Free resource · use it on this page
POS Loyalty Migration Checklist
The 30-day launch plan as a working task list, including the failure test and the paper-card conversion step that most launches skip.
How to use this checklist
This is the 30-day migration plan as a task list, meant to be worked through on the page or printed and kept by the till. It assumes you are adding a loyalty program on top of a POS you already run, not replacing the POS. One person should own the whole list, even if they delegate individual tasks. Each week gates the next: do not start week two until week one is done, and do not announce anything publicly until the soft launch in week four is clean.
Two steps here get skipped in most launches, which is why they are explicit. The first is the failure test, a set of register scenarios you run before launch day, because loyalty programs do not fail on the demo, they fail at 12:40pm with a queue. The second is the paper-card conversion step: if you have an existing punch-card program, your most loyal customers are holding stamps, and how you honor them decides whether the new program starts with goodwill or resentment.
Days 1 to 7: Groundwork
Decisions and facts to lock down before anything gets built. Everything in this week is paperwork; nothing touches the register yet.
Name a single launch owner
One person with authority to make calls on scripts, signage, and the go/no-go decision. A committee launch stalls in week three.
Confirm your POS integration mode in writing
Direct API integration, receipt or QR scan, or a standalone tablet beside the till. Get the vendor to state it in an email, not a sales call.
Count your active paper cards
Estimate how many part-filled punch cards are in circulation. This number sizes the conversion step in week three and the cost of honoring stamps.
Write the earn rule in one sentence
For example: one stamp per visit with any purchase, or one point per unit of currency spent. If it does not fit in one sentence, staff cannot explain it.
Cost-check the reward against margin
Price the free item or discount at its cost to you, not its retail price, and check the earn rule pays for itself at your average ticket size.
Define how a redemption is verified and marked
Exactly what the staff member does at the register so a reward cannot be redeemed twice. Vague redemption rules become shrinkage.
Decide what customer data you actually need
Often nothing beyond the pass install itself. Every extra field you require at signup lowers enrollment and raises your privacy obligations.
Write the privacy line
One plain sentence for signage and the pass fine print: what you collect, what you use it for, and how someone opts out or asks for deletion.
Confirm who can edit the program after launch
You should be able to change the earn rule, reward, and pass design yourself. If every change requires a vendor ticket, note that as a risk now.
Book the failure test session
Put a one-hour slot on the calendar in week two, at the actual register, with the launch owner and at least one till staff member present.
Days 8 to 14: Build it, then try to break it
Build the pass and the earn flow, then spend more time breaking it than admiring it. The failure test table below is the core of this week.
Build the pass with real branding and the earn rule on it
The one-sentence rule from week one should appear on the pass itself, so the program explains itself without staff.
Install the pass on one iPhone and one Android phone
Apple Wallet and Google Wallet render, update, and notify differently. Testing on only one is the most common shortcut and it always costs a launch-day surprise.
Run one full earn-to-redeem cycle yourself
Enroll, earn to the reward threshold, redeem, and confirm the pass resets correctly. Do it as a customer would, not through an admin panel.
Run the full failure test
Work through every scenario in the table below at the real register. Record a pass, a fix, or a written workaround for each row before moving on.
Test a refund against an earning transaction
Refund a purchase that earned a stamp or points and confirm what happens. If points survive the refund, decide in writing whether you accept that.
Time the loyalty step at the register
Run three transactions with the loyalty step included and time them. Judge against your own peak-hour queue; if it feels slow now, it will be slower with a line.
Test the update path
Change a stamp count or points balance and confirm the change appears on both test phones without the customer doing anything.
Send one test lock-screen message to your own devices only
Confirm what a campaign actually looks like on a lock screen before you ever send to customers. In PushNotice this is a campaign sent to a test pass; whatever your platform, never rehearse on a real list.
Fix what broke, then retest only what you changed
Keep a written list of failures and fixes. A fix that was never retested is a rumor, not a fix.
Draft the one-page till-side card
Enroll steps, earn steps, redeem steps, and the three scripts from week three, on a single laminated page that lives at the register.
The failure test: eight register scenarios
Run these at the actual register with a till staff member, not in an office. Each row needs a pass, a fix, or a written workaround before launch. This is the test most launches skip.
| Scenario | How to run it | Pass condition |
|---|---|---|
| Loyalty service unreachable at the register | Disconnect the tablet from wifi or block the loyalty app, then ring up a sale | The sale completes normally, and there is a defined way to credit the customer afterward, even if it is a manual note |
| Customer's phone is dead or left at home | Have someone ask to earn a stamp with no phone to show | Staff know the fallback without asking: a lookup path, a hold-the-receipt policy, or an honest no with the script for it |
| Refund of an earning purchase | Refund a transaction that earned points or a stamp | Points are reversed automatically, or the written policy explicitly accepts the small loss and staff know not to improvise |
| Duplicate earn attempt | Scan or enter the same transaction twice | The second attempt is rejected or flagged, or a till-side rule makes doubling up impractical |
| Redemption attempted twice | Redeem a reward, then try to redeem it again | The second redemption is impossible or clearly visible to staff before they hand over the reward |
| Peak-hour queue pressure | Run the loyalty step during your real rush, or simulate one with three staff in a line | The step adds seconds, not a queue; if staff start skipping it under pressure, it is too slow |
| Untrained cover staff | Give the till-side card to someone who missed training and have them run one enroll and one earn | They complete both using only the card, with no verbal help |
| Customer asks for their data to be deleted | Have someone ask at the till how to be removed from the program | Staff know who handles it and the deletion actually works when tested end to end |
Days 15 to 21: Staff training and paper-card conversion
The program is now technically sound; this week makes it operationally sound. The paper-card conversion policy is the emotional center of the migration, so it gets settled here, before launch, not improvised at the register.
Train every shift, not just the one you see most
The evening and weekend crews will run more loyalty transactions than the owner ever will. No shift launches untrained.
Rehearse the three scripts out loud
The enroll offer, the dead-phone fallback, and the graceful response to a customer who declines. Reading a script is not the same as saying it.
Set the paper-card conversion policy: honor stamps at face value
A customer with seven of ten stamps starts the new program with the equivalent of seven. Anything less punishes your best customers for your migration.
Define the conversion mechanic at the register
Staff take the paper card, credit the equivalent stamps or points into the new program on the spot, and keep or mark the card so it cannot be converted twice.
Set a conversion window with a hard end date
Long enough that regulars all cycle through, short enough that the paper program actually ends. Put the end date on signage from day one.
Keep a small stack of paper cards behind the till during the window
For the customer who wants nothing to do with phones. A quiet fallback preserves goodwill and costs almost nothing.
Print register and door signage with the install QR code
Test the QR from both a locked iPhone and an Android camera app, at arm's length, under your actual lighting.
Run a dry-run day where staff enroll each other
Every staff member installs the pass, earns, and redeems once as a customer. Their questions now are your customers' questions next week.
Name the exception contact
One phone number or chat channel staff use when something odd happens at the till, with an expected response time they can repeat to a waiting customer.
Check every piece of copy against the earn rule
Signage, pass text, receipt slips, and the till-side card must state the rule in identical words. Mismatched copy becomes an argument at the register.
Days 22 to 30: Soft launch, real launch, and the day-30 review
Launch quietly first. The soft launch exists to find the problems that only real customers can find, while the audience is still small enough that mistakes are cheap.
Soft launch to walk-ins only for two to three days
Signage up, staff offering enrollment, but no announcement anywhere. No social post, no email, no window banner yet.
Watch the first ten real enrollments end to end
Stand near the register. Where customers hesitate, squint, or hand the phone to staff is exactly what to fix before the public launch.
Convert the first paper cards and log the friction
The first conversions reveal whether the mechanic from week three actually works with a queue behind it. Adjust the script, not the policy.
Check installs, earns, and redemptions daily
Three numbers on a sticky note is enough. What matters is that earns roughly track transactions, and that the first redemption actually happens.
Fix, then confirm the fix in real traffic
Any change during soft launch gets verified on the next day's real transactions before you call it done.
Announce publicly only after a clean soft-launch day
One full day with no register incidents is the gate. Then the window sign, receipt slips, and social posts go out together.
Hold the first campaign until it can matter
A lock-screen message to nine passholders teaches you nothing and spends attention. Wait until the list is large enough that one message is a real test.
Run the day-30 review with till staff in the room
Compare enrollments to transactions, review every failure-test workaround that got used, and let staff name the one thing that annoys them daily.
Record the month-one baseline
Total installs, earn events, redemptions, and paper cards converted. Month two has nothing to improve against without these written down.
Choose the two things you will change, and schedule them
Not ten. Two changes with dates attached. Everything else goes on a list for the quarter.
Launch record
Fill this in as you go. If any line is blank on launch day, that line is your risk register.
Launch owner
The one person with go/no-go authority
POS integration mode
API, receipt or QR scan, or standalone tablet, as confirmed by the vendor in writing
Earn rule, in one sentence
Copied word for word from the signage
Active paper cards estimated
The count from week one that sized the conversion step
Failure test completed on
Date all eight scenarios passed or received a written workaround
Paper-card conversion window
Start date and the hard end date printed on signage
Soft launch dates
First quiet day through the first clean day
Public launch date
The day signage, receipt slips, and posts went out together
Day-30 review date and baseline numbers
Installs, earns, redemptions, and cards converted
From the guide: POS Loyalty Program: Add Loyalty Without Switching POS
This resource accompanies the full article — worth reading before you commit to a tool.