Loyalty glossary · 8. Technology and architecture (25)
Rate Limit
Rate limit is a cap on how many points or actions a member can accrue or perform in a set period, enforced by the programme's technology. It protects the system from abuse but also restricts legitimate earning and redemption, so the limit itself is a commercial decision.
Rate limits are commonly framed as a technical safeguard. They are really a pricing decision. They decide how much value a member can extract in a given time. That decision belongs to commercial teams, not just infrastructure engineers.
There are two distinct kinds of rate limit. Infrastructure limits protect servers from being overwhelmed by too many requests. Programme limits cap member earning or redemption actions regardless of server load. Treating the second kind as a technical detail is the most common mistake.
Consider a programme that caps accrual at 1000 points per month. A member earns 1 point per dollar spent. That member reaches the cap after spending exactly 1000 dollars. If the cap is cut to 500 points, the same member reaches it after 500 dollars. The rate limit does not change the earn rate. It changes how much spend is worth chasing.
A flat rate limit is a poor substitute for better design. Accrual rules can vary by tier or by activity, which is what activity-based-qualification does. Active-member-rate measures whether members are actually using the programme, and a rate limit that suppresses active members while leaving inactive ones untouched is a failure. The limit should be set where it stops abuse, not where it stops ordinary use.
The test for any rate limit is whether it blocks abuse without blocking legitimate use. Measure the share of members who hit the cap in a normal month. If that share is above a few percent, the cap is not protecting the programme. It is taxing the best members. The right response is to raise the limit or replace it with a smarter rule, not to ask members to spend less.