DM9: Growth engineering: referrals, sharing and lifecycle messages
Lifecycle messages people want
🎯You'll be able to
Design triggered messages (welcome, abandoned cart, reorder reminders) with consent, frequency caps and easy opt-out, and measure them with holdouts.
🏠The everyday picture
A tailor who calls to say your clothes are ready is helpful. One who calls every evening to ask whether you need more clothes is not.
📘What it really is
- Lifecycle messages are sent when something happens (a trigger), not on a fixed schedule:
- welcome after sign-up;
- abandoned cart after a cart is left (for example after a few hours);
- delivery and care tips after an order arrives;
- a reorder reminder when a consumable might run out;
-
win-back after a long gap.
-
Channels: email, push notifications, SMS and WhatsApp, each with its own consent and rules (DM4).
- Consent: service messages (order updates) are different from marketing messages (offers, reminders), which need separate consent.
- Frequency caps: a limit across all channels (for example no more than three marketing messages a week), and quiet hours at night.
- Easy opt-out: one click or a reply like "STOP"; honour it everywhere.
-
Measure with holdouts: keep a random share of eligible customers who don't get the message, and compare orders (DM7). Opens and clicks alone overstate the effect, because many recipients would have bought anyway.
-
Automate carefully (Module 88): triggers from events, a delay, a check that the person still qualifies (the cart wasn't bought in the meantime), and a log.
📝A worked example
Input: an abandoned-cart email for MittiMart, with an illustrative 10% holdout over four weeks: 2,000 carts emailed, 200 held out; 12% of emailed carts and 9% of held-out carts were bought within three days.
Steps:
- Incremental purchase rate: 12% − 9% = 3 points.
- Extra orders: about 3% of 2,000 = 60, not the 240 the email "influenced".
- Check guardrails: unsubscribes and complaints.
Expected output: a decision to keep the email (it adds about 60 orders a month in this example) and a note that 180 of the 240 buyers would probably have bought anyway. (Illustrative figures.)
🌍Where it's used
- Shops run welcome, abandoned-cart and win-back flows.
- Subscription services send renewal reminders.
- Banks and telecom companies use triggered SMS with DLT templates (DM4).
📝Case scenario
A cosmetics shop sent abandoned-cart messages by email, SMS and WhatsApp within an hour, plus a discount. Complaints rose, and customers learned to leave carts to get discounts. The shop moved to one email after four hours, no automatic discount, and a weekly cap; a holdout showed the extra orders barely changed while complaints fell.
🖱️Try it
Write a three-message welcome series for a new MittiMart customer: when each is sent, what it says, and how to opt out. Success: none of them is an offer unless the customer agreed to marketing.
⚠️Common mistakes
- Counting every buyer after a message as caused by it. Sign: "the email influenced 240 orders." Fix: a holdout.
- Messages on every channel at once. Sign: email, SMS and WhatsApp within an hour. Fix: one channel per trigger, and a cap.
- Not checking before sending. Sign: an abandoned-cart message after the person bought. Fix: re-check the trigger before sending.
🛠️For your project
Your capstone's design notes describe one lifecycle message with its consent, cap and holdout measure.
✅Quick check
How do you measure what an abandoned-cart email adds?
Show the answer
Hold out a random group that doesn't get it, and compare their purchase rate with those who did.
What's a frequency cap?
Show the answer
A limit on how many messages a person gets in a period, across channels.