Case Study 03
MPL Rummy — The 7-Day Retention Journey
Applying Behavioral Psychology to a Stagnant Day 7 Retention Metric
At a glance
- Problem
- Day 7 retention for new users had sat at 25%. Activation on day 0 was healthy; the loss came during the first week, so it was a retention problem, not the wrong users arriving.
- What I decided
- A 7-day task journey with visible rewards. On a 4-week deadline I shipped it without the interactive Claim moment and carried that into V2 — a cost that showed up later as the lowest completion on Task 1.
- Result
- In an A/B test splitting all new Rummy users 50/50 on Play Store, D7 retention rose 12.7% over control, and 90% of eligible users opened the journey. Day-0 conversion did not move: the lift was retention, not activation.
- My role
- Led the research, the hypothesis and the journey design, negotiated scope with engineering and ran UAT. A UX researcher and a second designer co-ran the interviews.
01 — Problem Statement
Problem Statement
New users don't return on Day 7 by accident — in a real-money game, trust hasn't been earned yet. If we didn't engineer a reason to come back inside the first week, the business would lose them before they ever saw what Rummy could offer.
Day 7 retention for new users had sat at 25% for long enough that it had stopped being news. What made it a project was someone putting a number on it: against an average CM1 revenue baseline of about ₹75L, the majority of new sign-ups were being lost before they built any habit around the product — and the cost of that was finally legible.
Before accepting the framing I wanted to know whether this was an acquisition problem wearing a retention mask. If the new users arriving were simply the wrong users, no amount of Day-2-to-7 design would hold them. The cohort data said otherwise: D0 activation was healthy and the drop came later, which pointed at the first week itself rather than at who was walking in.
The goal: design and ship a fix within a 4-week window — a system that gives new users a reason to return every day of their first week, before they have had time to develop trust or deep engagement with the platform on their own.
02 — My Role vs. Team Contribution
My Role vs. Team Contribution
I owned
- Led the UX research plan — 1:1 interviews and surveys with new users who had churned after Day 7
- Ran the competitor analysis across genre-relevant apps (Clash Royale, Angry Birds 2, Fruit Ninja, Duolingo)
- Formed the behavioral hypothesis and designed the 7-day task-based reward journey
- Negotiated implementation scope with engineering, including descoping the Claim feature to V2
- Ran UAT and signed off on the shipped experience
Team contributed
- UX Researcher + another Product Designer: co-ran interviews and surveys
- Engineering: built the reward-crediting and task-tracking backend; flagged the Claim feature as out of scope for V1
- Marketing: campaign support for the 7-day journey launch
03 — UX Research & Key Findings
UX Research & Key Findings
To validate why users were churning, I ran 1:1 interviews and surveys with new users who had already churned after Day 7, across different age groups — testing whether a 7-day task-based journey with visible rewards could change that trajectory.
Participants
New users who churned after Day 7, across different age groups
Method
1:1 user interviews and structured surveys
Key Findings
- Users lacked immediate rewards for their actions — unlike competitor apps, MPL Rummy offered no instant gratification, leaving them unmotivated to continue.
- The onboarding experience was confusing — with no guided flow, new users didn't understand the app's benefits and often churned out of frustration.
- Users preferred clear, structured tasks that led to tangible rewards — task-based engagement gave a sense of achievement and progression.
- Past task-based initiatives had shown higher engagement, reinforcing that a structured journey could work.
- Users repeatedly compared MPL Rummy unfavorably to competitors with more engaging reward systems.
- Users wanted something that encouraged a returning habit — small daily rewards accumulating into a bigger one.
The pain points behind those findings were consistent across interviews: no clear reward for continued play, an overwhelming first-run experience, and no structured onboarding to explain what to do next.
If new users are given a structured 7-day journey with clear tasks and visible, escalating rewards, they'll form a return habit before churning — closing the gap that competitor apps had already exploited with FOMO and progress-tracking mechanics.
The 7-day journey's entry point, surfaced as a task banner inside the existing Rummy lobby
The Rummy Carnival screen — Day 1–7 tasks with escalating, partially-locked rewards and a live countdown
04 — Competitor Analysis
Competitor Analysis
I benchmarked genre-relevant apps — Clash Royale, Angry Birds 2, Fruit Ninja, and Duolingo — to see how they engineered return behavior in their own onboarding and reward systems.
Where Competitors Nailed It
Clash Royale and Fruit Ninja unlock the next reward only once the current one is claimed — a visible, sequential journey that motivates the next step. Angry Birds 2 goes further: a task flips from a GO button to a CLAIM button the moment it's done, with a checkmark and a delightful collect animation confirming exactly what was won.
The Gap in MPL Rummy
None of this existed in the new-user flow — no visible sequencing, no explicit claim moment, and no animated confirmation of what a user had actually earned.
- Clash Royale — a King's Journey timeline where each completed task visibly advances progress, reinforcing a habit-forming loop
- Angry Birds 2 — quests flip from GO to CLAIM on completion, with a checkmark and a delightful collect animation
- Angry Birds 2 — a free starter treasure teaches the claim mechanic before asking anything of the user
- Fruit Ninja — rewards unlock strictly in sequence, so claiming one is what motivates the next
- Duolingo — upcoming quests stay visible but locked ("Revealed in 2 days"), keeping the journey ahead in view
The pattern held across all four: visible sequencing, an explicit claim moment with animated feedback, and upcoming rewards kept in view even while locked. That became the direct blueprint for the 7-day journey.
05 — User Flow
User Flow
I mapped the flow from the existing MPL home screen down to claiming a reward, keeping a cancel path back to the Rummy home page at every step so a hesitant new user is never trapped mid-journey.
Rummy Home Button
The entry point had to earn a tap on its own — a new user has no reason yet to know what's behind it.
Select a Task
Picking a task surfaces its own progress, so the user always knows exactly where they stand in the day's goal.
Seven Tasks, Seven Days
Except the first one, the rest are locked and unlock as the user progresses.
View the interactive prototype in FigmaThe full 7-day journey flow, as designed
The full flow — MPL Home → Rummy Home → Cash Back Dhamaka → pick a task → perform → claim → unlock the next, with a cancel path back to Rummy Home at every step
06 — The Hardest Decision
The Hardest Decision
Decision
Ship the 7-day task journey on time without the Claim feature — commit it to a V2 release
With a 4-week timeline, engineering flagged that the Claim feature — the interactive moment where users actively claim their reward — couldn't be built without extending the deadline. I agreed to descope it and carry it into V2 rather than delay the entire journey.
| Factor | Option A | Option B — Chosen |
|---|---|---|
| Reward delivery | Interactive Claim moment — higher delight, needs more backend work | Auto-credited rewards — ships on time, less delight |
| Scope vs. timeline | Full-featured V1 — richer experience, missed deadline | MVP journey — hits the 4-week deadline, enhance later |
| Risk | Push the deadline for completeness | Ship on time, validate the core hypothesis first |
What it cost is measurable, and it showed up where you would expect. Task 1 — the one a brand-new user meets first — had the lowest completion of the seven, climbing to a peak around Task 4. Shipping without the Claim moment meant the first task had no active, satisfying beat to it; rewards simply appeared. Lowering the Task 1 threshold in the follow-up lifted completion, and the Claim feature landing in V2 is the other half of that same fix.
07 — Rollout & Measurement
Rollout & Measurement
The journey launched on schedule, without the Claim feature, with an A/B test set up to measure impact against a control group.
What went wrong
Task-level data showed completion was lowest on Task 1 of 7, before climbing and peaking around Task 4, then tapering off through Task 7.
How it was corrected
We reduced the completion threshold for Task 1 in a follow-up version of the journey, which increased Task 1 completion and made more users eligible to reach Task 2.
The flow also handled a lapsed state — missing a day restarts the journey from Day 1
A win-celebration moment cross-promoting back into the journey and a leaderboard entry point
Retention improved before we'd even shipped every intended mechanic — which told us the core loop, not the polish, was doing the work.
08 — Quantitative Results
Quantitative Results
The journey was A/B tested twice on Play Store. The second round, started 16 June, split all new Rummy DAU 50/50 — 94,062 users saw the 7-day journey (test) against 76,679 who didn't (control).
- D7 Retention: Test group improved 12.7% over control on Play Store
- Carnival Page Visit Rate: 90% of eligible users opened the journey page on Play Store, versus 25% on the Pro app variant
- D0 DAU → PDAU Conversion: No change between test and control — the lift was specific to D7 retention, not overall activation
Day-wise completion rates in the weeks after launch
Task-level completion — lowest on Task 1, peaking around Task 4, tapering after
The production task ladder and rewards, with an actual completed-journey payout
09 — Next Steps
Next Steps
- Rolled the 7-day journey out by default to all new Play Store logins after the second, larger-scale A/B test (94,062 test vs. 76,679 control).
- Shipped the Claim feature in V2 — the interactive claiming moment engineering couldn't fit into the original 4-week window — which generated real delight for users.
- Planned a further A/B test measuring D15 retention specifically for users who completed the 7-day journey, across cohorts like resurrection and roller retention.
- Explored celebrating wins on a 7-day leaderboard and experimenting with variable rewards.
10 — Design Principles Demonstrated