Paddy Power - Beat the Drop Cross Sell
Project Context
Beat the Drop was one of Paddy Power's most popular free-to-play games — tens of thousands of daily active users, a loyal community, and strong retention metrics.
The brief was straightforward: introduce cross-sell functionality to convert free players into active bettors.
The real challenge was doing it without disrupting the product people had come to trust.
The Core Problem
The risk wasn't technical. It was relational.
Beat the Drop's user base had a strong identity around the free, low-pressure game format. Introducing betting prompts too aggressively could damage retention, devalue the product, or feel like a bait-and-switch.
The question wasn't how to add cross-sell. It was where and when it could live without breaking the thing that made the game valuable in the first place.
Approach
1. Ground the brief in real user sentiment
Rather than relying on internal assumptions, I went directly to where the community lived — Facebook groups and Reddit threads dedicated to the game.
I participated in discussions around new features to surface organic sentiment. The feedback was broadly positive about cross-sell in principle, but there were clear signals: relevance mattered. Generic betting prompts would land badly. Contextual, moment-specific prompts had a chance.
This shaped the entire design direction.
2. Analyse how competitors handled it
I reviewed cross-sell execution across Bet365, William Hill, BetBull, and DraftKings.
The pattern was consistent: the strongest approaches used subtle, contextually timed prompts — tied to live events, surfaced at natural pauses, and framed as value rather than interruption.
Disruptive overlay patterns correlated with negative user response. That ruled out the most obvious implementation routes.
3. Align stakeholders before designing
I brought together the squad, Product Owner, Marketing, and Trade in a cross-office workshop in Porto.
The goal wasn't ideation for its own sake — it was alignment on strategy before any screens were drawn. We mapped what the business needed, what the user would tolerate, and where those two things could meet.
Going into design with that clarity meant fewer revision cycles later.
Strategic Decisions
Protect the core journey
The main game experience remained untouched. No cross-sell prompts during active play. No interruption to the primary loop.
This wasn't a concession — it was the strategic call that made everything else viable.
Activate at the moment of selection
The intervention point was post-selection: the moment a user had committed to their choice but before they moved on.
At that point, intent was high, attention was available, and the cross-sell could feel additive rather than disruptive — surfacing related odds that let users extend their engagement on their own terms.
Skip wireframes, go direct to high-fidelity
Given my deep familiarity with the product and the trust already established with the squad, I moved straight to high-fidelity design.
Wireframing would have added overhead without meaningful discovery value at this stage. The team could respond more usefully to something that felt real.
Delivery Under Constraint
The primary constraint was confidence. The product had a large, vocal user base. Getting this wrong publicly — in a game with an active community — carried real reputational risk.
Rather than a staged rollout or internal UAT, I worked with the experimentation team to design an A/B test that would give us statistically meaningful signal before full deployment.
We ran a 50/50 split over 14 days — approximately 400,000 sessions — with the cross-sell variant visible to half the user base.
This approach preserved the ability to pull back quickly if the data turned negative, while generating the kind of evidence the business needed to commit to wider rollout.
Outcome
The A/B test exceeded internal projections.
20% click-through rate on cross-sell prompts
£1.5m additional revenue generated across the test period
The design was shipped and progressively expanded across almost all question variants.
More importantly: the game's core experience remained intact, and user trust in the product held.
Reflection
The outcome validated the strategic premise: restraint in where the cross-sell appeared was as important as the quality of the design itself.
The biggest risk in this project wasn't designing something that didn't convert. It was designing something that converted once but eroded the product long-term.
Getting the timing right — and being willing to leave most of the game untouched — was the decision that made the commercial result possible.