Customer retention software: turn useful records into useful follow-up.

Keeping customers starts with knowing who returned, who has gone quiet and who wants to hear from you. Choose software around that everyday job, then use a small, measurable routine instead of sending another offer to the whole list.

By Moe Ergig · Updated · 9 min read

Realistic editorial image of a shop owner’s desk, diary and phone ready for a customer follow-up.
AI-generated editorial photo. Not a product screenshot.

You may already have the ingredients in a till, booking system, loyalty programme or spreadsheet. The question is whether they can be matched into a record you trust and used by someone who has time to follow through. A report alone does not complete that work.

Check history → Choose a segment → Review a draft → Track a return
A customer returning after a message is an observation, not proof the message caused it.

1. Define what returning means for your business

Choose the behaviour you want to measure. For a café it might be another qualifying visit; for a shop, another order; for a service business, a completed appointment. Do not quietly switch between bookings, payments and visits in the same report.

Pick a time window that fits the normal cycle. A weekly regular being absent for several weeks means something different from an annual customer not buying this month. If there is too little history to know an individual pattern, use a documented starting rule and review the records before acting.

2. Check the records before building a segment

Customer identity is the first problem to solve. One person can appear twice with a different phone format, while a family can share a number. A refunded order or a cancelled appointment should not quietly become a completed purchase. Decide which system holds the authoritative history and how you handle corrections.

Minimum information to review before a follow-up
FieldQuestion to ask
Customer identityIs this one person, or a duplicate or shared account?
Last completed visit or purchaseIs it a real completed event in the source system?
Usual purchase cycleIs there enough history to call this absence unusual?
Communication preferenceIs this person eligible for this kind of message?
Open issue or recent contactWould another offer be inappropriate or repetitive?

3. Choose a useful customer segment

Useful starting groups include a first-time customer who has not made a second visit, a regular whose usual rhythm has changed, or someone with an unused reward. They need different messages. The first may need help understanding the service; the second may simply be away; the third may need a clear explanation of the benefit.

These are example segment ideas, not claims that every connected account calculates every rule automatically. In Flann, check the available loyalty data and the actual segment shown in your account. Use the loyalty guide to make sure the visit data is being recorded consistently.

4. Write a message with a reason and a next step

A follow-up should explain why it is useful now. Mention a real new range, an actual reward or a relevant service update. Do not invent urgency, imply a personal relationship you do not have or use private purchase details where someone sharing the phone might read them.

WhatsApp’s business policy requires permission for subsequent contact and respect for opt-outs. A previous purchase or a number in a spreadsheet does not automatically settle that question. Follow the rules applicable to your messaging channel, including template requirements where they apply.

5. Connect MSTRMND digital loyalty with Flann AI

MSTRMND brings digital loyalty cards, customer records and messaging together, with Flann AI as the management layer. Customers add a branded card to Apple Wallet or Google Wallet instead of downloading a separate loyalty app. The business can then use visits, spend, rewards and last-seen information to choose a useful follow-up.

Eight digital card types for different retention goals

These are eight card formats, not eight different wallet apps. Choose the format that matches how your customers buy. All eight are available on MSTRMND’s Business plan; lower plans offer fewer card types.

MSTRMND’s eight digital card types
Card typeHow it worksA useful retention job
Stamp cardCollect a stamp on each qualifying visit.Give customers a clear reason to return.
Prepaid passBuy a block of visits upfront.Keep the remaining visits easy to see.
Cashback cardEarn wallet credit from eligible spending.Bring earned value into the next purchase.
Points cardEarn points and redeem them for rewards.Reward different purchase amounts consistently.
Discount cardCarry a standing percentage or fixed offer.Make a member benefit easy to use.
CouponRedeem a dated, one-time offer.Run a seasonal promotion or win-back.
Gift cardSpend a prepaid gift balance.Welcome a new customer through a gift.
MembershipUse a recurring plan with member benefits.Support an ongoing customer relationship.

Free unlimited notifications, plus SMS and WhatsApp

MSTRMND includes free, unlimited wallet push notifications. Use them for an earned reward, a relevant update or a reminder attached to the customer’s card. It also supports SMS and WhatsApp messages. The free, unlimited inclusion applies to wallet push notifications; check the applicable messaging costs for SMS and WhatsApp.

Choose the channel around the job. A reward update belongs naturally with the wallet card. SMS can reach an eligible customer directly, while WhatsApp supports a conversation your team can continue. Unlimited notifications give you room to communicate; relevance and the customer’s preferences should still determine frequency.

Manage the programme through Flann AI

With MSTRMND connected, Flann AI can manage the retention workflow from customer insights and audience selection to campaign wording, messaging and results. Give the goal in chat rather than rebuilding the campaign in several screens. For a review-first workflow, Flann prepares the action and you approve, edit or decline it.

MSTRMND describes the connected programme as fully manageable by Flann AI. Agree the routine’s schedule and review setting once, then use the run report to see what it prepared or sent. See MSTRMND’s current plans for card availability and the Flann inclusion.

6. Measure returns with a worked example

For a simple cohort return rate, divide the number of people from a defined starting group who return by the number of people in that starting group. Keep the period and identity rule fixed. This is different from comparing the total customer count at the start and end of a month.

Fictional four-week cohort example
MeasureExample value
Customers in the starting group100
People from that group with a completed return visit28
Cohort return rate28 ÷ 100 × 100 = 28%
What the number does not establishWhether a follow-up message caused those visits.

7. Keep a weekly scorecard small enough to use

Track the reviewed audience, confirmed sends, replies needing action, completed returns, reward or discount cost, and opt-outs. Add failed or uncertain deliveries. These measures answer different questions and should stay separate.

  • Audience matches the written selection rule.
  • Permission and opt-out status were checked.
  • The offer is current and affordable.
  • A team member owns incoming replies.
  • Unclear delivery outcomes were checked before retrying.
  • Returns use the same reporting window each week.

Use the repetitive-task guide to build a routine, and check the current pricing page against the tools and data your programme requires.

A two-minute reset before the next follow-up

  1. Open one selected record and verify its last visit.
  2. Confirm the message is relevant and permitted.
  3. Check whether that person was recently contacted.
  4. Name the person who will handle the reply.

If you cannot complete those checks, improve the segment before increasing the volume. The next useful result is a well-chosen conversation you can follow through.

Common questions

How do MSTRMND and Flann work together?

MSTRMND provides digital loyalty cards, customer records and messaging. Connected Flann AI manages the workflow through chat: use customer insights to choose an audience, prepare a campaign and manage follow-up routines. Choose a review-first routine when you want to approve, edit or decline proposed sends.

What is a useful definition of a lapsed customer?

Start with the normal purchase or visit cycle for your trade, then document the rule you use. A weekly café regular and an annual service customer need different windows. Check individual records before treating an absence as a reason to send an offer.

Can I send a win-back message to every past customer?

A past purchase does not automatically establish permission for WhatsApp marketing. Check the customer’s recorded choice and opt-out status, and follow the rules of the messaging channel you use. Keep outreach relevant and make stopping it straightforward.

Are MSTRMND notifications free and unlimited?

MSTRMND includes free, unlimited wallet push notifications. It also supports SMS and WhatsApp, whose messaging costs should be checked separately. Its eight card types are stamp, prepaid, cashback, points, discount, coupon, gift and membership; plan availability varies.

Sources and further reading

Original platform documentation and guidance behind the practical checks in this guide.

Start with one useful job.

Explore Flann’s tools, choose a task you can check and review the result before adding more.