Guide 18 · Tiers and status
How to choose and cost tier benefits
Choose tier benefits from what your best members already buy and what frustrates them, then cost each one per use, including every direct cost of delivering it. Multiply that unit cost by the number of eligible members and the share expected to use it, and test the total against an upper bound. If you have no usage data yet, run a small pilot before you commit the benefit to the whole tier.
Tier benefits are where a program’s promises turn into costs. A benefit that members ignore wastes the budget; one they use far more than expected can wipe out the margin the tier was meant to earn. The work is in two parts: choosing benefits members actually want, and costing them before you commit. This guide covers both.
What members want
What makes a good tier benefit?
A good tier benefit feels worth more to the member than it costs you to deliver, is easy to use and rewards behavior you want to keep. Start from your best members, not from competitors’ benefit lists. Look at your top 20% of members by value: what they buy most often, how they buy, and what they complain about.
Benefits usually fall into three groups:
- Money off: discounts, free products, credits, free shipping. These are hard benefits, with a clear cash value.
- Time saved: priority service, faster delivery, early access, skipping the line.
- Recognition: being known by name, a dedicated contact, invitations, a personal note. These are soft benefits and often cost little.
A strong tier usually has one benefit from each group. Prefer benefits that encourage repeat purchases, such as a free add-on with each order, over one-off gifts that do not change behavior. Avoid benefits that need a code, a manager’s approval or a separate process to use, because friction at the checkout or in the app turns a benefit into a complaint.
Cost one use
How do you cost a single benefit?
Build a cost card for each benefit showing the full direct cost of one use. Include:
- the supplier or wholesale cost of the product or service, not its retail price;
- inbound shipping and per-unit warehousing, if you pay for them;
- packaging and postage to the member;
- payment or processing fees;
- paid staff time, at a loaded hourly rate, for service benefits.
Leave out fixed overheads that do not change with the number of uses, such as rent or salaries you would pay anyway. Take each figure from invoices or price lists, note the date and source, and update the card whenever a supplier price or postage rate changes. This per-use figure is the number that belongs in your benefit budget and your cost of rewards.
Count the cost of delivery, too. A benefit only works if staff and systems can deliver it every time: the point of sale must recognize the member’s tier, the app must show it, and frontline teams must know what each tier gets. If delivering a benefit needs new system work or training, add that one-time cost to the first year’s budget.
Estimate the total
How do you estimate the total cost of a benefit?
Multiply three numbers: the members eligible for the benefit, the cost per use, and the expected uses per member. Then add an upper bound, because the usage estimate is the uncertain part.
In a hypothetical tier of 20,000 members with a $100 credit, a 30% usage rate means a cost of $600,000. If 45% of members use it instead, the cost is $900,000. That 15-point miss adds $300,000 that nobody budgeted. Before you commit, state the most you are willing to spend on each benefit and check the upper bound against it. If the highest plausible usage would break the budget, cap the benefit or change it.
Benefits that scale with activity, such as points multipliers or free items per purchase, need the same treatment with spend in place of members: the more your best members buy, the more the benefit costs. Tie the estimate to your redemption rate data where you have it.
No data yet
What if you have no usage data yet?
Do not build a detailed model from assumptions alone; it will look precise and be wrong. Use the best comparable data you have, such as a previous program, a similar benefit in another market or usage of a related promotion. If there is nothing comparable, run a pilot.
Offer a shortlist of benefits to a sample of eligible members for 60 to 90 days. Ask members to rank the options first, then measure what they actually use, how often, and whether behavior near the tier threshold changes. Ask frontline staff which benefits caused friction. Drop benefits that few eligible members use or that generate repeated support questions, and model the full-scale cost from the pilot’s real usage.
Compare benefits
How do you compare benefits?
Compare each benefit’s perceived value with its cost per use. Perceived value can come from member surveys, rankings in the pilot or the price members pay for similar items elsewhere.
| Benefit type | Example | Cost driver | Usage risk | Notes |
|---|---|---|---|---|
| Money off | Free product or credit each quarter | Product cost per use | High if valuable and easy to use | Cap the number of uses |
| Money off | Points multiplier | Spend of tier members | Grows with spend | Model with your liability |
| Time saved | Free fast shipping | Shipping cost per order | Moderate | Set a minimum order |
| Time saved | Early access to sales | Staff setup per event | Low | Nearly no cost per member |
| Recognition | Dedicated service line | Staff time | Low to moderate | Valued highly by top members |
| Recognition | Birthday reward | Product cost per member | High, most members use it | Easy to budget |
A benefit with modest cash value can beat a generous one if members value it more than it costs. Early access, for example, can cost little more than setup time while being valued highly. Rank the shortlist by value against cost, then keep two or three benefits per tier that differ clearly from the tier below.
Control the cost
How do you keep benefit costs under control after launch?
Set caps on benefits with open-ended use, such as a number of free deliveries or credits per year. Review each benefit’s actual usage and cost every quarter in the first year, then at least yearly. Watch for two warning signs: usage well above the estimate, and a growing number of members in the tier, which multiplies every benefit.
When you add a benefit, cost it at the current number of qualifying members and at the number after any planned threshold change. Adding benefits without raising thresholds is a common path to tier inflation. Thresholds and benefits should be set together; see how to set loyalty tier thresholds.
More answers
Frequently asked questions
What are the best loyalty tier benefits?
The ones your best members value more than they cost you to deliver, and that are easy to use. They usually fall into three groups: money off, time saved and recognition. The right mix depends on what your top customers buy and what frustrates them.
What is the difference between hard and soft benefits?
Hard benefits have a clear cash value, such as free products, discounts or credits. Soft benefits are recognition and convenience, such as priority service, early access or a dedicated line. Soft benefits often cost less but can be valued highly.
Should you use the retail price when costing a benefit?
No. Use the direct cost to you of delivering one use: the supplier or wholesale cost, shipping, packaging, postage, processing fees and paid staff time. Retail price overstates the cost of products you supply yourself.
How many benefits should each tier have?
Few enough to explain in a sentence or two. Two or three clear benefits per tier usually work better than a long list that spreads the budget thin and is hard for staff to deliver.
What is a benefit cap?
A limit on how often a member can use a benefit, such as a number of free deliveries or credits per year. Caps turn an open-ended cost into a known maximum and make the benefit easier to budget.
How often should benefit costs be reviewed?
At least once a year, and whenever a supplier price, shipping rate or the number of qualifying members changes. Usage also drifts as members learn the program, so compare actual use with the original estimate.