NevTan Engage lets you create automated email, push, SMS, and WhatsApp customer journeys, segment audiences, and deliver personalized campaigns powered by unified customer data.
You have users who signed up, maybe bought once, then went quiet. Re-engaging them is one of the higher-ROI activities available to a growth team, because they already know you — but most campaigns fail because they're generic, badly timed, or sent to people who should have been let go.
This guide covers the framework that works, and the part most guides skip: deciding who not to message.
Define inactivity precisely, segment by value rather than treating all lapsed users alike, give a real reason to return, automate the sequence with exit conditions, and measure incremental lift against a holdout. The counterintuitive part:mailing deeply inactive users damages the deliverability of everything else you send, so re-engagement has to be paired with a sunset policy, not run instead of one.
What You Need Before Starting
A precise definition of inactive. For a mobile app, no session in 30 days. For ecommerce, no purchase in 90. The number matters less than it being written down and measured consistently.
Access to behavioural events — last login, last purchase, last engagement — and the ability to segment on them. Re-engagement without behavioural data is a newsletter with a guilt-trip subject line.
At least two channels. Single-channel re-engagement reaches only the people who still check that channel, which is a shrinking set by definition — someone who stopped opening email is not reachable by more email.
A holdout group. Set aside 5–10% of each inactive segment that receives nothing. Lapsed users return on their own all the time, and without a control you'll attribute organic returns to your campaign. This is the single most important measurement decision in re-engagement and the one most often skipped.
A sunset policy. Decide now what happens to people who don't respond. Not later, when you'll be tempted to keep messaging them.
Step 1: Define Inactivity and Segment by Value
Query for users who haven't performed a key action inside your window, then split them — because "lapsed" describes a date, not a problem.
Tier | Definition | What they need |
|---|---|---|
Never activated | Signed up, never completed the core action | Onboarding help, not a win-back offer |
One-time, lapsed | Activated or purchased once, never returned | A reason to return; often the largest group |
High value, lapsed | Multiple purchases, now quiet | Personal outreach, VIP treatment — worth the cost |
Long dormant | No engagement in 6–12 months | Probably a sunset candidate, not a campaign target |
RFM scoring (recency, frequency, monetary) is the standard way to rank these, and behavioural segmentation is what keeps the segments updating as people move between them. Build them so users enter and exit automatically — see segments documentation.
Pro Tip: A user who never activated is a fundamentally different problem from one who was active and stopped. The first has never seen your value and needs onboarding; the second has seen it and chose to leave. Sending both "we miss you" wastes the message on the first group and insults the second.
Step 2: Check Whether You Should Message Them at All
This step isn't in most re-engagement guides, and skipping it is how a win-back campaign costs more than it recovers.
Mailing deeply inactive addresses damages your sender reputation. Mailbox providers weight engagement heavily. A large send to people who haven't opened in a year produces low engagement, elevated complaints, and — the serious risk — possible hits on recycled spam traps, which are abandoned addresses that providers reactivate specifically to catch senders mailing stale lists. One trap hit can affect placement for your entire list, including your best customers.
The practical rule: re-engagement is worth running on recently lapsed users and expensive on long-dormant ones. Draw a line — often around 6–12 months of no engagement — beyond which you suppress rather than campaign.
Sequence it this way:
Recently lapsed (30–90 days) → full re-engagement journey
Moderately lapsed (90 days–6 months) → one short sequence, then decide
Long dormant (6+ months) → one final attempt at most, then suppress
Suppressed doesn't mean deleted. Keep the record, stop the sending. If they return on their own, you'll know.
Deliverability fundamentals covers why this matters more than the campaign itself: a re-engagement send that recovers 200 users while degrading inbox placement for 50,000 is a net loss you won't see in the campaign report.
Step 3: Choose Channels and Sequence Them
Escalate from least to most intrusive, and only to channels the person actually consented to.
Day | Channel | Condition |
|---|---|---|
1 | All consented | |
4 | Push | App installed, notifications enabled |
8 | SMS | Consented to SMS; higher-value segments only |
12 | Opted in, in markets where it's standard |
Space it wider than feels natural. A common mistake is compressing a four-touch sequence into a week, which reads as desperation to someone who already disengaged. Spreading the same four messages across two weeks also stays inside a sensible frequency cap, which a Day 1/3/5 sequence does not.
Each message must add something new. Four variations of "come back" is one message sent four times. Change the angle each touch, not just the channel.
Respect the cost asymmetry. Email is nearly free at scale; SMS and WhatsApp carry per-message costs plus registration requirements. Use the paid channels on segments where the recovered value justifies them — which is another argument for value-tiered segmentation rather than one blanket campaign.
The channel comparison covers which fits where, and WhatsApp carries its own template approval process before you can initiate contact.
Pro Tip: Exit anyone who re-engages mid-sequence, immediately. A "we miss you" SMS arriving after someone just bought tells them nobody is paying attention — and it's the most visible failure in the whole journey.
Step 4: Give a Real Reason to Return
"We miss you" is a feeling, not an offer. The message has to answer what's in it for them.
Four angles that work:
Angle | Example | Best for |
|---|---|---|
New value | "The feature you asked for is live" | Users who left over a gap |
Unfinished business | "Your cart is still here" | Recent abandoners |
Exclusive incentive | "20% off, just for you" | High-value lapsed, used sparingly |
Social proof | "10,000 people tried this last month" | Never-activated users |
On discounts: they work, and they attract people who return for the discount and leave after it. Try non-monetary incentives first — new features, exclusive content, saved progress. When you do discount, measure what those users do over the following 90 days, not just whether they converted this week.
Keep copy tight: under 160 characters for SMS (one segment — an emoji drops that to 70), under about 100 for push, one clear CTA for email. Personalisation should change the offer, not just insert a first name.
Pro Tip: Give a reason, plainly stated. "Here's 20% off because it's been a while" outperforms a bare discount code — people respond better to offers that make sense than to offers that appear from nowhere. This is a real effect, though the popular statistics attached to it come from lab studies on small in-person requests and don't transfer cleanly to marketing copy.
Step 5: Automate with Triggers, Branches, and Exits
Manual re-engagement doesn't scale and doesn't run consistently.
Build a journey that triggers on segment entry, with:
Delays of three to four days between touches
Branches — opened but didn't click gets a different angle than no engagement at all
Exit conditions on conversion, and on any meaningful activity, not just purchase
Frequency caps enforced at platform level so someone in two journeys doesn't get both
Built-in flows cover the standard win-back pattern; custom flows handle anything specific to your model.
Where converts go next matters. Someone who returns should move into onboarding or loyalty, not back into the general list. A reactivated user who receives the same broadcast stream that lost them the first time usually lapses again — real-time behavioural messaging is what makes the second relationship different from the first.
Pro Tip: Add a final message with a genuine deadline before the journey ends. It performs well — and it also gives you a clean decision point: no response after the last chance means sunset, no further deliberation needed.
Step 6: Measure Incremental Lift, Not Reactivations
This is where most re-engagement reporting goes wrong.
Your holdout is the measurement. If 14% of your campaign group returns and 3% of your holdout returns, your campaign caused 11 points, not 14. Reporting the raw 14% overstates your result by more than a third, and that inflated number then becomes the benchmark you're measured against next quarter.
Track:
Metric | What it tells you |
|---|---|
Incremental reactivation rate | Campaign group minus holdout. The real number |
Incremental revenue | Campaign group revenue minus holdout revenue, scaled |
Revenue per reactivated user | Whether the segment is worth the channel cost |
Channel-level performance | Where to spend the SMS budget |
90-day retention of reactivated users | Whether they stayed |
That last row separates a good campaign from an expensive one. If reactivated users churn again within 30 days, you've bought a brief spike — usually a sign the incentive attracted deal-seekers rather than people who wanted the product.
Review performance in automation reports and campaign reports; the metrics glossary defines each so your comparisons hold. Test one variable at a time — what to test and when covers sequencing, and send timing is worth isolating from copy.
A note on significance: how long a test takes depends on your baseline rate and the effect size you're trying to detect, not on a fixed user count. Calculate required sample before you start rather than running until the numbers look good.
Worked Example: A Fitness App
A modelled example showing the structure, not a customer result.
50,000 users with no app session in 45 days, split by prior behaviour:
Segment | Size | Sequence |
|---|---|---|
Completed ≥1 workout | 30,000 | Email D1 → push D4 → SMS D8 → WhatsApp D12 with premium offer |
Never worked out | 20,000 | Onboarding-focused: "Your first workout takes 7 minutes" |
A 10% holdout in each, receiving nothing.
The structural lesson is in the segment split, not the rates. The two groups needed different messages because they represent different problems. Group one knows what a workout in your app feels like; group two never found out. Sending both a "come back" message would have wasted it on half the audience.
How to read your own results. Take the campaign group's reactivation rate, subtract the holdout's, and multiply the difference by segment size. That's what the campaign produced. Then multiply by revenue per reactivated user, and subtract channel costs — SMS and WhatsApp are per message, so a sequence that reaches 30,000 people on four channels has a real cost that the email-only version doesn't.
Expect channel results to differ in instructive ways. The channel with the highest click rate is often not the one producing the most total reactivations, because the cheap broad channel reaches more people. Both numbers matter: one tells you where to invest per message, the other tells you what actually moved the total.
See customer case studies for production accounts.
Matching Strategy to Your Business
Mobile-first apps. Push and in-app lead, since the app is where value lives. Email as backup for people with notifications disabled — see SaaS and app teams.
Ecommerce. Email and SMS are the workhorses; browse and cart data give you the most specific hooks available. Ecommerce brands usually find unfinished-business angles outperform discounts.
Markets where WhatsApp dominates. India, Brazil, Indonesia, Mexico, much of Latin America — prioritise it, with template approval arranged in advance.
B2B SaaS. Email plus human outreach. SMS reads as intrusive in most B2B contexts, and the buying cycle rewards patience over urgency.
Small lists (under 10,000 inactive). High-touch and personalised beats automated. At that size a human can look at the high-value names.
Large lists. Automation and segmentation are mandatory, and so is the sunset policy — scale makes the deliverability risk larger, not smaller.
Why Re-Engagement Works
Most users don't leave because they dislike you. They leave because you stopped being relevant, or they simply forgot. App and ecommerce retention curves both drop steeply in the first days and weeks after acquisition, which means a large share of any user base is reachable-but-inactive at any time.
Three mechanisms do the work:
Reciprocity — give something of value and people feel inclined to respond. Loss aversion — "your saved progress expires" motivates more reliably than "come back." Relevance — a message referencing what they actually did beats a generic appeal.
Multi-channel works because attention profiles differ. Email is checked routinely but crowded. Push is immediate and easily disabled. SMS and WhatsApp are near-certain to be seen and correspondingly easy to resent. Sequencing increases the chance of reaching someone at a moment they're receptive — and reaches people who abandoned one channel but not another, which is the real mechanism rather than repetition.
A caution on the "98% open rate" figures you'll see quoted for SMS and WhatsApp: SMS has no open tracking at all. Carriers confirm delivery to a handset; nobody can measure whether it was read. Those numbers conflate delivery with reading. Measure what's observable — delivery, clicks, conversions, opt-outs.
Common Mistakes
1. One message for everyone. A never-activated user and a lapsed VIP need different things.
2. No holdout. You'll credit your campaign with organic returns and set yourself an impossible benchmark.
3. Over-messaging. Compressing four touches into five days reads as desperation. Space them, cap frequency, honour opt-outs — chronic over-sending is one of the biggest mistakes in customer retention.
4. Ignoring channel preference. Someone who hasn't opened email in six months isn't reachable by email. Switch channels or stop.
5. Giving up after one touch. Most reactivations come on the second or third message.
6. Mailing deeply dormant users. Spam traps and low engagement damage placement for your whole list. Suppress rather than campaign past your threshold.
7. Optimising for reactivation instead of retention. A returning user who churns in 30 days cost you money.
8. No exit conditions. Messaging someone who already returned is the most visible failure in the sequence.
Frequently Asked Questions
What's a good reactivation rate?
It varies enormously by industry, channel, how long people have been inactive, and how you define reactivation — which makes cross-industry benchmarks close to useless. The number that means something is your lift over holdout. A campaign returning 14% where the holdout returns 3% produced 11 points. Measure your own baseline and improve on it.
How many messages should a sequence have?
Three to five touches across two or three channels over roughly two weeks. If there's no response after the last one, stop — continuing damages deliverability and irritates people who may still buy later through another route.
Should I offer a discount?
Strategically. Discounts work and they attract deal-seekers who churn again. Try non-monetary incentives first, reserve discounts for high-value lapsed users, and measure 90-day retention rather than immediate reactivation.
How do I re-engage users who never open emails?
Switch channels — push, SMS, or WhatsApp, where they've consented. If email is your only option and they still hold valid consent, a short re-permission message is reasonable; if they don't respond, suppress them. Note that if you lack a lawful basis to email someone, you also lack one to email them asking for permission — our compliance guide covers where that line sits.
What's the best send time?
Your own data beats any published benchmark. Look at when your active users engage and send in those windows, then test against it.
How do I measure ROI?
Incremental revenue — campaign group minus holdout — less campaign costs including per-message SMS and WhatsApp charges. Then track reactivated users' value over 90 days to see whether the return held.
Can I re-engage users who opted out?
No. Opted-out contacts cannot receive marketing messages. Transactional messages about an active account remain permissible, but re-engagement is marketing by definition. Respect the opt-out permanently.
When should I stop and just sunset them?
When the deliverability cost exceeds the recovery value — usually somewhere past 6–12 months of no engagement, depending on your purchase cycle. Suppression isn't deletion: keep the record, stop the sending, and let them come back on their own terms.
Start With Your Highest-Value Segment
Re-engagement rewards precision more than volume. Define inactivity, split by value, give people a real reason, and measure against a holdout so you know what you actually caused.
And set the sunset line before you start — the users you shouldn't message are as important to identify as the ones you should.
NevTan Engage runs these journeys across email, push, SMS, and WhatsApp from one unified profile, with consent per channel, platform-level frequency caps, and automatic suppression — so a sequence stops the moment someone returns.
Start free — no credit card required.
