Win-Back and Reactivation Campaigns SKILL.md

This is the exact file agents read, including the credit header and licence notice. Save it as win-back-reactivation/SKILL.md.

---
name: win-back-reactivation
description: "Win back lapsed customers without damaging deliverability or margins: define lapse from the purchase cycle, segment lapsed groups by value, recency and reason, lead with what changed, set tiered offer guardrails, design email, paid audience, SMS, WhatsApp and direct mail sequences, send in safe batches, measure reactivation against a holdout and sunset non-responders. Use whenever someone asks about win-back emails, reactivation campaigns, lapsed customers, dormant users, former subscribers, re-engagement, or how to bring old customers back."
license: CC-BY-4.0
metadata:
  author: Afeez Adebayo
  author_url: https://afeezadebayo.com
  author_linkedin: https://www.linkedin.com/in/afeez-adebayo/
  author_x: https://x.com/za_icon
  source: https://afeezadebayo.com/skills-md/marketing/crm-lifecycle/win-back-reactivation/
  category: marketing
  subcategory: crm-lifecycle
  version: "1.0.0"
  created: "2026-10-10"
  updated: "2026-10-10"
---

# Win-Back and Reactivation Campaigns

Lapsed customers already know the brand, have bought before and often left for reasons that have since changed. Winning some of them back can cost far less than acquiring new customers. But win-back done badly does damage: blasting old addresses hurts deliverability, deep discounts train customers to lapse on purpose, and messages to people who clearly left look desperate.

This skill defines who counts as lapsed for your business, segments them by value and reason, designs win-back sequences with the right channel and offer, protects deliverability and measures true reactivation against a holdout.

## Operating principles
1. **Define lapse by your purchase cycle.** A coffee subscriber is lapsed after weeks; a mattress buyer may not buy again for years.
2. **Lead with what changed,** not with a discount: new products, fixed problems, new features.
3. **Segment by value and reason.** High-value lapsed customers deserve different treatment from one-time bargain buyers.
4. **Protect the sender reputation.** Old, unengaged addresses carry spam trap and complaint risk.
5. **Know when to stop.** After a clear final message, suppress non-responders.

## Defining lapse
| Business | Typical lapse signal |
|---|---|
| Consumables e-commerce | No purchase in 2 to 3 times the usual replenishment window |
| Fashion or general retail | No purchase in 6 to 12 months |
| Subscriptions | Cancelled, or paused past the pause period |
| SaaS | Cancelled, or account inactive for 60 to 90 days |
| Apps | No session in 30 days (varies by app type) |
| Email subscribers (no purchases) | No clicks or visits in 90 to 180 days |
Use your own data: plot the distribution of time between purchases and set lapse where returning becomes unlikely.

## When to use, and when not to
Use it for former customers, cancelled subscribers, dormant accounts and long-inactive subscribers.

Don't use it for customers still active but showing risk (`retention-churn-prevention`) or for sunset policy design across all subscribers (`lifecycle-segmentation` and `email-deliverability`).

## Inputs to ask for, in one message
1. Purchase or usage history to define lapse windows.
2. Size of lapsed groups by value and time since last activity.
3. Known reasons customers left (exit surveys, reviews, support).
4. What has changed since (products, prices, features, service).
5. Channels and consent available for lapsed customers.
6. Past win-back results and offers used.

## Workflow

### Step 1. Define and size lapsed groups
Split by time since last purchase or activity (for example 3 to 6, 6 to 12, 12 to 24 months) and by value (RFM or lifetime spend).

### Step 2. Understand reasons
Use exit data and reviews to group reasons: price, quality issue, moved to competitor, no longer needed, service problem.

### Step 3. Decide the offer policy
Tiered by value and reason. High-value customers who left over service may get a personal apology and fix; price-led lapsers may get a time-limited incentive; recent lapsers may need only a reminder.

### Step 4. Design the sequence
Three to four messages over three to six weeks, escalating gently: what's new, social proof or a reason to return, an offer where justified, a final message.

### Step 5. Choose channels
Email to engaged-enough addresses; paid social and search audiences for unreachable or unconsented customers (where permitted); SMS or WhatsApp only with consent; direct mail for high-value lapsed customers.

### Step 6. Protect deliverability
Send to the most recently engaged lapsed groups first and in smaller batches; watch bounces and complaints; don't mail addresses inactive for years without re-permission.

### Step 7. Measure against a holdout
Reactivation rate, revenue, and second-purchase rate after reactivation, compared with an untreated holdout.

### Step 8. Sunset
Suppress non-responders from marketing after the final message.

## Win-back sequence example (e-commerce)
| Message | Timing | Intent |
|---|---|---|
| 1 | Day 0 | "Here's what's new since you last visited" (products, improvements) |
| 2 | Day 7 | Reviews or best-sellers in their category; reason to return |
| 3 | Day 14 | Time-limited incentive (for price-led or mid-value groups only) |
| 4 | Day 28 | "Should we stop emailing?" with preference options |

## SaaS win-back
- Remind them what they built: "Your 42 projects are still here."
- Show what changed: features that address the reason they left.
- Offer an easy restart: restore data, a short trial of a higher plan, or a call.
- Time it: some customers return when budgets reset or seasons change.

## Offer guardrails
| Rule | Why |
|---|---|
| Cap discount depth by margin and customer value | Prevents unprofitable reactivation |
| Exclude recent lapsers from incentives | They may return on their own |
| Track "lapse to get a discount" behaviour | Customers learn patterns quickly |
| Prefer value adds over price cuts | Free shipping, bonus item or upgrade often cost less |

## Finding the lapse point from data
1. For customers with at least two purchases, calculate the days between consecutive orders.
2. Plot the distribution and find where most repeat purchases have already happened (for example the 80th or 90th percentile).
3. Plot the probability of buying again against days since last purchase; set lapse where it drops sharply.
4. Set separate windows for product lines with different cycles.
5. Review the windows every six months.

## Win-back by reason
| Reason for leaving | Lead message | Offer |
|---|---|---|
| Price | Value comparison, lower-cost option or bundle | Time-limited incentive within margin |
| Quality or bad experience | Acknowledge, explain the fix, invite back | Service gesture (replacement, credit) |
| Switched to competitor | What's different now, proof from switchers | Only if economics allow |
| No longer needed | Light touch, new use cases or gift ideas | Usually none |
| Forgot or drifted | Reminder of what they liked, what's new | Usually none or free delivery |

## Paid audiences for unreachable lapsed customers
- Upload hashed customer lists to ad platforms where your privacy notice and local law allow it.
- Exclude anyone who asked not to be contacted.
- Use different creative for lapsed customers than for prospects: recognise the relationship.
- Cap frequency and run for a limited period.
- Measure with a holdout group excluded from the audience.

## Metrics after reactivation
| Metric | Why |
|---|---|
| Reactivation rate vs holdout | True effect of the campaign |
| Second purchase within the normal cycle | Whether reactivation sticks |
| Margin per reactivated customer | Whether offers pay back |
| Complaint and unsubscribe rate | Deliverability and goodwill cost |
| Time to next lapse | Whether you trained discount-waiting |

## Edge cases and judgement calls
| Situation | What to do |
|---|---|
| Customer left after a complaint | Personal message acknowledging it, explain what changed |
| Consent expired or unclear | Use paid audiences where lawful, not email or SMS |
| Very old addresses | Re-permission or suppress; don't blast |
| B2B account churned | Contact the decision-maker with relevant changes; check for new stakeholders |
| Seasonal buyers | Time win-back to the season, not a fixed date |

## Handoffs to related skills
- Lapse segments: `lifecycle-segmentation`
- Deliverability risk: `email-deliverability`
- Message writing: `lifecycle-email-copy-design`
- SMS and WhatsApp: `sms-whatsapp-lifecycle`
- Holdouts: `lifecycle-experimentation`
- Preventing future lapse: `retention-churn-prevention`

## Worked example
A pet supplies retailer had 65,000 customers with no order in over six months. Segmenting by value and time showed 9,000 high-value customers who had lapsed within the last year. They received a three-message sequence leading with a new subscription option and a personal note from customer service, with free delivery rather than a discount. Lower-value lapsers got a shorter sequence with a modest incentive. Customers inactive for over two years were sent a re-permission message only. Against a holdout, the high-value group reactivated at several times the control rate without a price discount, and complaint rates stayed below normal campaign levels.

## Output format
```
# Win-back plan: <business>, <date>
## Lapse definition and lapsed group sizes
## Reasons for lapse
## Offer policy by value and reason
## Sequences by group (timing, channel, intent)
## Deliverability safeguards
## Measurement with holdout
## Sunset rules
```

## Common mistakes
- One discount for everyone who stopped buying.
- Mailing years-old addresses in bulk.
- Ignoring why customers left.
- No holdout, so reactivation is overstated.
- Endless win-back emails with no final message.

## Questions to ask before you design
- What is a normal gap between purchases for this business?
- What has changed since these customers left?
- What margin can a win-back offer afford?

## Self-check before you answer
- Is lapse defined from data for this business?
- Are groups segmented by value and reason?
- Are deliverability safeguards in place?
- Is there a holdout and a sunset rule?

## Further reading
- Klaviyo, win-back flow guides: https://help.klaviyo.com/
- Google Postmaster Tools for monitoring complaints: https://postmaster.google.com/

---
Created by Afeez Adebayo, Marketing, SEO & Growth Expert
Website: https://afeezadebayo.com
LinkedIn: https://www.linkedin.com/in/afeez-adebayo/
X: https://x.com/za_icon
Source: https://afeezadebayo.com/skills-md/marketing/crm-lifecycle/win-back-reactivation/

Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may copy, share and adapt this skill, including for commercial use, as long as you credit Afeez Adebayo and link to the source URL above. Keep this notice in any copy or adapted version.