numbers being harder to process than text
Ran informal testing with 5 people, a text-only version felt too unfamiliar, iterated to a hybrid of text and smaller numbers, added swipe-to-send for more deliberate commitment
Overview
Transferring funds to accounts can sometimes be annoying, you start the process and you are met with all of these infuriating "confirm" screens; "confirm bank account", "confirm recipient", "confirm transfer amount", "confirm your identity to send".
It's genuinely annoying, sometimes you're in a hurry and just want to quickly send money to a merchant you've just bought from without having to deal with all that friction.
I had one of those days a while back, I opened the bank app to send money to a merchant and as usual I was met with a cascade of confirm screens, and like most people I blindly spammed the "next" button; I just wanted to get this over with as quickly as possible and leave the seller's store.
When I was done with that particularly unpleasant encounter, I decided to look into ways to make the experience better.
I had two goals;
a. Shortening the flow by redesigning all the screens involved, optimising for ease & speed.
b. Find a new way to present the amount that's about to be sent without using numbers.
Keep In Mind
For the purpose of this exploration I will be using the Opay app and trying to improve it's specific user-flow since it's the one I'm most familiar with.
OPAY'S CURRENT FLOW
First Objective
For the first goal I decided to cut out entirely the first “Confirm Recipient” screen.
My rationale is simple, confirm screens are by design meant to slow users down, i.e force you to show restraint and actually look through the details, but doing so requires willpower a resource that we have a limited amount of, so the longer the flow is the more willpower it would take to get through, and sometimes that reservoir is already low or even empty before the app is opened (like say a user who has had a particularly hectic day at the market and just wants to send funds to a merchant and go home) it's why sometimes we enter that autopilot state where we just blindly tap the next button without even looking at the details carefully.
To help fix this I decided to shorten the user-flow so the steps become less demanding and users end up completing the process faster while also using up less of that reservoir in the process.
I also made certain key information about the users current position in the flow bolder & persistent the whole time so that the user always knows. Some key UI elements where also repositioned.
Second Objective
As for the second goal I wanted to do a bit of research before trying anything new and it turns out that using numbers to represent the amount that’s about to be sent might not be the best option because according to the research, reading numbers is way harder for the human brain to do compared to letters1
According to the research we make around 2.5 times more fixations when reading short numbers compared to short words and that difference becomes even larger the longer the numbers get, as much as 7x in some cases.
What a Fixation is
A fixation is the short moment when the eyes stop moving so that the brain can process the information it sees. More or longer fixations usually means that the information the brain is decoding is more complex.
Saccades was also measured and we need 3x more saccades when reading short numbers and as much as 9x more when reading longer numbers compared to text.
What are Saccades
Saccades are rapid jumps our eyes make between points. We do this because unlike words that the brain processes as a whole, numbers are broken up into chunks which the eyes then move through before reaching the end of the number.
Think of it like this. We need to break numbers up into chunks before we can understand them, the longer the number the more the chunks. Any time our eyes move between chunks that’s saccades, when our eyes stop on a chunk and starts processing the numbers within it that’s fixation.
Lastly, fixation duration and saccades amplitude was measured, long numbers have about a 20% increase in duration when compared to similar words. There is also a .83 character drop in saccades amplitude when reading long numbers too.
All of this suggests that users are likely to benefit from it if the confirmation screens used text as the main medium and not numbers, but the implementation approach was important though.
At first I used a text only approach and when I ran informal tests with five people, they said it felt too different and that despite numbers being harder it was more familiar, so I iterated and came up with a hybrid instead, now the UI has both text and numbers, although the numbers were made smaller so as not to be the main focus.
I also replaced the “Transfer” button with a “Swipe-To-Send” interaction, this is because unlike a tap which can become reflexive, a swipe needs a continuous gesture and a higher level of intentional engagement and commitment. The swipe doesn’t replace authentication though, users would still need their PIN or biometric after.
Summary
So! In conclusion;
A shorter flow - reduces the likelihood of a user entering an autopilot state.
Written amounts - make the critical information easier to consciously verify.
A swipe-to-send gesture - requires a more deliberate physical commitment than a tap.
Biometric/PIN - remains the actual authorization mechanism.
In the end this was an exploration and wasn't widely tested, so I can't say that these ideas would scale across every user or financial product. But based on the existing research, and the prototypes I designed, I believe this approach has the potential to reduce cognitive load while making transaction flows more engaging.
SHOTS OF THE NEW FLOW