Case Study
How Allwyn used Piano Analytics to understand what 50,000 player journeys were telling them
Case Study
How Allwyn used Piano Analytics to understand what 50,000 player journeys were telling them
Case Study
How Allwyn used Piano Analytics to understand what 50,000 player journeys were telling them
From data overload to behavioral clarity – at 200 million events a month
Allwyn Entertainment operates one of Europe's most complex digital gaming platforms – six products, one site, and player behaviors that couldn't be more different. Since gaming regulations block Allwyn from using audience targeting or remarketing on platforms like Google, the team can only run untargeted campaigns externally. That makes understanding player behavior on their own platform the only real way to grow. Piano Analytics gave them the infrastructure to do it – and the self-service tools to spread that understanding across 150+ colleagues.
What 50,000 player journeys taught Allwyn about what players actually want
From data overload to behavioral clarity – at 200 million events a month
Allwyn Entertainment operates one of Europe's most complex digital gaming platforms – six products, one site, and player behaviors that couldn't be more different. Since gaming regulations block Allwyn from using audience targeting or remarketing on platforms like Google, the team can only run untargeted campaigns externally. That makes understanding player behavior on their own platform the only real way to grow. Piano Analytics gave them the infrastructure to do it – and the self-service tools to spread that understanding across 150+ colleagues.
What 50,000 player journeys taught Allwyn about what players actually want
Allwyn Entertainment is one of Europe's leading lottery and gaming operators, running the Czech Republic's dominant lottery business as well as the UK National Lottery. It competes with more than 27 online casino providers in the Czech market and serves players across 8,000+ physical branches and multiple digital channels.
All six products share a single platform and a single registration flow – but the players behind them couldn't be more different. Traditional lotteries run three draws a week with physical balls and jackpots in the hundreds of millions. Digital lotteries are drawn by computer every minute with smaller prizes. Scratch cards, sports betting, online casino, and live casino – where real dealers run tables via video – each attract their own audience with their own habits. Every player signs up the same way. After that, they diverge completely.
About Allwyn
About Allwyn
Allwyn Entertainment is one of Europe's leading lottery and gaming operators, running the Czech Republic's dominant lottery business as well as the UK National Lottery. It competes with more than 27 online casino providers in the Czech market and serves players across 8,000+ physical branches and multiple digital channels.
All six products share a single platform and a single registration flow – but the players behind them couldn't be more different. Traditional lotteries run three draws a week with physical balls and jackpots in the hundreds of millions. Digital lotteries are drawn by computer every minute with smaller prizes. Scratch cards, sports betting, online casino, and live casino – where real dealers run tables via video – each attract their own audience with their own habits. Every player signs up the same way. After that, they diverge completely.
Key Results
Key Results
Key Results
Key results
200M+
Events tracked per month (up from 40M in 2018)
150+
Active Piano Analytics users
200+
Custom Boards built for individual users
200+
Automated data validation test cases run daily
3
Core player behavior patterns identified across 50,000 journey combinations
200M+
Events tracked per month
(up from 40M in 2018)
150+
Active Piano
Analytics users
200+
Custom Boards built for
individual users
200+
Automated data validation
test cases run daily
3
Core player behavior patterns identified
across 50,000 journey combinations
Allwyn's platform is built for speed. Most core products take just 1-3 clicks to place a bet – intentionally frictionless, intentionally fast. That efficiency created an unexpected problem: the journey from landing on the site to completing a bet was so short that the team couldn't tell whether a player had decided not to bet or had simply not found what they were looking for.
The team analyzed a single lottery product, tracking 36 key actions across the betting flow (such as which page a player came from, whether they opened a bet ticket, and whether they completed checkout). They found 50,000 distinct sequences across 200,000 bets, with every combination representing a different path through the site. "We started with 40 million events per month. Now we are above 200 million – with 50,000 specific combinations of 36 milestones or events," said Tomáš Rauch, Web Analytics Lead at Allwyn. There was no practical way to act on that volume of variation.
Besides, product managers and brand teams were generating a steady stream of ad hoc data requests. The analytics team was buried in repetitive reporting. And with no ability to use audience targeting externally, there were no shortcuts – understanding what players wanted on their own platform was the only lever available.
Challenge
Challenge
Allwyn's platform is built for speed. Most core products take just 1-3 clicks to place a bet – intentionally frictionless, intentionally fast. That efficiency created an unexpected problem: the journey from landing on the site to completing a bet was so short that the team couldn't tell whether a player had decided not to bet or had simply not found what they were looking for.
The team analyzed a single lottery product, tracking 36 key actions across the betting flow (such as which page a player came from, whether they opened a bet ticket, and whether they completed checkout). They found 50,000 distinct sequences across 200,000 bets, with every combination representing a different path through the site. "We started with 40 million events per month. Now we are above 200 million – with 50,000 specific combinations of 36 milestones or events," said Tomáš Rauch, Web Analytics Lead at Allwyn. There was no practical way to act on that volume of variation.
Besides, product managers and brand teams were generating a steady stream of ad hoc data requests. The analytics team was buried in repetitive reporting. And with no ability to use audience targeting externally, there were no shortcuts – understanding what players wanted on their own platform was the only lever available.
The team stopped trying to count every interaction and started focusing on which behaviors actually influenced outcomes – completions, product choices, repeat visits. By removing every action that didn't indicate intent, the team went from 50,000 combinations to three patterns clear enough to build on.
Solution
Solution
The team stopped trying to count every interaction and started focusing on which behaviors actually influenced outcomes – completions, product choices, repeat visits. By removing every action that didn't indicate intent, the team went from 50,000 combinations to three patterns clear enough to build on.
Finding the patterns that matter
The team worked through the 36 tracked actions and removed the ones that didn't indicate a real decision – logging in, for example, is something every player does regardless of intent. Piano's user-level data model made this possible: by tying behavioral events back to consented player IDs across sessions, the team could see how individual players actually moved through the site rather than inferring behavior from aggregated traffic. What remained revealed three distinct ways players place bets: some go straight to their profile and repeat a previous bet; some browse a page listing all available lotteries and place a quick bet from there; others navigate to a dedicated page for a specific game. These three patterns held consistently enough across the player base to build on.
Personalized analytics at scale
"Early attempts to train colleagues on data interpretation didn't land. Using Piano's custom Boards, the team built a personalized view for each colleague."
Piano's custom Boards gave each colleague their own view of the data at the level of detail they needed. Today, 150+ active users access Piano Analytics, with 200+ Boards built across the organization. Rather than routing every data question through the analytics team, product managers and brand managers can explore independently.
Automated threshold alerts meant the team no longer had to check routine metrics manually. A product manager whose key metric is landing page bounce rate now opens their Board once a week and sees a single number – green or red. If something is wrong, they catch it immediately rather than discovering it days later.
"It's easy to have a lot of configurations done by yourself."
A dual infrastructure built for trust
Piano handles behavioral insight and self-service exploration. An export of 80 selected columns feeds Azure and Power BI for enterprise-level reporting – giving different teams the format that works for them. "These two ways of connecting people with the data are essential for us," Rauch says.
To make sure everyone was working from data they could trust, Allwyn partnered with implementation partner Cross Masters to build an automated validation layer using a tool called Voila. More than 200 test cases run daily, comparing current data against a six-week rolling average and flagging anything outside a ±20% threshold.
Finding the patterns that matter
The team worked through the 36 tracked actions and removed the ones that didn't indicate a real decision – logging in, for example, is something every player does regardless of intent. Piano's user-level data model made this possible: by tying behavioral events back to consented player IDs across sessions, the team could see how individual players actually moved through the site rather than inferring behavior from aggregated traffic. What remained revealed three distinct ways players place bets: some go straight to their profile and repeat a previous bet; some browse a page listing all available lotteries and place a quick bet from there; others navigate to a dedicated page for a specific game. These three patterns held consistently enough across the player base to build on.
Personalized analytics at scale
"Early attempts to train colleagues on data interpretation didn't land. Using Piano's custom Boards, the team built a personalized view for each colleague."
Piano's custom Boards gave each colleague their own view of the data at the level of detail they needed. Today, 150+ active users access Piano Analytics, with 200+ Boards built across the organization. Rather than routing every data question through the analytics team, product managers and brand managers can explore independently.
Automated threshold alerts meant the team no longer had to check routine metrics manually. A product manager whose key metric is landing page bounce rate now opens their Board once a week and sees a single number – green or red. If something is wrong, they catch it immediately rather than discovering it days later.
"It's easy to have a lot of configurations done by yourself."
A dual infrastructure built for trust
Piano handles behavioral insight and self-service exploration. An export of 80 selected columns feeds Azure and Power BI for enterprise-level reporting – giving different teams the format that works for them. "These two ways of connecting people with the data are essential for us," Rauch says.
To make sure everyone was working from data they could trust, Allwyn partnered with implementation partner Cross Masters to build an automated validation layer using a tool called Voila. More than 200 test cases run daily, comparing current data against a six-week rolling average and flagging anything outside a ±20% threshold.
Finding the patterns that matter
The team worked through the 36 tracked actions and removed the ones that didn't indicate a real decision – logging in, for example, is something every player does regardless of intent. Piano's user-level data model made this possible: by tying behavioral events back to consented player IDs across sessions, the team could see how individual players actually moved through the site rather than inferring behavior from aggregated traffic. What remained revealed three distinct ways players place bets: some go straight to their profile and repeat a previous bet; some browse a page listing all available lotteries and place a quick bet from there; others navigate to a dedicated page for a specific game. These three patterns held consistently enough across the player base to build on.
Personalized analytics at scale
"Early attempts to train colleagues on data interpretation didn't land. Using Piano's custom Boards, the team built a personalized view for each colleague."
Piano's custom Boards gave each colleague their own view of the data at the level of detail they needed. Today, 150+ active users access Piano Analytics, with 200+ Boards built across the organization. Rather than routing every data question through the analytics team, product managers and brand managers can explore independently.
Automated threshold alerts meant the team no longer had to check routine metrics manually. A product manager whose key metric is landing page bounce rate now opens their Board once a week and sees a single number – green or red. If something is wrong, they catch it immediately rather than discovering it days later.
"It's easy to have a lot of configurations done by yourself."
A dual infrastructure built for trust
Piano handles behavioral insight and self-service exploration. An export of 80 selected columns feeds Azure and Power BI for enterprise-level reporting – giving different teams the format that works for them. "These two ways of connecting people with the data are essential for us," Rauch says.
To make sure everyone was working from data they could trust, Allwyn partnered with implementation partner Cross Masters to build an automated validation layer using a tool called Voila. More than 200 test cases run daily, comparing current data against a six-week rolling average and flagging anything outside a ±20% threshold.
From reporting bottleneck to strategic function
With routine reporting handled through Piano Boards, Allwyn's analytics team shifted from answering "how many clicked?" to asking "what should we do about it?". Product managers who previously submitted ad hoc data requests now explore independently and flag anomalies themselves.
"It's the shift from reporting to decisions. It's that easy."
Smarter experimentation, less wasted effort
The player patterns that Piano surfaced changed how Allwyn runs experiments. For example, players who always repeated the same bet had little interest in trying other products. Allwyn tested several ways to change that. They ran a Golden Wheel loyalty program designed to reward betting activity with coins that players could spend across other games – but this group engaged with it less than any other segment. They tried adding a promotional banner on the site pushing their usual betting options further down the page – which also didn’t work. The resistance was consistent across every experiment.
So, they changed the approach. Instead of pushing other products before a player had even placed their bet, the team introduced recommendations after the bet was complete – when players were more receptive. Those recommendations were based on what each player type had actually responded to in the past. For some players, the best next offer turned out to be the same game they'd just played. "We can now better understand how the specific groups of players are behaving and give them what they want," says Rauch. "It's not about finding the unicorn. It's about working out where the unicorn isn't, so you stop investing in segments that won't respond, and focus on the ones that will."
"The biggest impact for us internally is that we understand the players more."
Looking ahead
Allwyn is working toward a satisfaction metric built from behavioral data rather than survey responses – one that reflects how players actually experience the platform, not just what they say about it afterward. The team is also continuing to automate more of the day-to-day measurement work, freeing up time for the analysis to do more strategic, meaningful work.
Results
Results
From reporting bottleneck to strategic function
With routine reporting handled through Piano Boards, Allwyn's analytics team shifted from answering "how many clicked?" to asking "what should we do about it?". Product managers who previously submitted ad hoc data requests now explore independently and flag anomalies themselves.
"It's the shift from reporting to decisions. It's that easy."
Smarter experimentation, less wasted effort
The player patterns that Piano surfaced changed how Allwyn runs experiments. For example, players who always repeated the same bet had little interest in trying other products. Allwyn tested several ways to change that. They ran a Golden Wheel loyalty program designed to reward betting activity with coins that players could spend across other games – but this group engaged with it less than any other segment. They tried adding a promotional banner on the site pushing their usual betting options further down the page – which also didn’t work. The resistance was consistent across every experiment.
So, they changed the approach. Instead of pushing other products before a player had even placed their bet, the team introduced recommendations after the bet was complete – when players were more receptive. Those recommendations were based on what each player type had actually responded to in the past. For some players, the best next offer turned out to be the same game they'd just played. "We can now better understand how the specific groups of players are behaving and give them what they want," says Rauch. "It's not about finding the unicorn. It's about working out where the unicorn isn't, so you stop investing in segments that won't respond, and focus on the ones that will."
"The biggest impact for us internally is that we understand the players more."
Looking ahead
Allwyn is working toward a satisfaction metric built from behavioral data rather than survey responses – one that reflects how players actually experience the platform, not just what they say about it afterward. The team is also continuing to automate more of the day-to-day measurement work, freeing up time for the analysis to do more strategic, meaningful work.
... if you're dealing with complex customer journeys and high data volume:
Key takeaways
... if you're dealing with complex customer journeys and high data volume:
Key takeaways
Start with the decision, not the data
Before pulling any report, ask what decision it needs to support. Rauch's team learned to ask that every stakeholder who came to them with a data request.
Accept that everything isn't for everyone
The most valuable insight isn't always finding the audience that responds. It's identifying the ones that don't – and stopping the investment there. Eliminating low-probability conversions frees your team to focus on segments that are actually worth pursuing.
Build trust before expecting action
Data only drives decisions when people trust it. Allwyn built that trust two ways: technically, through 200+ automated daily validation tests; and relationally, by working alongside product and brand teams rather than simply reporting to them.
Empower, don't gatekeep
When 150+ colleagues can access their own Piano Boards and answer their own questions, the analytics team becomes a strategic resource and spends less time answering basic data requests.
Document everything
Allwyn tracks changes, revisits old assumptions, and maintains detailed records of what's been tested and why. "Have good documentation. This is the key because everything gets broken all the time."
Iterate constantly and celebrate small wins
There was no single breakthrough moment. The progress came from continuous testing, regular refinement, and a willingness to learn from experiments that didn't work. "It's small chunks of work that build to something bigger.”
Start with the decision, not the data
Before pulling any report, ask what decision it needs to support. Rauch's team learned to ask that every stakeholder who came to them with a data request.
Accept that everything isn't for everyone
The most valuable insight isn't always finding the audience that responds. It's identifying the ones that don't – and stopping the investment there. Eliminating low-probability conversions frees your team to focus on segments that are actually worth pursuing.
Build trust before expecting action
Data only drives decisions when people trust it. Allwyn built that trust two ways: technically, through 200+ automated daily validation tests; and relationally, by working alongside product and brand teams rather than simply reporting to them.
Empower, don't gatekeep
When 150+ colleagues can access their own Piano Boards and answer their own questions, the analytics team becomes a strategic resource and spends less time answering basic data requests.
Document everything
Allwyn tracks changes, revisits old assumptions, and maintains detailed records of what's been tested and why. "Have good documentation. This is the key because everything gets broken all the time."
Iterate constantly and celebrate small wins
There was no single breakthrough moment. The progress came from continuous testing, regular refinement, and a willingness to learn from experiments that didn't work. "It's small chunks of work that build to something bigger.”
Start with the decision, not the data
Before pulling any report, ask what decision it needs to support. Rauch's team learned to ask that every stakeholder who came to them with a data request.
Accept that everything isn't for everyone
The most valuable insight isn't always finding the audience that responds. It's identifying the ones that don't – and stopping the investment there. Eliminating low-probability conversions frees your team to focus on segments that are actually worth pursuing.
Build trust before expecting action
Data only drives decisions when people trust it. Allwyn built that trust two ways: technically, through 200+ automated daily validation tests; and relationally, by working alongside product and brand teams rather than simply reporting to them.
Empower, don't gatekeep
When 150+ colleagues can access their own Piano Boards and answer their own questions, the analytics team becomes a strategic resource and spends less time answering basic data requests.
Document everything
Allwyn tracks changes, revisits old assumptions, and maintains detailed records of what's been tested and why. "Have good documentation. This is the key because everything gets broken all the time."
Iterate constantly and celebrate small wins
There was no single breakthrough moment. The progress came from continuous testing, regular refinement, and a willingness to learn from experiments that didn't work. "It's small chunks of work that build to something bigger.”
Empower every team to understand and influence customer behavior.
Platform
Industries
Company
Empower every team to understand and influence customer behavior.
Platform
Industries
Company
Empower every team to understand and influence customer behavior.
Platform
Industries
Company