— Improving Feature Adoption.
- Optimisation of an existing feature.
- Successful split test over multiple rounds.
- 160% increase in bet slip adds.
- Client
- Sky Bet Italia
01/ Summary
- — Services
- Split testing
- Conversion optimisation
- User-interface design
- Iterative design
- Improving design maturity
- User research
- — Deliverables
- Multiple designs
- Pushed the implementation of Google Optimise for split testing
- The first split test at Sky Bet Italia
- — Outcomes
- 118% increase in traffic on the previous busiest day
- 145% increase in overall traffic to subject pages
- 160% increase in users adding bets to their bet split from the subject pages
- 137% increase in users placing bets from the subject pages
- A noticeable improvement in design maturity
02/ Understanding the problem
What is a Coupon?
In 🇮🇹 Italy, Coupon is what we call Accumulators in the U.K. and are sometimes referred to as "Accas". It is one of the most popular types of football bets.
An Accumulator bet combines bets from multiple games into one larger bet to create more significant potential winnings. Examples of markets are "Both Teams to Score" or "First Goalscorer".
Low adoption rates
After designing and launching the Coupon feature for our users, the team realised we were experiencing low amounts of traffic.
The adoption rate was extremely low compared to other newly launched features. The expectation was much higher.
Benchmarking low adoption
I wanted to benchmark what it meant to have a low adoption rate compared to other product features.
Of course, our expectation was a much higher adoption rate, but I also needed to know what a high adoption rate would be.
03/ The solution
Adding more tracking
- From which pages do users add bets?
- What is the standard rate of adoption for features?
- From which pages do users place bets?
Looking for trends
Using Google Analytics, I began to identify trends in usage and saw a massive spike at weekends. Next, I wanted to understand Coupon's everyday use cases, when users were likely to use them, and how.
Surveying users
- What do users call the Coupon feature in their own words?
- How would they describe it to their friend?
- How often do they use Coupon on other betting websites?
- Does our offer differ from what they're used to seeing?
- What would they change or improve?
Survey results
From the survey, a few things were obvious. First, users did not understand the offering we were displaying or couldn't find it on the website.
So we began to focus on improving two areas: increasing the feature's prominence and adding detail to the page to explain what a Coupon is.
Split testing
No part of Sky Bet Italia had ever been split tested. So this adoption problem felt like a fantastic opportunity to implement a split-testing tool.
We decided on Google Optimise as it was easy to use and fit nicely with our tech stack.
Three challenger designs, one original
The focus of the designs was prominence and explanation.
🔵 Original Design
Theory
Having the familiar name of Coupon would attract users.
In addition, only showing a few markets would mean that those interested in seeing more would use the expansion button.
🟠 Design 1
Theory
Some of the surveyed users referred to this feature as "Filtri Rapidi", which translates to "Quick Filters".
By using language that was familiar to the user, it would improve adoption due to recognition.
🟢 Design 2
Theory
By having a grey box around the component, I felt it would draw more attention due to the
Von Restorff Effect (making one item stand out).
🟣 Design 3
Theory
Similar to design 2, I used language drawn from the survey with users. For example, "Multipla Rapida" translates to "Quick Multiple".
A few other key differences in this design are the use of the yellow icon, the bold header, and the descriptive text underneath the feature.
The descriptive text translates to: "Events and bets selected for you, the fastest way to build your multiples".
Pushing to 40% of users
I pushed the challenger designs to 40% of the user base, which was 10% of the user base per design, including the original.
To reach statistical significance, we ran the test for three weeks, which allowed us to smooth for weekend spikes as previously identified.
- 🔵 Original Design
- 🟠 Design 1
- 🟢 Design 2
- 🟣 Design 3
Merging the high performing-designs
Design 2 and Design 3 performed the best. I merged the best parts of these to create Design 4.
The results
The test was concluded after three weeks as Design 4 had reached significance.
It was shown as overwhelmingly successful and was thus released to 100% of the user base. We kept monitoring its performance through Google Optimiser.
04/ Outcomes
The impact over three weeks
- ✅ 118% increase in traffic for the previous busiest day
- ✅ 145% increase in overall traffic to all coupon pages
- ✅ 160% increase in users adding to their bet slip from coupon pages
- ✅ 137% increase in users placing bets from coupon pages
- ✅ Significant increases in the margin as accumulators are typically high-margin bets
After understanding potential reasons why the feature wasn't gaining a high adoption rate through user research, I created three designs to challenge the original.
Finally, I took the two highest-performing designs and merged their best features to create Design 4.
To better understand the shortcomings of the newly launched feature, I surveyed our users.
As a result, we gained valuable information about how users currently use this feature, what they call it, when it's used, and on which other services the users use it.
Available from January 2027
Looking for a product, service or design systems leader?
I'm open to product design, service design, design systems and design leadership roles. Remote anywhere, or hybrid across the UK.