LIVE Typesy · 34% → 39.5% 12-week retention

The one behavior that
predicts retention. Found, then driven.

We test your product data, identify the activation milestone that actually moves retention, then drive users toward it with staged emails you approve.

48-hour turnaround · no setup before the call

sendlyr · retention engine running

12-week retention

34.0%

baseline · n = 24,118

With Sendlyr signal

39.5%

milestone met in 14d

Lift

+5.5pp

p < 0.001

Time to insight

48h

no integration needed

Live in production with

Typing app · 24k+ users · subscription

How it works

Three steps. One milestone.

No instrumentation, no new tracking. We work with the behavioral events you already capture in Mixpanel, Amplitude, Firebase, or your warehouse.

01 · Connect

Amplitude 2.1M events

Mixpanel 1.4M events

Firebase 680k events

Pipe your event data

Read-only access to the analytics you already use. We pull the last 90 days and join it to your retention outcomes.

02 · Analyze

signal_a62%

signal_b71%

signal_c92%

signal_d48%

signal_e55%

Test behaviors against retention

We score every candidate early behavior on lift, coverage, and statistical significance. The winner becomes the milestone.

03 · Drive

New user · day 1queued

Activated · day 5sent

Stuck · day 7staged

Reaching milestonesent

Staged emails, your approval

Each user is routed by behavior toward the milestone. Every email waits in a review queue — you approve, reject, or regenerate.

Before Sendlyr

Writing emails is the easy part. Knowing what to drive is the hard part.

Most teams optimize toward a guessed activation metric. Well-written sequences push users toward the wrong behavior — and retention never moves.

email_2 · onboardingdraft 7

tonew users

subject

tone unclear cta missing 2h 14m

Symptom 01

Driving the wrong milestone

The activation metric was picked because it sounded right — never tested against whether it actually predicts who stays.

Wait between emails editing

2 days

Typing…

2d 5d 3d

Symptom 02

Open-rate theater

Tweaking cadence and copy lifts opens. None of it lifts retention if the target behavior isn't the one that predicts staying.

01signup

02send email

03wait 2d

04follow-up

Symptom 03

Building before knowing

Flows shipped before the analysis. Nobody can tell whether the journey is aimed at the right moment, only that it sends.

After Sendlyr

Find the milestone first. Then write the lifecycle.

We test behaviors against retention, surface the one that predicts staying, and stage every email for your review before anything sends.

✦

Step 01

Milestone from data, not guesswork

Every candidate behavior scored against 12-week retention. The winning signal becomes the target the lifecycle drives toward.

New0

Stuck0

High intent0

Churn risk0

Activated0

Step 02

Users routed by behavior

Same product, different message. Users who reach the milestone get reinforcement. Users stuck before it get the unblock email.

✓

queued · 8am tomorrow

Step 03

You approve every send

Nothing leaves the queue without your review. Read the draft, see the reasoning, approve or reject. Judgment stays in the loop.

Case · Typesy

We don't guess the activation moment. We test it.

For Typesy, we compared early product behaviors against 12-week retention and selected the milestone with the strongest useful lift and enough coverage to matter.

One milestone reached in the first 14 days retained at 39.5% — a 5.5pp lift over baseline. We mask the others to protect customer data.

Candidate signal Retention Decision

Key action B in 14 days 37.9% 81% cov · drop

Key action A in 14 days 36.1% 82% cov · drop

Threshold met 39.5% 79% cov · primary

12-week retention · Typesy

34%

Before

→

39.5%

After

Typing app · eReflect · n = 24,118 trials

Why Sendlyr

Every other tool assumes you already know which behavior matters.

Sendlyr finds it first. Emails, segments, and the review queue come after the data shows which milestone is worth driving.

01 · The milestone

Find what predicts retention

Test candidate behaviors against retention outcomes. Pick the one with real lift and meaningful coverage.

No activation guesswork

02 · The lifecycle

Build the sequence around it

Every email exists to move a user closer to the proven milestone. Not a generic checklist. Not a templated drip.

Email becomes a retention lever

03 · The control

Approve every send

Each email waits in a review queue. Approve, reject, or regenerate before it reaches a single user.

Human judgment in the loop

◆ Get started

See your signal on your users in 48 hours.

Book a discovery call. We'll look at your event data, test behaviors against retention, and show you the milestone that predicts who stays — before we write a single email.

Book a discovery call →

No commitment No setup before the call 48-hour turnaround

Blog