How to Assign Points by Revenue Band: A Step-by-Step Guide for Loyalty Programs
05/11/2026
6
To assign points by revenue band, you group orders into ranges of order value and give each range its own earning rate, so bigger baskets earn more. Draw the bands from your own order data, choose a calculation that avoids sudden jumps at the edges, and check the cost of each band against your margin before launch. Then write clear rules for discounts, tax, returns and rounding, and test the bands against a control group.
Most points programs start with one simple rule, such as 1 point for every dollar spent. It is easy to explain, but it treats a $15 order and a $150 order the same way. Many retailers want to reward the larger basket a little more, both to say thank you and to give shoppers a reason to add one more item.
This guide walks through six steps, using one worked example from start to finish.
What is a revenue band in a loyalty program?
A revenue band is a range of order value, such as $30 to $59.99, that earns its own points rate. When a customer checks out, the system looks at the order value, finds the matching band and awards points at that band’s rate.
Most loyalty platforms support this kind of rule. Square’s Loyalty API documentation, for example, describes spend-based rules such as one point for every $2 spent, visit rules with a minimum purchase amount, and a setting for whether tax counts toward the purchase amount.
Revenue bands are different from customer tiers. A band looks at one order. A tier looks at how much a customer has spent over a year and gives them a status such as Silver or Gold. The two can work together, but they solve different problems.
Read more:
Step 1. Set the goal for your bands
Start by writing down the one behavior you want to change. For most retailers, it is one of these:
- Raise average order value: encourage shoppers to add a second or third item.
- Reward high-value orders: say thank you for large baskets without raising the rate for everyone.
- Protect margin: keep low-value orders on a lower rate so rewards stay affordable.
The goal decides where the bands sit. If you want bigger baskets, the first band edge should sit just above your typical order. If you only want to thank big spenders, the top band can start much higher.
Step 2. Draw your bands from your order data
Export the last 6 to 12 months of completed orders and sort them by order value. Then look at a few points in the list, called percentiles.
In our example, a home and lifestyle retailer has 10,000 orders a month and an average order of about $39. Half of its orders are under $30, about a third fall between $30 and $60, and the top 15% are $60 or more. That suggests three bands:
| Band | Order value | Share of orders |
| Band 1 | Under $30 | 50% |
| Band 2 | $30 to $59.99 | 35% |
| Band 3 | $60 and above | 15% |
Three simple tips help here. Use round numbers that customers can remember. Keep to two or three bands, because more bands are harder to explain at checkout. And place the first edge a little above your middle order, so many customers are within reach of it.
Step 3. Choose how points are calculated
This is the step most guides skip. There are three common ways to turn a band into points, and they behave very differently at the band edges. In each example below, the base rate is 1 point per $1, and 100 points are worth $1.
Model A: Flat points per band. Every order in a band earns the same number of points, for example 20, 60 or 150 points. It is easy to understand, but the jump between bands is large. A $29 order earns 20 points and a $30 order earns 60.
Model B: Whole-order rate. The whole order earns the rate of the band it lands in: 1 point per $1 in Band 1, 1.5 in Band 2 and 2 in Band 3. This feels generous, but there is still a cliff. A $59 order earns 88 points, while a $60 order earns 120.
Model C: Step-by-step rate. Each part of the order earns the rate for its own band, similar to how many tax systems work. The first $30 earns 1 point per $1, the next $30 earns 1.5, and anything above $60 earns 2. A $60 order earns 75 points and a $100 order earns 155.
| Order value | Model B: whole-order rate | Model C: step-by-step rate |
| $29 | 29 points | 29 points |
| $30 | 45 points | 30 points |
| $59 | 88 points | 73 points |
| $60 | 120 points | 75 points |
| $100 | 200 points | 155 points |
Model C has no cliffs, so customers are never surprised that spending $1 more earned 30 extra points. It is also cheaper. Model B is easier to explain on a poster. Many brands choose Model C for the system and describe it simply as “earn more on every dollar over $30”.
Step 4. Check the cost of each band
Every band has a cost, so turn the points into a reward percentage before launch. The effective reward rate is the value of the points divided by the order value.
Using the example retailer’s monthly order mix, here is what each model would cost. The figures are illustrative assumptions, not benchmarks, and they assume every point is eventually redeemed.
| Model | Monthly reward cost | Share of $389,500 monthly revenue |
| Flat 1 point per $1 (no bands) | $3,895 | 1.00% |
| Model B: whole-order rate | $6,055 | 1.55% |
| Model C: step-by-step rate | $4,855 | 1.25% |
Now ask how much extra spending the bands must create to pay for themselves. Model B costs $2,160 a month more than a flat rate. With a 40% gross margin, the retailer needs about $5,400 of extra sales each month to break even. Model C costs $960 more, so it needs about $2,400 of extra sales.
If you expect bands to lift order value by more than that, they are worth testing. If not, lower the top rate, raise the band edges, or use a smaller step.
Step 5. Write the edge-case rules before launch
Bands raise questions a single rate never did. Decide the answers in writing, and make sure every channel applies them the same way.
- Edges: say exactly where each band starts, for example “$30.00 and above”.
- What counts as order value: decide whether you use the amount after discounts and coupons, and whether tax, shipping and gift cards are included.
- Points redeemed on the order: decide whether the band uses the amount before or after points are spent.
- Returns: if part of an order is returned, recalculate the band. A $65 order that becomes $40 after a return moves from Band 3 to Band 2, so some points must be taken back.
- Rounding: decide whether fractional points round down or to the nearest point, and whether you round per order or per band. The examples in this guide round down.
- Split orders: decide whether two orders on the same day count separately, so customers cannot game the bands by combining or splitting baskets.
Read more:
Step 6. Launch, test and adjust
Roll the bands out to a share of members first and keep a similar group on the old flat rate. After a full buying cycle, compare average order value, order frequency and reward cost between the two groups.
Also look at where orders now land. If many orders bunch just above a band edge, the band is working. If orders bunch just below it, the edge may be slightly too high. And if total orders fall while order value rises, customers may be waiting to buy more at once, which can hurt visit frequency.
Read more:
How SupremeTech builds earning rules that hold up
Band rules only work if the system calculates them the same way every time. For a Japanese restaurant corporation, SupremeTech built a loyalty points system on AWS where the POS sends each purchase to a central Point Gateway through an API. The gateway checks and applies coupons, and points are earned, used or canceled in real time, with the result synced back to the mobile app.
That kind of central rules engine is what keeps a band rule, and its return and rounding logic, consistent across stores, apps and online orders.
Conclusion
Assigning points by revenue band comes down to six decisions: what you want to change, where the bands sit, how points are calculated, what it costs, how edge cases work and how you will test it. Start with two or three bands drawn from your own data, prefer a step-by-step rate to avoid cliffs, and check the cost before you promise anything to customers.
If your earning rules need to work the same way across your POS, online store and app, SupremeTech team can help you design and build them.
Frequently Asked Questions
It means grouping orders into ranges of order value and giving each range its own points rate, so larger orders earn more points than smaller ones.
Two or three bands are usually enough. More bands are harder to explain at checkout and harder to test.
A revenue band applies to a single order. A loyalty tier is based on how much a customer spends over a longer period, usually a year, and gives them a status level.
Use a step-by-step rate, where each part of the order earns the rate for its own band. The total then rises smoothly instead of jumping at each edge.
Recalculate the order value after the return. If it falls into a lower band, take back the difference in points according to the rules you published.
