Re-engaging Lapsed Users to Drive Repeat Activity
How Levered designed a targeted re-engagement campaign that increased weekend driver hours by 1%, generating an estimated $5.2 million in additional annual revenue.
At a glance
+1%
Weekend hours lift
$5.2M
Annual revenue impact
Near-miss messaging
Primary behavior trigger
Capabilities used
The problem
• Drivers who earned $100+ in a weekend frequently failed to return the following weekend — a specific behavioral signal, not a broad lapse definition. • That early retention leak meant lost supply exactly when riders needed it most: peak weekend demand. • The challenge wasn't identifying the segment — it was finding a behavioral nudge that was testable at marketplace scale without promo spend. • A blunt "come back" message wasn't enough; the intervention needed to understand driver psychology to move the needle.
The fulcrum
The highest-leverage change we could make
• Designed a 3-arm split: 34% control (no comms), 33% "Gained" framing (what you earned last weekend), 33% "Near-miss" framing (how close you were to the next earnings tier). • Same underlying data — different psychological frame — so the experiment could isolate which mental model actually pulls drivers back online. • Push notifications only, no promo spend — a pure messaging test with no incentive cost. • Experiment was designed to produce a durable insight, not just a one-time lift.
Applying force
What happened, and what we scaled into
• Near-miss framing: +0.72% driver hours (p=0.080) — outperformed simple earnings reminders. • Gained framing: +0.56% driver hours (p=0.176) — directionally positive but weaker. • ~+1% weekend hours in the targeted segment → ~$5.2M estimated incremental annual revenue at Lyft scale. • Near-miss psychology is now a proven retention lever — a durable insight applicable to future driver lifecycle work.
Why this matters
• A 1% lift in weekend hours sounds incremental. At Lyft's marketplace scale, that's $5M+ annually from a push notification with no promo cost. • The insight — that near-miss framing outperforms simple earnings reminders — only exists because the experiment was designed to test psychology, not just send a message. • Most re-engagement campaigns pick a message and blast it. This one was built to learn something, and the learning compounds across every future retention program. • If you have a retention leak and you're not sure why users aren't coming back, the answer is usually in the framing — not the channel.
Most engagements fail at the handoff between strategy and execution. Levered doesn't hand off — we own both.